Bailey's Industrial Oil and Fat Products, 6 Volume Set

BAILEYS INDUSTRIAL
OIL AND FAT

PRODUCTS
Sixth Edition
Volume 1
Edible Oil and Fat Products:
Chemistry, Properties, and
Health Effects
Edited by
Fereidoon Shahidi
Memorial University of Newfoundland
Baileys Industrial Oil and Fat Products is available online at

http://www.mrw.interscience.wiley.com/biofp
A John Wiley & Sons, Inc., Publication

Copyright # 2005 by John Wiley & Sons, Inc. All rights reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or
by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as
permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior
written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to
the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400, fax
978-646-8600, or on the web at www.copyright.com. Requests to the Publisher for permission should be
addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ
07030, (201) 748-6011, fax (201) 748-6008.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in
preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. No warranty may be created or extended by sales
representatives or written sales materials. The advice and strategies contained herein may not be suitable
for your situation. You should consult with a professional where appropriate. Neither the publisher nor
author shall be liable for any loss of profit or any other commercial damages, including but not limited to
special, incidental, consequential, or other damages.
For general information on our other products and services please contact our Customer Care
Department within the U.S. at 877-762-2974, outside the U.S. at 317-572-3993 or fax 317-572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print,
however, may not be available in electronic format.
Library of Congress Cataloging-in-Publication Data:
Shahidi, Fereidoon.
Baileys industrial oil & fats products. 6th ed./edited by Fereidoon Shahidi.
p. cm.
A Wiley-Interscience publication.
Includes bibliographical references and index.
Contents: v. 1. Edible oil and fat products: chemistry, properties, and health effects v. 2. Edible
oil and fat products: edible oils v.3. Edible oil and fat products: specially oils and oil products v. 4.
Edible oil and fat products: products and applications v. 5. Edible oil and fat products: processing
technologies v. 6. Industrial and nonedible products from oils and fats.
ISBN 0-471-38460-7 (set) ISBN 0-471-38552-2 (v. 1) ISBN 0-471-38551-4 (v. 2)
ISBN 0-471-38550-6 (v. 3) ISBN 0-471-38549-2 (v. 4) ISBN 0-471-38548-4 (v. 5)
ISBN 0-471-38546-8 (v. 6)
1. Oils and fats, I. Title: Industrial oil & fats products. II. Title: Baileys industrial
oil and fats products. III. Bailey, Alton Edward, 1907-1953. IV. Title.
TP670.S46 2004
665dc22 2004043351
Printed in the United States of America

10 9 8 7 6 5 4 3 2 1
Contributors
R. G. ACKMAN: Canadian Institute of Fisheries Technology, Dalhousie University,

Halifax, Nova Scotia, Canada, Fish Oils.
YVONNE T. V. AGUSTIN: Coconut Oil.
KLAUS A. ALEXANDERSEN: Margarine Processing Plants and Equipment.
DAN ANDERSON: A Primer on Oils Processing Technology.
YUSOF BASIRON: Palm Oil.
MARIA LUZ J. BENDANO: Coconut Oil.
ANTHONY P. BIMBO: International Fisheries, Kilmarnock, Virginia, Rendering.
MICHAEL J. BOYER: AWT-Agribusiness and Water, Cumming, Georgia, Environ-
mental Impact and Waste Management.
D. D. BROOKS: Oil-Dri Corporation, Chicago, Illinois, Adsorptive Separation of
Oils.
MICHAEL R. BURKE: Soaps.
ELIAS C. CANAPI: Coconut Oil.
VANCE CAUDILL: Packaging.
ARMAND B. CHRISTOPHE: Ghent University Hospital, Ghent, Belgium, Structural
Effects on Absorption, Metabolism, and Health Effects of Lipids.
MICHAEL M. CHRYSAN: Margarines and Spreads.
W. DE GREYT: De Smet Technologies & Services, Brussels, Belgium, Deodoriza-
tion.
NURHAN TURGUT DUNFORD: Oklahoma State University, Stillwater, Oklahoma, Germ
Oils from Different Sources.
SEVIM Z. ERHAN: National Center for Agricultural Utilization Research, Peoria,
Illinois, Vegetable Oils as Lubricants, Hydraulic Fluids, and Inks.
N.A.M. ESKIN: University of Manitoba, Winnipeg, Manitoba, Canada, Canola Oil.
S. ESWARANANDAM: University of Arkansas, Fayetteville, Arkansas, Edible Films
and Coatings From Soybean and Other Protein Sources.
v
vi CONTRIBUTORS
WALTER E. FARR: Walter E. Farr & Associates, Olive Branch, Mississippi, Hydroge-
nation: Processing Technologies.
DAVID FIRESTONE: United States Food and Drug Administration, Washington, DC,
Olive Oil.
BRENT D. FLICKINGER: Archer Daniels Midland Company, Decatur, Illinois, Diacyl-
glycerols.
GREGORIO C. GERVAJIO: Fatty Acids and Derivatives from Coconut Oil.
MARIA A. GROMPONE: Sunflower Oil.
FRANK D. GUNSTONE: Vegetable Oils.
MONOJ K. GUPTA: MG Edible Oil Consulting International, Richardson, Texas,
Frying of Foods and Snack Food Production; Frying Oils.
OZLEM GUCLU-USTUNDAG: University of Alberta, Edmonton, Alberta, Canada,
Supercritical Technologies for Further Processing of Edible Oils.
MICHAEL J. HAAS: Eastern Regional Research Center, Agricultural Research Ser-
vice, Wyndmoor, Pennsylvania, Animal Fats.
EARL G. HAMMOND: Iowa State University, Ames, Iowa, Soybean Oil.
RICHARD W. HARTEL: University of Wisconsin, Madison, Wisconsin, Crystallization
of Fats and Oils.
BERNHARD HENNIG: University of Kentucky, Lexington, Kentucky, Dietary Lipids
and Health.
ERNESTO HERNANDEZ: Texas A&M University, College Station, Texas, Pharmaceu-
tical and Cosmetic Use of Lipids.
P. B. HERTZ: Agriculture and Agri-Food Canada, Saskatoon, Saskatchewan, Canada,
Vegetable Oils as Biodiesel.
NAVAM S. HETTIARACHCHY: University of Arkansas, Fayetteville, Arkansas, Edible
Films and Coatings From Soybean and Other Protein Sources.
DAVID HETTINGA: Butter.
STEVEN E. HILL: Cooking Oils, Salad Oils, and Dressings.
CHI-TANG HO: Rutgers University, New Brunswick, New Jersey, Flavor Com-
ponents of Fats and Oils.
LUCY SUN HWANG: National Taiwan University, Taipei, Taiwan, Sesame Oil.
LAWRENCE A. JOHNSON: Iowa State University, Ames, Iowa, Soybean Oil.
LYNN A. JONES: Collierville, Tennessee, Cottonseed Oil.
AFAF KAMAL-ELDIN: SLU, Uppsala, Sweden, Minor Components of Fats and Oils.
Y.K. KAMATH: Leather and Textile Uses of Fats and Oils.
RAKESH KAPOOR: Bioriginal Food and Science Corp., Saskatoon, Saskatchewan,
Canada, Conjugated Linoleic Acid Oils; Gamma Linolenic Acid Oils.
M. KELLENS: De Smet Technologies & Services, Brussels, Belgium, Deodorization.
TIMOTHY G. KEMPER: Oil Extraction.
C. CLAY KING: Texas Womens University, Denton, Texas, Cottonseed Oil.
CONTRIBUTORS vii
DAVID D. KITTS: University of British Columbia, Vanuouver, British Columbia,

Canada, Toxicity and Safety of Fats and Oils.
XIAOHUA KONG: Agri-Food Materials Science Centre, University of Alberta Edmonton,
Alberta, Canada, Vegetable Oils in Production of Polymers and Plastics.
S. SEFA KOSEOGLU: Extraction and Refining Program, A Division of Filtration and
Membrane World LLC, College Station, Texas, Membrane Processing of Fats
and Oils.
R. G. KRISHNAMURTHY: Cooking Oils, Salad Oils, and Dressings.
PAUL KRONICK: Leather and Textile Uses of Fats and Oils.
YONG LI: Purdue University, West Lafayette, Indiana, Dietary Lipids and Health.
K. F. LIN: Paints, Varnishes, and Related Products.
LAN LIN: Extraction and Refining Program, A Division of Filtration and Membrane
World LLC, College Station, Texas, Membrane Processing of Fats and Oils.
GARY R. LIST: Iowa State University, Ames, Iowa, Storage, Handling, and Transport
of Oils and Fats.
JERROLD W. LITWINENKO: University of Guelph, Guelph, Ontario, Canada, Fat Crys-
tal Networks.
EDMUND E. LUSAS: Fats and Oils in Feedstuffs and Pet Foods.
JESSE L. LYNN, JR.: Detergents and Detergency.
T. MAG: University of Manitoba, Winnipeg, Manitoba, Canada, Canola Oil.
LINDA J. MALCOLMSON: Canadian International Grains Institute, Winnipeg, Manitoba,
Canada, Flavor and Sensory Aspects.
ALEJANDRO G. MARANGONI: University of Guelph, Guelph, Ontario, Canada, Fat
Crystal Networks.
NOBORU MATSUO: Kao Corporation, Tochigi, Japan, Diacylglycerols.
W. W. MCCALLEY: Agriculture and Agri-Food Canada, Saskatoon, Saskatchewan,
Canada, Vegetable Oils as Biodiesel.
D. JULIAN MCCLEMENTS: The University of Massachusetts, Amherst, Massachusetts,
Lipid Emulsions.
B.E. MCDONALD: University of Manitoba, Winnipeg, Manitoba, Canada, Canola
Oil.
THOMAS A. MCKeon, USDA-ARS Western Regional Research Center, Albany,
California, Transgenic Oils.
SERPIL METIN: Cargill Inc., Minneapolis, Minnesota, Crystallization of Fats and Oils.
DOUGLAS J. METZROTH: Shortenings: Science and Technology.
HOMAN MIRALIAKBARI: Memorial University of Newfoundland, St. Johns,
Newfoundland, Canada, Tree Nut Oils.
ROBERT A. MOREAU: United States Department of Agriculture, Agricultural
Research Service, Corn Oil.
EVANGEKUBE A. MORO: Coconut Oil.
viii CONTRIBUTORS
HARIKUMAR NAIR: Bioriginal Food & Science Corp., Saskatoon, Saskatchewan,

Canada, Gamma Linolenic Acid Oils.
SURESH S. NARINE: Agri-Food Materials Science Centre, University of Alberta,
Edmonton, Alberta, Canada, Vegetable Oils in Production of Polymers and
Plastics.
RICHARD D. OBRIEN: Plano, Texas, Cottonseed Oil; Shortenings: Types and Formu-
lations.
FRANK T. ORTHOEFER: Rice Bran Oil.
JOHN W PARRY: University of Maryland, College Park, Maryland, Oils from Herbs,
Spices, and Fruit Seeds.
HAROLD E. PATTEE: North Carolina State University, Raleigh, North Carolina,
Peanut Oil.
ECONOMICO PEDROSA, JR.: Coconut Oil.
M. D. PICKARD: By-Product Utilization.
A. PROCTOR: University of Arkansas, Fayetteville, Arkansas, Adsorptive Separation
of Oils.
ROMAN PRZYBYLSKI: University of Manitoba, Winnipeg, Manitoba, Canada, Canola
Oil; Flax Oil and High Linolenic Oils.
COLIN RATLEDGE: Lipid Research Centre, University of Hull, Hull, United Kingdom,
Oils from Microorganisms.
MARTIN REANEY: Bioriginal Food and Science Corp., Saskatoon, Saskatchewan,
Canada, Conjugated Linoleic Acid Oils.
M. J. T. REANEY: Agriculture and Agri-Food Canada, Saskatoon, Saskatchewan,
Canada, Vegetable Oils as Biodiesel.
MIAN N. RIAZ: Texas A&M University, College Station, Texas, Extrusion Proces-
sing of Oilseed Meals for Food and Feed Production.
GEOFFREY G. RYE: University of Guelph, Guelph, Ontario, Canada, Fat Crystal
Networks.
KIYOTAKA SATO: Graduate School of Biosphere Science, Hiroshima University,
Higashi-Hiroshima, Japan, Polymorphism in Fats and Oils.
K. M. SCHAICH: Rutgers University, New Brunswick, New Jersey, Lipid Oxidation:
Theoretical Aspects.
KEITH SCHROEDER: CC Engineering Ltd., Glycerine.
CHARLIE SCRIMGEOUR: Scottish Crop Research Institute Dundee, Scotland, Chemistry
of Fatty Acids.
S. P. J. NAMAL SENANAYAKE: Martek Biosciences Corporation, Winchester, Kentucky,
Dietary Fat Substitutes; Modification of Fats and Oils via Chemical and Enzy-
matic Methods.
FEREIDOON SHAHIDI: Memorial University of Newfoundland, St. Johns,
Newfoundland, Canada, Antioxidants: Regulatory Status; Antioxidants: Science,
Technology, and Applications; Citrus Oils and Essences; Dietary Fat Substitutes;
CONTRIBUTORS ix
Flavor Components of Fats and Oils; Lipid Oxidation: Measurement Methods;

Marine Mammal Oils; Modification of Fats and Oils via Chemical and Enzy-
matic Methods; Novel Separation Techniques for Isolation and Purification of
Fatty Acids and Oil By-Products; Quality Assurance of Fats and Oils; Tree
Nut Oils.
JOSEPH SMITH: Safflower Oil.
VIJAI K.S. SHUKLA: International Food Science Center, Lystrup, Denmark, Con-
fectionery Lipids.
VIJAI K.S. SHUKLA: Iowa State University, Ames, Iowa, Storage, Handling, and
Transport of Oils and Fats.
CLYDE E. STAUFFER: Emulsifiers for the Food Industry; Fats and Oils in Bakery
Products.
CAIPING SU: Iowa State University, Ames, Iowa, Soybean Oil.
BERNARD F. SZUHAJ: Szuhaj & Associates LLC, Fort Wayne, Indiana, Lecithins.
DENNIS R. TAYLOR: DR Taylor Consulting, Port Barrington, Illinois, Bleaching.
FERAL TEMELLI: University of Alberta, Edmonton, Alberta, Canada, Supercritical
Technologies for Further Processing of Edible Oils.
MICHAL TOBOREK: University of Kentucky, Lexington, Kentucky, Dietary Lipids and
Health.
SATORU UENO: Graduate School of Biosphere Science, Hiroshima University,
Higashi-Hiroshima, Japan, Polymorphism in Fats and Oils.
PHILLIP J. WAKELYN: National Cotton Council, Washington, DC, Cottonseed Oil.
PETER J. WAN: USDA, ARS, New Orleans, Lowsiana, Cottonseed Oil.
P. K. J. P. D. WANASUNDARA: Agriculture and Agri-Food Canada, Saskatoon
Research Center, Saskatoon, Saskatchewan, Canada, Antioxidants: Science,
Technology, and Applications; Novel Separation Techniques for Isolation and
Purification of Fatty Acids and Oil By-Products.
UDAYA N. WANASUNDARA: POS Pilot Plant Corporation, Saskatoon, Saskatchewan,
Canada, Novel Separation Techniques for Isolation and Purification of Fatty
Acids and Oil By-Products.
TONG WANG: Iowa State University, Ames, Iowa, Soybean Oil; Storage, Handling,
and Transport of Oils and Fats.
BRUCE A. WATKINS: Purdue University, West Lafayette, Indiana, Dietary Lipids and
Health.
JOCHEN WEISS: The University of Massachusetts, Amherst, Massachusetts, Lipid
Emulsions.
NEIL D. WESTCOTT: Bioriginal Food and Science Corp., Saskatoon, Saskatchewan,
Canada, Conjugated Linoleic Acid Oils.
PAMELA J. WHITE: Iowa State University, Ames, Iowa, Soybean Oil.
MAURICE A. WILLIAMS: Anderson Corporation, Cleveland, Ohio, Recovery of Oils
and Fats from Oilseeds and Fatty Materials.
x CONTRIBUTORS
JAMES P. WYNN: Martek Biosciences Corporation, Columbia, Maryland, Oils from

Microorganisms.
LIANGLI (LUCY) YU: University of Maryland, College Park, Maryland, Oils from
Herbs, Spices, and Fruit Seeds.
YING ZHONG: Memorial University of Newfoundland, St. Johns, Newfoundland,
Canada, Antioxidants: Regulatory Status; Citrus Oils and Essences; Lipid
Oxidation: Measurement Methods; Marine Mammal Oils.
KEQUAN ZHOU: University of Maryland, College Park, Maryland, Oils from Herbs,
Spices, and Fruit Seeds.
Preface
Oils and fats are important components of foods, and they, or their derivatives and
products thereof, play an important role in non-food applications. In food, oils and
fats provide a concentrated source of energy as well as a carrier of fat-soluble com-
ponents. They also serve as a heat transfer medium for food processing and render
desirable texture and flavor as well as mouthfeel to products. Oils and fats originate
from plant and animal sources. Although plant sources include oilseeds, tropical
fruits, and alga, the latter may originate from land-based animals, fish, marine
mammals, and derived sources. The main components of food lipids are triacylgly-
cerols, but minor components are also important for quality characteristics, stabi-
lity, and application areas. Both the type of fatty acids and their degree of
unsaturation as well as the type and content of minor components affect the keeping
quality of the oil, and certain minor components such as phytosterols might also be
used for fingerprinting and authentification of the source materials.
The physical state of fats and oils and their crystal structures are important for
application of such products. In addition, formulation of products for special appli-
cations such as bakery, confectionary, frying, salad dressing, margarines, and
spreads requires special characteristics that make the products suitable for such pur-
poses. Thus, each source material will be important for its physical and chemical
characteristics and hence suitability as a food component.
Recent developments in the area of oils and fats has led to the production of spe-
cialty lipids from novel sources such as fruit seeds, nuts, and other minor plant
sources. In addition, preparation of structured lipids for a myriad of applications
has been of interest. Minor components of oils and fats may be isolated during pro-
cessing and used as nutraceutical and functional food ingredients. Examples are
lecithin, phytosterols, tocopherols, and tocotrienols, among others. Obviously, the
health-promoting potential of such products is also of interest.
The processing technologies employed for production of fats and oils, and asso-
ciated components, to make them shelf-stable with acceptable sensory characteris-
tics and flavor as well as secondary processing technologies for production of
specific products are important considerations in this area. Food commodities
xi
xii PREFACE
may be produced, and some components may also be used in animal feed and other
applications. There are many areas where oils and fats are used for non-food pur-
poses. Thus, detergents, soaps, glycerine and polymers, inks, lubricants, and biodie-
sel may be derived from fatty acids and their derivatives. Many applications would
provide alternatives to the use of synthetic material or environmentally friendly
substitutes in non-food applications.
The sixth edition of Bailey provides a comprehensive description of topics rele-
vant to the oils and fats industry in six volumes as compared with five volumes in
the fifth edition. The additional volume (volume 3) is mainly on specialty oils and
fats and their byproducts or minor components as well as on those of low-calorie fat
substitutes and structured lipids. An article on fish oils and one on marine mammal
oils are also included in this volume. However, the material covered in other
volumes is often substantially different from the available in the fifth edition as
new articles are introduced, and when the title appears the same, substantial
updating of the references and introduction of new material has occurred; new
authors in some cases have made these contributions. Thus, the first volume
includes three new articles on crystallization and physical properties of oils and
fats. There are also new articles on antioxidant theory and regulatory status as
well as on mechanisms and measurements of lipid oxidation. A new article has
been introduced on quality assurance of oils and fats. Meanwhile, the second
volume presents the main sources of food lipids, and new articles on sesame oil
and rice bran oil have been introduced. The fourth volume provides a description
of application areas, and here again new articles on confectinary lipids as well as on
frying oils and snack food production have been added. The fifth volume on proces-
sing technologies introduces new articles on supercritical, membrane, and extrusion
technologies. Finally, the sixth volume on nonedible uses of fats and oils has new
articles on biodiesel, hydrolic fluids, lubricants, inks, as well as pharmaceutical and
cosmetic uses of lipids. An article on the use of soybean oil in edible film and adhe-
sive production is also included. Thus, the sixth edition is substantially different
from what was available in the fifth edition.
I am indebted to many authors for their state-of-the-art contributions as well as
to primary and secondary reviewers for different articles. The advisory committee
members served an important role in providing invaluable comments. In addition,
staff from John Wiley and Sons provided considerable help in different aspects
related to production and assembly of the work. This series serves as a primary
source of and as a compendium of information on oils and fats for the industry,
academia and government scientists, and technical personnel, and as a reference
for senior undergraduate and graduate students in food science, nutrition, dietetics,
biochemistry, and related disciplines. An integrated table of contents allows better
search of materials of interest, and the last volume has a cumulative index. Exten-
sive bibliography throughout the series also provides the reader with the opportu-
nity to consult primary references for additional information.
FEREIDOON SHAHIDI
Contents
Contributors ........................................................................ v
Preface ............................................................................... xi
Volume 1. Edible Oil and Fat Products: Chemistry,

Properties, and Health Effects ......................... 1: I
1.1 Chemistry of Fatty Acids ...................................................... 1:1
1.1.1 Introduction ..................................................... 1:1
1.1.2 Composition and Structure .............................. 1:2
1.1.3 Hydrolysis, Esterification, and Ester
Exchange ........................................................ 1:10
1.1.4 Oxidation ......................................................... 1:15
1.1.5 Reduction ........................................................ 1:25
1.1.6 Production of Surface Active Compounds
and Oleochemicals .......................................... 1:27
1.1.7 Modifying Fatty Acid Structure ......................... 1:32
1.1.8 Novel Chemistry for Functionalizing the Alkyl
Chain ............................................................... 1:36
References .................................................................... 1:39
1.2 Crystallization of Fats and Oils ............................................ 1:45
1.2.1 Introduction ..................................................... 1:45
1.2.2 Lipid Phase Behavior ....................................... 1:46
1.2.3 Crystallization Behavior ................................... 1:57
1.2.4 Controlling Crystallization ................................ 1:68
1.2.5 Summary ......................................................... 1:73
References .................................................................... 1:73
xiii
xiv Contents
1.3 Polymorphism in Fats and Oils ............................................ 1:77

1.3.1 Introduction ..................................................... 1:77
1.3.2 Basic Concepts of Polymorphism of Fats ........ 1:78
1.3.3 Polymorphism of Monoacid Triacylglycerols .... 1:85
1.3.4 Polymorphism of Mixed-acid
Triacylglycerols ................................................ 1:90
1.3.5 Fat Mixtures and Polymorphism ...................... 1:100
1.3.6 Polymorphism of Natural Fats ......................... 1:108
1.3.7 Summary ......................................................... 1:115
References .................................................................... 1:116
1.4 Fat Crystal Networks ............................................................ 1:121
1.4.1 Introduction ..................................................... 1:121
1.4.2 Mechanical Properties of Milkfat ...................... 1:122
1.4.3 Lipid Composition ............................................ 1:123
1.4.4 Processing Conditions ..................................... 1:125
1.4.5 Nucleation and Crystal Growth ........................ 1:126
1.4.6 Mechanical Properties ..................................... 1:148
1.4.7 Assessing the Validity of the Model:
Correlating Experimentally Determined
Parameters ...................................................... 1:154
Acknowledgments .......................................................... 1:158
References .................................................................... 1:158
1.5 Animal Fats ........................................................................... 1:161
1.5.1 Introduction ..................................................... 1:161
1.5.2 Sources, Fatty Acid Content, and
Acylglycerol Structure ...................................... 1:162
1.5.3 Acylglycerol Structure and Its Relationship to
Functionality and Use ...................................... 1:168
1.5.4 Quality Indicators for Edible Fats ..................... 1:171
1.5.5 Regulatory and Commercial Classifications
of Animal Fats ................................................. 1:174
Contents xv
1.5.6 Patterns and Trends in the Production and

Use of Animal Fats .......................................... 1:179
1.5.7 Processing of Animal Fats ............................... 1:182
1.5.8 Antioxidants in Animal Fats ............................. 1:194
1.5.9 Characteristics of Animal Fat-based
Shortenings and Frying Fats ............................ 1:195
1.5.10 Recent Developments ..................................... 1:197
References .................................................................... 1:205
1.6 Vegetable Oils ...................................................................... 1:213
1.6.1 Introduction ..................................................... 1:213
1.6.2 Biosynthesis .................................................... 1:213
1.6.3 Minor Components .......................................... 1:217
1.6.4 Classification of Vegetable Oils ....................... 1:219
1.6.5 The Major Vegetable Oils and Fats ................. 1:224
1.6.6 Speciality and Minor Oils ................................. 1:232
1.6.7 Modification of Oils and Fats ........................... 1:242
1.6.8 Technological Procedures Used for Lipid
Modification ..................................................... 1:244
1.6.9 Biological Methods of Lipid Modification .......... 1:251
1.6.10 Production and Trade Statistics ....................... 1:258
1.6.11 Conclusion ....................................................... 1:259
References .................................................................... 1:259
1.7 Lipid Oxidation: Theoretical Aspects ................................... 1:269
1.7.1 Introduction ..................................................... 1:269
1.7.2 Initiation (LH L ) ........................................ 1:273
1.7.3 Propagation ..................................................... 1:304
1.7.4 Termination ..................................................... 1:333
1.7.5 Expanded Integrated Reaction Scheme .......... 1:341
References .................................................................... 1:343
xvi Contents
1.8 Lipid Oxidation: Measurement Methods .............................. 1:357

1.8.1 Introduction ..................................................... 1:357
1.8.2 Methods for Measuring Lipid Oxidation ............ 1:358
1.8.3 Measurement of Oxygen Absorption ................ 1:359
1.8.4 Measurement of Reactant Change .................. 1:360
1.8.5 Measurement of Primary Products of
Oxidation ......................................................... 1:361
1.8.6 Measurement of Secondary Products of
Oxidation ......................................................... 1:366
1.8.7 Measurement of Free Radicals ........................ 1:373
1.8.8 Other Methods ................................................. 1:374
1.8.9 Measurement of Frying Fat Deterioration ........ 1:377
1.8.10 Methods for Measuring Antioxidant Activity ..... 1:378
1.8.11 Conclusions and Recommendations ................ 1:380
References .................................................................... 1:380
1.9 Flavor Components of Fats and Oils ................................... 1:387
1.9.1 Free Radical Autoxidation of Lipids ................. 1:387
1.9.2 Hydroperoxides of Fatty Acids or Their
Esters .............................................................. 1:388
1.9.3 Major Volatile Compounds of Commercial
Fats and Oils ................................................... 1:395
References .................................................................... 1:408
1.10 Flavor and Sensory Aspects ................................................ 1:413
1.10.1 Introduction ..................................................... 1:413
1.10.2 Sensory Methods ............................................. 1:413
1.10.3 Factors Affecting Sensory Measurements ....... 1:415
1.10.4 Experimental Design and Statistical
Analysis ........................................................... 1:416
1.10.5 Sensory Testing Facility .................................. 1:416
1.10.6 Sample Preparation and Presentation ............. 1:418
1.10.7 Reference Samples ......................................... 1:420
Contents xvii
1.10.8 Selection and Training of Panelists .................. 1:422

1.10.9 Monitoring and Motivation of Panelists ............ 1:423
1.10.10 Sensory Evaluation of Oils ............................... 1:423
1.10.11 Sensory Evaluation of Oil-containing Foods .... 1:426
1.10.12 Sensory Evaluation of Frying Oils/Room
Odor ................................................................ 1:426
1.10.13 Sensory Evaluation of Fried Foods .................. 1:426
1.10.14 Electronic Nose ............................................... 1:427
1.10.15 Gas Chromatography-olfactometery ................ 1:427
1.10.16 Conclusions ..................................................... 1:427
References .................................................................... 1:428
1.11 Antioxidants: Science, Technology, and Applications ......... 1:431
1.11.1 An Antioxidant Definition ............................... 1:431
1.11.2 History of Antioxidants and Their Use .............. 1:432
1.11.3 Scope of Using Antioxidants in Food ............... 1:433
1.11.4 Oxidation of Fats and Oils and Mechanism of
Antioxidants ..................................................... 1:434
1.11.5 Classification of Antioxidants ........................... 1:436
1.11.6 Evaluation of Antioxidant Activity ..................... 1:445
1.11.7 Commonly Used Antioxidants in Foods ........... 1:455
1.11.8 Estimation and Analysis of Antioxidants in
Foods .............................................................. 1:474
1.11.9 Technological Considerations in Using
Antioxidants ..................................................... 1:474
1.11.10 Regulatory Status and Safety Issues of
Synthetic and Natural Antioxidants .................. 1:475
1.11.11 Safety Considerations of Antioxidants Used
in Foods .......................................................... 1:476
References .................................................................... 1:483
1.12 Antioxidants: Regulatory Status ........................................... 1:491
1.12.1 Introduction ..................................................... 1:491
1.12.2 Synthetic Antioxidants ..................................... 1:492
xviii Contents
1.12.3 Natural Antioxidants ........................................ 1:503

1.12.4 Conclusions ..................................................... 1:508
References .................................................................... 1:509
1.13 Toxicity and Safety of Fats and Oils .................................... 1:513
1.13.1 Introduction ..................................................... 1:513
1.13.2 Adverse Effects of Fats and Associated
Constituents .................................................... 1:515
1.13.3 Adverse Effects of Some Natural
Constituents in Fats and Oils ........................... 1:530
1.13.4 Bioactive Lipid-soluble Constituents ................ 1:533
1.13.5 Chemical Reactions in Fats ............................. 1:535
1.13.6 Adverse Products from Overheated Fats and
Oils .................................................................. 1:543
1.13.7 Toxic Substances Produced during Smoking,
Charbroiling, and Barbecuing of Foods ............ 1:545
1.13.8 Potential Hazards from Government
Approved Antioxidants ..................................... 1:547
1.13.9 Manufacturing Hazards in Processing Crude
Oils and Fats ................................................... 1:550
1.13.10 Conclusions ..................................................... 1:552
References .................................................................... 1:552
1.14 Quality Assurance of Fats and Oils ..................................... 1:565
1.14.1 Introduction ..................................................... 1:565
1.14.2 Oil Composition ............................................... 1:568
1.14.3 Minor Components .......................................... 1:570
1.14.4 Unsaponfiables Matter ..................................... 1:570
1.14.5 Characteristics of Fats and Oils ....................... 1:571
1.14.6 Color and Appearance ..................................... 1:571
1.14.7 Oxidative Quality and Stability Tests ............... 1:572
1.14.8 Carbonyl Compounds ...................................... 1:573
1.14.9 Polymers and Polar Components .................... 1:573
1.14.10 Antioxidants ..................................................... 1:574
Contents xix
1.14.11 Adulteration ..................................................... 1:574

1.14.12 Pollutants ........................................................ 1:574
References .................................................................... 1:574
1.15 Dietary Lipids and Health ..................................................... 1:577
1.15.1 Introduction ..................................................... 1:577
1.15.2 PUFA Biochemistry ......................................... 1:582
1.15.3 Molecular Actions ............................................ 1:583
1.15.4 Fat and Chronic Diseases ............................... 1:586
1.15.5 Role of Dietary Fat in Cardiovascular
Disease and Atherosclerosis ........................... 1:589
References .................................................................... 1:600
Volume 2. Edible Oil and Fat Products: Edible Oils.. ......... 2: I

2.1 Butter .................................................................................... 2:1
2.1.1 Introduction ..................................................... 2:1
2.1.2 Chemical Composition ..................................... 2:2
2.1.3 Modification of Milkfat ...................................... 2:12
2.1.4 Quality Control ................................................. 2:21
2.1.5 Butter Manufacture .......................................... 2:26
2.1.6 Butter Fat Products .......................................... 2:45
2.1.7 Economics ....................................................... 2:52
References .................................................................... 2:55
2.2 Canola Oil ............................................................................. 2:61
2.2.1 Introduction ..................................................... 2:61
2.2.2 Origin ............................................................... 2:62
2.2.3 Development of Canola ................................... 2:62
2.2.4 Composition .................................................... 2:63
2.2.5 Physical Properties .......................................... 2:74
2.2.6 Canola Oil Extraction and Processing .............. 2:76
2.2.7 Nutritional Properties of Canola Oil .................. 2:93
2.2.8 Major Food Uses ............................................. 2:99
xx Contents
2.2.9 Nonfood Uses of Standard Canola Oil ............. 2:109

2.2.10 Production of Oilseeds and Oils ....................... 2:113
References .................................................................... 2:116
2.3 Coconut Oil ........................................................................... 2:123
2.3.1 The Coconut Palm ........................................... 2:123
2.3.2 The Fruit .......................................................... 2:126
2.3.3 Copra .............................................................. 2:128
2.3.4 Oil Extraction ................................................... 2:129
2.3.5 Refining ........................................................... 2:132
2.3.6 Coconut Oil Composition ................................. 2:135
2.3.7 Chemical and Physical Tests ........................... 2:137
2.3.8 Uses ................................................................ 2:141
2.3.9 Storage ............................................................ 2:142
2.3.10 Economics ....................................................... 2:143
References .................................................................... 2:146
2.4 Corn Oil ................................................................................ 2:149
2.4.1 Overview ......................................................... 2:149
2.4.2 Extraction and Refining ................................... 2:151
2.4.3 Composition .................................................... 2:155
2.4.4 Properties of Corn Oil ...................................... 2:165
2.4.5 Major Food Uses of Corn Oil ........................... 2:167
2.4.6 Nonfood Uses of Corn Oil ................................ 2:168
2.4.7 Conclusions ..................................................... 2:168
References .................................................................... 2:169
2.5 Cottonseed Oil ...................................................................... 2:173
2.5.1 Introduction ..................................................... 2:173
2.5.2 Cottonseed Oil Industry Development ............. 2:174
2.5.3 Cottonseed Oil Properties ................................ 2:185
2.5.4 Cottonseed Handling, Oil Extraction and
Processing ....................................................... 2:207
Contents xxi
2.5.5 Regulatory Considerations: Cottonseed Oil

Extraction and Processing ............................... 2:235
2.5.6 Cottonseed Oil Products Finishing
Treatment ........................................................ 2:242
2.5.7 Cottonseed Oil Utilization ................................ 2:246
2.5.8 Liquid Oils ....................................................... 2:248
2.5.9 Shortenings ..................................................... 2:256
2.5.10 Margarine and Spreads ................................... 2:267
2.5.11 Other Cottonseed Oil Uses .............................. 2:272
References .................................................................... 2:272
2.6 Flax Oil and High Linolenic Oils ........................................... 2:281
2.6.1 Introduction ..................................................... 2:281
2.6.2 Flax ................................................................. 2:282
2.6.3 Perilla Oil ......................................................... 2:292
2.6.4 Camelina ......................................................... 2:294
2.6.5 Chia ................................................................. 2:298
References .................................................................... 2:298
2.7 Olive Oil ................................................................................ 2:303
2.7.1 Introduction and History ................................... 2:303
2.7.2 Statistics and Definitions ................................. 2:306
2.7.3 Extraction Technology ..................................... 2:310
2.7.4 Refining of Olive Oils ....................................... 2:314
2.7.5 Refining of Pomace Oil .................................... 2:316
2.7.6 Olive Oil Components ...................................... 2:317
2.7.7 Analysis of Olive Oils ....................................... 2:320
References .................................................................... 2:328
2.8 Palm Oil ................................................................................ 2:333
2.8.1 Introduction ..................................................... 2:333
2.8.2 Chemical and Physical Properties of Palm
Oil .................................................................... 2:339
2.8.3 Production Process ......................................... 2:351
xxii Contents
2.8.4 Refining and Fractionation ............................... 2:371

2.8.5 End Uses ......................................................... 2:388
2.8.6 Potential Developments ................................... 2:410
2.8.7 Nutritional Effects of Palm Oil .......................... 2:411
2.8.8 Prospects of Palm Oil and Market
Requirements .................................................. 2:417
References .................................................................... 2:423
2.9 Peanut Oil ............................................................................. 2:431
2.9.1 Peanut Origin and History ................................ 2:431
2.9.2 Global .............................................................. 2:432
2.9.3 Environmental and Genotype Effects on the
Composition Peanuts ...................................... 2:445
2.9.4 Modification of Oil Characteristics through
Breeding .......................................................... 2:445
2.9.5 Oil Color .......................................................... 2:446
2.9.6 Peanut Oil Evaluation and Composition ........... 2:447
2.9.7 Uses ................................................................ 2:453
2.9.8 Dietary Aspects ............................................... 2:454
2.9.9 Allergenicity ..................................................... 2:455
References .................................................................... 2:455
2.10 Rice Bran Oil ........................................................................ 2:465
2.10.1 Introduction ..................................................... 2:465
2.10.2 Composition of Rice and Rice Bran Lipids ....... 2:466
2.10.3 Milling of Rice .................................................. 2:470
2.10.4 Enzymes in Rice Bran ..................................... 2:473
2.10.5 Stabilization of Rice Bran ................................ 2:475
2.10.6 Rice Bran to Rice Bran Oil ............................... 2:477
2.10.7 Refining of the Oil ............................................ 2:478
2.10.8 Dewaxing ........................................................ 2:479
2.10.9 Degumming and Deacidification ...................... 2:479
Contents xxiii
2.10.10 Bleaching, Hydrogenation and

Deodorerization ............................................... 2:480
2.10.11 Winterization .................................................... 2:481
2.10.12 Co-products from Processing .......................... 2:481
2.10.13 Composition of Refined Rice Bran Oil .............. 2:482
2.10.14 Rice Bran Oil Nutrition ..................................... 2:483
2.10.15 Rice Bran Oil Utilization ................................... 2:485
2.10.16 Rice Oil Production (Potential) ......................... 2:486
2.10.17 Summary ......................................................... 2:487
References .................................................................... 2:487
2.11 Safflower Oil ......................................................................... 2:491
2.11.1 History and Botanical Description .................... 2:491
2.11.2 Physical and Chemical Properties ................... 2:505
2.11.3 Processing ....................................................... 2:511
2.11.4 Economics and Marketing ............................... 2:514
2.11.5 Quality Assessment ......................................... 2:522
2.11.6 Storage and Transportation ............................. 2:525
2.11.7 Unique Uses .................................................... 2:527
References .................................................................... 2:530
2.12 Sesame Oil ........................................................................... 2:537
2.12.1 Introduction ..................................................... 2:537
2.12.2 Botany of Sesame ........................................... 2:538
2.12.3 World Production ............................................. 2:541
2.12.5 Sesame Lignans and Lignan Glycosides ......... 2:549
2.12.6 Processing ....................................................... 2:555
2.12.7 Nutritional Characteristics ................................ 2:564
Reference ...................................................................... 2:570
2.13 Soybean Oil .......................................................................... 2:577
2.13.1 Introduction ..................................................... 2:577
2.13.2 Composition of Soybeans ................................ 2:578
xxiv Contents
2.13.3 Physical Properties of Soybean Oil .................. 2:584

2.13.4 Grading ........................................................... 2:589
2.13.5 Recovery of Oil from Soybeans ....................... 2:591
2.13.6 Qualities of Soybean Oils and Meals
Extracted by Different Methods ........................ 2:600
2.13.7 Soy Protein Ingredients ................................... 2:603
2.13.8 Basic Processing Operations ........................... 2:604
2.13.9 Alternative Refining Methods ........................... 2:611
2.13.10 Coproducts and Utilization ............................... 2:612
2.13.11 Food and Biobased Product Uses of
Soybean Oil ..................................................... 2:614
2.13.12 Oxidative Quality of Soybean Oil ..................... 2:629
2.13.13 Dietary Fatty Acids and Their Health Effects ... 2:638
References .................................................................... 2:641
2.14 Sunflower Oil ........................................................................ 2:655
2.14.1 Historical Review ............................................. 2:655
2.14.2 Sunflower Crops .............................................. 2:658
2.14.3 Chemical and Physical Properties of Regular
Sunflower Oil ................................................... 2:664
2.14.4 Sunflower Seed of Modified Fatty Acid
Composition .................................................... 2:674
2.14.5 Extraction and Processing of Sunflower Oil ..... 2:685
2.14.6 Hydrogenation of Regular Sunflower Oil .......... 2:700
2.14.7 Storage and Deterioration of Sunflower Oil ...... 2:703
2.14.8 Uses of Sunflower Oil ...................................... 2:707
2.14.9 World Production and Distribution of
Sunflower Oil ................................................... 2:713
2.14.10 Sunflower Oil Extraction and Processing
by-products ...................................................... 2:719
References .................................................................... 2:725
Contents xxv
Volume 3. Edible Oil and Fat Products: Specialty

Oils and Oil Products ......................................... 3: I
3.1 Conjugated Linoleic Acid Oils .............................................. 3:1
3.1.1 Introduction ..................................................... 3:1
3.1.2 Metabolism ...................................................... 3:2
3.1.3 Physiological Actions of CLA ........................... 3:3
3.1.4 Strategies to Increase Dietary Intake of CLA ... 3:8
3.1.5 Commercial Production of CLA ........................ 3:8
3.1.6 Analysis of Conjugated Linoleic Acids ............. 3:21
3.1.7 Conclusions ..................................................... 3:29
References .................................................................... 3:29
3.2 Diacylglycerols ..................................................................... 3:37
3.2.1 Introduction ..................................................... 3:37
3.2.2 Comparison of DAG Oil vs. TAG Oil ................ 3:39
3.2.3 Summary ......................................................... 3:44
References .................................................................... 3:46
3.3 Citrus Oils and Essences ..................................................... 3:49
3.3.1 Introduction ..................................................... 3:49
3.3.2 Oil Extraction ................................................... 3:50
3.3.4 Storage of Citrus Oils ...................................... 3:58
3.3.5 Applications of Citrus Oils and Essences ......... 3:61
3.3.6 Challenges ...................................................... 3:62
3.3.7 Conclusions ..................................................... 3:63
References .................................................................... 3:63
3.4 Gamma Linolenic Acid Oils .................................................. 3:67
3.4.1 Introduction ..................................................... 3:67
3.4.2 Sources of GLA ............................................... 3:68
3.4.3 Extraction of Oil ............................................... 3:76
3.4.4 Metabolism of GLA .......................................... 3:80
3.4.5 Cardiovascular Effects ..................................... 3:83
xxvi Contents
3.4.6 Cancer ............................................................. 3:87

3.4.7 Immune Function and Autoimmune
Diseases ......................................................... 3:93
3.4.8 Skin Conditions ............................................... 3:99
3.4.9 Diabetes .......................................................... 3:103
3.4.10 Premenstrual Syndrome (PMS) ....................... 3:106
3.4.11 Infant Nutrition and Development .................... 3:107
3.4.12 Drug/Nutrient Interactions ................................ 3:108
3.4.13 Safety of GLA-containing Oils .......................... 3:109
3.4.14 Current Research Focus .................................. 3:110
References .................................................................... 3:111
3.5 Oils from Microorganisms .................................................... 3:121
3.5.1 General Introduction and Background
Information ...................................................... 3:121
3.5.2 Commercial Microbial Oils ............................... 3:130
3.5.3 SCOs in Current (2003) Production ................. 3:137
3.5.4 Prospects for Production of Other PUFAs by
Microorganisms ............................................... 3:145
3.5.5 The Future of Microbial Oils ............................. 3:147
References .................................................................... 3:150
3.6 Transgenic Oils .................................................................... 3:155
3.6.1 Introduction ..................................................... 3:155
3.6.2 Technology for Altering Fatty Acid
Composition .................................................... 3:158
3.6.3 Canola from Traditional Breeding of Oilseed
Crops ............................................................... 3:159
3.6.4 High-oleic Sunflower from Mutagenesis of
Oilseed Crops .................................................. 3:160
3.6.5 Applications of High-oleate Oils ....................... 3:160
3.6.6 Altered Polyunsaturate Content through
Mutagenesis .................................................... 3:162
Contents xxvii
3.6.7 Improved Oil Composition of a Transgenic

Soybean .......................................................... 3:162
3.6.8 Improved Industrial Use from Genetically
Engineering Oilseed Crops .............................. 3:163
3.6.9 Future Directions for Transgenic Oilseeds ....... 3:164
3.6.10 Potential New Oils for Food, Feed, and
Industruial Use ................................................ 3:165
3.6.11 Issues Related to Transgenic Oilseeds ............ 3:167
References .................................................................... 3:172
3.7 Tree Nut Oils ........................................................................ 3:175
3.7.1 Introduction ..................................................... 3:175
3.7.2 Almond ............................................................ 3:176
3.7.3 Hazelnut .......................................................... 3:179
3.7.4 Pecan .............................................................. 3:182
3.7.5 Walnut ............................................................. 3:183
3.7.6 Pistachio .......................................................... 3:185
3.7.7 Brazil Nut ......................................................... 3:186
3.7.8 Pine Nut .......................................................... 3:186
3.7.9 Macadamia Nut ............................................... 3:187
3.7.10 Cashew Nut ..................................................... 3:188
3.7.11 Use of Defatted Tree Nut Meals and Other
Byproducts as Protein Sources ........................ 3:188
3.7.12 Concluding Remarks ....................................... 3:190
References .................................................................... 3:190
3.8 Germ Oils from Different Sources ........................................ 3:195
3.8.1 Introduction ..................................................... 3:195
3.8.2 Wheat Germ .................................................... 3:196
3.8.3 Corn Germ Oil ................................................. 3:207
3.8.4 Rice Bran Oil ................................................... 3:215
3.8.5 Oat and Barley Oil ........................................... 3:223
3.8.6 Conclusions ..................................................... 3:227
xxviii Contents
References .................................................................... 3:227

3.9 Oils from Herbs, Spices, and Fruit Seeds ........................... 3:233
3.9.1 Introduction ..................................................... 3:233
3.9.2 Edible Seed Oils Rich in -linolenic Acid
(18:3n3) ........................................................... 3:234
3.9.3 Edible Seed Oils Rich in -linolenic Acid
(18:3n6) ........................................................... 3:239
3.9.4 Edible Seed Oils Rich in Linoleic Acid
(18:2n6) ........................................................... 3:241
3.9.5 Edible Seed Oils Rich in Oleic Acid
(18:1n-9) .......................................................... 3:248
3.9.6 Other Special Seed Oils of Fruit, Spice, and
Herb ................................................................ 3:254
3.9.7 Summary ......................................................... 3:255
References .................................................................... 3:256
3.10 Marine Mammal Oils ............................................................ 3:259
3.10.1 Introduction ..................................................... 3:259
3.10.2 Lipid Classes ................................................... 3:260
3.10.3 Fatty Acid Composition .................................... 3:262
3.10.4 Oxidative Stability ............................................ 3:267
3.10.5 Processing ....................................................... 3:268
3.10.6 Production of 3 Fatty Acid Concentrates ....... 3:269
3.10.7 Applications ..................................................... 3:271
3.10.8 Health Benefits and Disease Prevention .......... 3:272
3.10.9 Health Effects of DPA ...................................... 3:273
3.10.10 Comparison of Fish Oil and Marine Mammal
Oil .................................................................... 3:274
Reference ...................................................................... 3:275
3.11 Fish Oils ................................................................................ 3:279
3.11.1 Introduction ..................................................... 3:279
3.11.2 Why Do We Still Have Fish Oils? ..................... 3:282
Contents xxix
3.11.3 Fish Oil Fatty Acids and Gas-liquid

Chromatography .............................................. 3:284
3.11.4 Saturated, Isomeric Monoenoic, and Unusual
Fatty Acids ...................................................... 3:290
3.11.5 Polyunsaturated Fatty Acids ............................ 3:293
3.11.6 Fish Oil Production and Quality ....................... 3:296
3.11.7 Concentrates of Fish Oil Omega-3 Products .... 3:303
3.11.8 The Other Oils ................................................. 3:309
3.11.9 Conclusion ....................................................... 3:311
References .................................................................... 3:313
3.12 Minor Components of Fats and Oils .................................... 3:319
3.12.1 Introduction ..................................................... 3:319
3.12.2 The Chemistry of Minor Lipid Components ...... 3:321
3.12.3 Significance of Minor Lipid Components .......... 3:336
3.12.4 Analysis of Minor Lipid Components ................ 3:346
References .................................................................... 3:347
3.13 Lecithins ............................................................................... 3:361
3.13.1 Introduction ..................................................... 3:361
3.13.2 Sources of Phospholipids ................................ 3:362
3.13.3 Nomenclature, Classification, Structure and
Composition, and Chemical/Physical
Properties ........................................................ 3:372
3.13.4 Manufacture, Fractionation, and Purification
of Lecithins ...................................................... 3:384
3.13.5 Food-grade Lecithin Products, Uses ................ 3:400
3.13.6 Animal Feeds, Uses ........................................ 3:420
3.13.7 Nonfood and Industrial Uses ........................... 3:427
3.13.8 Availability and Economics .............................. 3:438
3.13.9 Regulatory Aspects ......................................... 3:439
3.13.10 Future Prospects ............................................. 3:440
xxx Contents
References .................................................................... 3:440

3.14 Lipid Emulsions .................................................................... 3:457
3.14.1 Introduction ..................................................... 3:457
3.14.2 Definitions ....................................................... 3:458
3.14.3 Droplet Characteristics .................................... 3:460
3.14.4 Emulsion Preparation ...................................... 3:468
3.14.5 Physicochemical Properties of Food
Emulsions ........................................................ 3:480
3.14.6 Conclusions ..................................................... 3:496
References .................................................................... 3:497
3.15 Dietary Fat Substitutes ......................................................... 3:503
3.15.1 Introduction ..................................................... 3:503
3.15.2 Reduced-calorie Structured Lipids as Fat
Substitutes ...................................................... 3:510
3.15.3 Fat Substitutes Based on Esters and Ethers .... 3:516
3.15.4 Fat Mimetics Based on Carbohydrates ............ 3:528
3.15.5 Fat Mimetics Based on Proteins ...................... 3:530
References .................................................................... 3:532
3.16 Structural Effects on Absorption, Metabolism, and
Health Effects of Lipids ........................................................ 3:535
3.16.1 Introduction ..................................................... 3:535
3.16.2 Stereospecific Effects of Fat Digestion and
Related Phenomena ........................................ 3:537
3.16.3 Differences in Metabolism between Esterified
and Free Fatty Acids ....................................... 3:543
3.16.4 Postprandial Effects ......................................... 3:545
3.16.5 Effects of Stereospecific Structure of Dietary
Acylglycerols on Chylomicron Clearing and
Tissue Targeting .............................................. 3:547
3.16.6 Effects of Stereospecific Structure of Dietary
Triacylglycerols on Their Health-related
Nutritional Effects ........................................... 3:547
Contents xxxi
3.16.7 Effects Related to Triacylglycerol

Hydrophobicity ................................................. 3:548
3.16.8 Effect of Acylglycerol Structure on Nutritional
Effects ............................................................. 3:549
3.16.9 Semi-synthetic Food Fats without
Acylglycerol Structure ...................................... 3:550
3.16.10 Conclusion ....................................................... 3:550
References .................................................................... 3:551
3.17 Modification of Fats and Oils via Chemical and
Enzymatic Methods .............................................................. 3:555
3.17.1 Introduction ..................................................... 3:555
3.17.2 Hydrogenation ................................................. 3:556
3.17.3 Fractionation .................................................... 3:557
3.17.4 Blending .......................................................... 3:558
3.17.5 Interesterification ............................................. 3:558
3.17.6 Hydrolysis and Esterification ............................ 3:569
3.17.7 Lipases in Lipid Modification ............................ 3:571
3.17.8 Modification of Fats and Oils to Produce
Structured Lipids ............................................. 3:579
References .................................................................... 3:582
3.18 Novel Separation Techniques for Isolation and
Purification of Fatty Acids and Oil by-products .................... 3:585
3.18.1 Introduction ..................................................... 3:585
3.18.2 Methods of Obtaining Fatty Acids .................... 3:586
3.18.3 Separation of Byproduct Components ............. 3:607
References .................................................................... 3:614
Volume 4. Edible Oil and Fat Products: Products

and Applications ................................................... 4: I
4.1 Frying Oils ............................................................................ 4:1
4.1.1 Introduction ..................................................... 4:1
4.1.2 Role of Oil or Fat in Frying ............................... 4:3
xxxii Contents
4.1.3 Applications of Frying Oil ................................. 4:3

4.1.4 Selection of Frying Oil ..................................... 4:4
4.1.5 The Frying Process ......................................... 4:5
4.1.6 Chemical Reactions Occurring in Oil during
Frying .............................................................. 4:7
4.1.7 Sources of Free Radicals ................................ 4:11
4.1.8 Polymerization ................................................. 4:11
4.1.9 Complexity of Oil Reactions in Frying .............. 4:12
4.1.10 Analytical Requirements for Fresh Frying
Oil .................................................................... 4:14
4.1.11 Impact of Tocopherols and Tocotrienols in
Frying Oil ......................................................... 4:16
4.1.12 Factors Affecting Frying Oil Quality ................. 4:17
4.1.13 Quality Standards of Oils (Table 3) .................. 4:19
4.1.14 Comments on Palm Oil .................................... 4:20
4.1.15 Enhancement of Frying Oil Performance ......... 4:21
4.1.16 Storage and Transportation of Frying Oil ......... 4:22
4.1.17 Hydrogenation and Trans-fat ........................... 4:23
4.1.18 Alternatives to Trans-fats ................................. 4:25
4.1.19 Frying Shortening versus Frying Oils ............... 4:27
4.1.20 Summary ......................................................... 4:28
References .................................................................... 4:29
4.2 Margarines and Spreads ...................................................... 4:33
4.2.1 Historical Development of Margarine (47) ...... 4:35
4.2.2 U.S. Trends ..................................................... 4:36
4.2.3 Regulatory Status in the United States ............ 4:37
4.2.4 Product Characteristics ................................... 4:42
4.2.5 Oils Used in Vegetable Oil Margarines and
Spreads ........................................................... 4:45
4.2.6 Other Common Ingredients ............................. 4:58
4.2.7 Processing ....................................................... 4:63
Contents xxxiii
4.2.8 Low-calorie Spreads ........................................ 4:68

4.2.9 Balanced Spreads ........................................... 4:70
4.2.10 Deterioration and Shelf Life ............................. 4:71
References .................................................................... 4:73
4.3 Shortenings: Science and Technology ................................ 4:83
4.3.1 Introduction ..................................................... 4:83
4.3.2 Plastic Theory .................................................. 4:88
4.3.3 Formulation ..................................................... 4:89
4.3.4 Manufacturing Processes and Equipment ........ 4:94
4.3.5 Shortening Production Systems ....................... 4:106
4.3.6 Analytical Evaluation and Quality Control ........ 4:114
4.3.7 Packaging and Storage ................................... 4:116
4.3.8 Innovations ...................................................... 4:117
References .................................................................... 4:123
4.4 Shortenings: Types and Formulations ................................. 4:125
4.4.1 Introduction ..................................................... 4:125
4.4.2 Shortening Attributes ....................................... 4:132
4.4.3 Base Stock System ......................................... 4:137
4.4.4 Shortening Formulation ................................... 4:140
4.4.5 Shortening Crystallization ................................ 4:147
4.4.6 Plasticized Shortening Consistency ................. 4:149
4.4.7 Liquid Opaque Shortenings ............................. 4:153
4.4.8 Shortening Chips and Flakes ........................... 4:155
References .................................................................... 4:156
4.5 Confectionery Lipids ............................................................. 4:159
4.5.1 Introduction ..................................................... 4:159
4.5.2 Chemistry of Chocolate ................................... 4:159
4.5.3 Characteristics of Cocoa Butter ....................... 4:160
4.5.4 Confectionery Fats .......................................... 4:165
4.5.5 Hard Butters .................................................... 4:168
4.5.6 Lauric CBS ...................................................... 4:168
xxxiv Contents
4.5.7 Non-lauric CBS ................................................ 4:170

4.5.8 Cocoa Butter Equivalents (CBEs) .................... 4:170
4.5.9 Organic Chocolate and Confectionery ............. 4:171
4.5.10 Recipe Engineering and Oil Processing ........... 4:172
4.5.11 Conclusions and Future Prospectives .............. 4:173
References .................................................................... 4:173
4.6 Cooking Oils, Salad Oils, and Dressings ............................. 4:175
4.6.1 Introduction ..................................................... 4:175
4.6.2 Natural and Processed Cooking and Salad
Oils .................................................................. 4:176
4.6.3 Stability of Salad and Cooking Oils .................. 4:178
4.6.4 Quality Evaluation of Salad and Cooking
Oils .................................................................. 4:180
4.6.5 Additives for Salad and Cooking Oils ............... 4:183
4.6.6 Nutrition-oriented Salad and Cooking Oils ....... 4:184
4.6.7 New Salad and Cooking Oils ........................... 4:184
4.6.8 Oil-based Dressings ........................................ 4:185
4.6.9 Viscous or Spoonable Dressings ..................... 4:185
4.6.10 Pourable Dressings ......................................... 4:191
4.6.11 Reduced-calorie Dressings .............................. 4:196
4.6.12 Fat-free Dressings ........................................... 4:198
4.6.13 Refrigerated Dressings .................................... 4:200
4.6.14 Heat-stable Dressings ..................................... 4:200
References .................................................................... 4:201
4.7 Fats and Oils in Bakery Products ........................................ 4:207
4.7.1 Introduction ..................................................... 4:207
4.7.2 Bread and Rolls ............................................... 4:208
4.7.3 Layered Doughs .............................................. 4:210
4.7.4 Cakes .............................................................. 4:212
4.7.5 Cake Donuts .................................................... 4:215
4.7.6 Cookies ........................................................... 4:216
Contents xxxv
4.7.7 Pie Crust, Biscuits ........................................... 4:219

4.7.8 Specifications for Bakery Shortenings ............. 4:219
References .................................................................... 4:227
4.8 Emulsifiers for the Food Industry ......................................... 4:229
4.8.1 Emulsifiers as Amphiphiles .............................. 4:229
4.8.2 Surfaces and Interfaces in Foods .................... 4:231
4.8.3 Surface Activity ................................................ 4:231
4.8.4 Emulsions ........................................................ 4:236
4.8.5 Foams ............................................................. 4:242
4.8.6 Wetting ............................................................ 4:243
4.8.7 Physical State of Emulsifier Plus Water ........... 4:244
4.8.8 Emulsifiers for Food Applications ..................... 4:248
4.8.9 Interactions with Other Food Components ....... 4:256
4.8.10. Some Food Applications .................................. 4:261
References .................................................................... 4:266
4.9 Frying of Foods and Snack Food Production ...................... 4:269
4.9.1 Introduction ..................................................... 4:269
4.9.2 Frying Process ................................................ 4:270
4.9.3 Types of Restaurant Fryers ............................. 4:270
4.9.4 Safety Issues ................................................... 4:277
4.9.5 Improving Restaurant Fryer Operation ............. 4:278
4.9.6 Measuring Oil Quality in the Fryer ................... 4:279
4.9.7 Physical Testing Devices ................................. 4:280
4.9.8 Chemical Tests ................................................ 4:282
4.9.9 Filtration and Treatment of Oil ......................... 4:284
4.9.10 Industrial Frying ............................................... 4:285
4.9.11 The Purpose of Frying Foods .......................... 4:288
4.9.12 Difference between the Frying and Other
Cooking Methods ............................................. 4:289
4.9.13 What Happens during Frying? .......................... 4:289
4.9.14 Types of Industrial Fryers ................................ 4:290
xxxvi Contents
4.9.15 Criteria for Fryer Selection ............................... 4:294

4.9.16 Components in an Industrial Frying System ..... 4:295
4.9.17 Continuous Potato Chip Process ..................... 4:296
4.9.18 Extruded Products ........................................... 4:299
4.9.19 Heat Wave Fryers ............................................ 4:303
4.9.20 Evolution of the Frying Industry into Diverse
Products .......................................................... 4:304
4.9.21 Low Oil Snacks ................................................ 4:304
4.9.22 Generation of Fines and Their Removal .......... 4:305
4.9.23 Terminology Used in Industrial Frying ............. 4:307
4.9.24 Fryer Capacity ................................................. 4:307
4.9.25 Key Points in Determining the Fryer Size ........ 4:308
4.9.26 Heat Load Requirement ................................... 4:310
4.9.27 Air Requirement (for Combustion) ................... 4:313
4.9.28 Product Bulk Density ....................................... 4:313
4.9.29 Oil Quality Management .................................. 4:313
4.9.30 Fryer Sanitation ............................................... 4:314
4.9.31 Summary ......................................................... 4:314
References .................................................................... 4:315
4.10 Fats and Oils in Feedstuffs and Pet Foods ......................... 4:317
4.10.1 History ............................................................. 4:319
4.10.2 Information Sources, Authorities, and
Obligations ...................................................... 4:320
4.10.3 Availability, Characteristics, and
Composition .................................................... 4:323
4.10.4 Digestion Metabolism and Fats Feeding
Requirements .................................................. 4:341
4.10.5 Fat Utilization Practices ................................... 4:367
References .................................................................... 4:386
4.11 By-product Utilization ........................................................... 4:391
4.11.1 By-products of Seed Processing ...................... 4:391
Contents xxxvii
4.11.2 By-products of Oil Refining .............................. 4:405

References .................................................................... 4:413
4.12 Environmental Impact and Waste Management ................. 4:417
4.12.1 Introduction ..................................................... 4:417
4.12.2 Process Components and Major Wastewater
Sources ........................................................... 4:418
4.12.3 Process Factors Affecting Wastewater
Generation and Characteristics ....................... 4:421
4.12.4 Air Emissions, Sources, and Controls .............. 4:423
4.12.5 Solid and Hazardous Wastes, Sources, and
Controls ........................................................... 4:427
4.12.6 Current Issues ................................................. 4:429
4.12.7 Wastewater Treatment Processes and
Technologies ................................................... 4:431
References .................................................................... 4:440
Volume 5. Edible Oil and Fat Products: Processing

Technologies ........................................................ 5: I
5.1 A Primer on Oils Processing Technology ............................ 5:1
5.1.1 Introduction ..................................................... 5:1
5.1.2 Storage ............................................................ 5:2
5.1.3 Preparation ...................................................... 5:5
5.1.4 Mechanical Extraction ..................................... 5:9
5.1.5 Solvent Extraction ............................................ 5:11
5.1.6 Degumming, Lecithin Processing, and
Physical Refining Pretreatment ........................ 5:16
5.1.7 Caustic Refining .............................................. 5:20
5.1.8 Bleaching ........................................................ 5:25
5.1.9 Dewaxing ........................................................ 5:29
5.1.10 Hydrogenation ................................................. 5:33
5.1.11 Interesterification ............................................. 5:37
5.1.12 Fractionation .................................................... 5:39
xxxviii Contents
5.1.13 Deodorization and Physical Refining ............... 5:42

5.1.14 Shortening and Margarine Manufacturing ........ 5:48
5.1.15 Acidulation ....................................................... 5:53
5.1.16 Summary ......................................................... 5:54
References .................................................................... 5:55
General References ....................................................... 5:56
5.2 Oil Extraction ........................................................................ 5:57
5.2.1 Evolution of Oil Extraction ............................... 5:57
5.2.2 Seed Preparation ............................................. 5:63
5.2.3 Mechanical Extraction ..................................... 5:71
5.2.5 Summary ......................................................... 5:97
References .................................................................... 5:97
5.3 Recovery of Oils and Fats from Oilseeds and Fatty
Materials ............................................................................... 5:99
5.3.1 Introduction ..................................................... 5:99
5.3.2 Mechanical Pretreatment ................................. 5:102
5.3.3 Heat Pretreatment ........................................... 5:109
5.3.4 Mechanical Expression of Oil .......................... 5:128
5.3.6 Types of Extractors .......................................... 5:160
5.3.7 Recovery of Solvent ........................................ 5:172
5.3.8 Obtaining Oil from Fruit Pulps .......................... 5:180
Acknowledgements ........................................................ 5:182
References .................................................................... 5:182
5.4 Storage, Handling, and Transport of Oils and Fats ............. 5:191
5.4.1 Introduction ..................................................... 5:191
5.4.2 Storage and Handling ...................................... 5:200
5.4.3 Deterioration Processes .................................. 5:220
5.4.4 Conclusions and Future Perspectives .............. 5:226
References .................................................................... 5:226
Contents xxxix
5.5 Packaging ............................................................................. 5:231

5.5.1 Conceptual Design of Packaging Systems for
Oil Products ..................................................... 5:231
5.5.2 Preliminary Engineering .................................. 5:238
5.5.3 Final Engineering ............................................ 5:241
5.5.4 Packaging Systems Components .................... 5:244
5.5.5 Edible Oil Operations ....................................... 5:258
5.5.6 Bulk Packaging Operations ............................. 5:262
5.5.7 Summary ......................................................... 5:263
5.5.8 Future Considerations ..................................... 5:264
References .................................................................... 5:264
5.6 Adsorptive Separation of Oils .............................................. 5:267
5.6.1 Definitions ....................................................... 5:267
5.6.2 Mathematical Models ....................................... 5:268
5.6.3 Mechanisms .................................................... 5:279
5.6.4 Mechanics ....................................................... 5:280
References .................................................................... 5:282
5.7 Bleaching .............................................................................. 5:285
5.7.1 Introduction ..................................................... 5:285
5.7.2 Background and Historical Perspective ........... 5:286
5.7.3 Adsorptive Purification Agents Description/
Preparation/Properties ..................................... 5:287
5.7.4 Trace Constituents in Lipid Oils and Fats ........ 5:297
5.7.5 Adsorptive Purification Process: General
Description ...................................................... 5:314
References .................................................................... 5:335
5.8 Deodorization ....................................................................... 5:341
5.8.1 Introduction ..................................................... 5:341
5.8.2 Deodorization Principle .................................... 5:343
5.8.3 Refined Oil Quality ........................................... 5:349
5.8.4 Deodorizer Technology .................................... 5:363
xl Contents
5.8.5 Commercial Deodorizer Systems .................... 5:376

5.8.6 Future Challenges ........................................... 5:381
References .................................................................... 5:382
5.9 Hydrogenation: Processing Technologies ........................... 5:385
5.9.1 Introduction ..................................................... 5:385
5.9.2 Is There a Future for Hydrogenation? .............. 5:390
5.9.3 Research on Trans-reduction by
Hydrogenation ................................................. 5:390
5.9.4 The Trans-fat Issue ......................................... 5:391
5.9.5 Hydrogen Supply for Hydrogenation ................ 5:394
References .................................................................... 5:395
5.10 Supercritical Technologies for Further Processing of
Edible Oils ............................................................................ 5:397
5.10.1 Introduction ..................................................... 5:397
5.10.2 Definition and Properties of Supercritical
Fluids ............................................................... 5:398
5.10.3 Historic Development and Commercial
Applications ..................................................... 5:399
5.10.4 Solubility Behavior of Lipid Components .......... 5:400
5.10.5 Supercritical Fluid Processing of Fats and
Oils .................................................................. 5:408
5.10.6 Novel Process Development: an Integrated
Approach ......................................................... 5:423
References .................................................................... 5:424
5.11 Membrane Processing of Fats and Oils .............................. 5:433
5.11.1 Introduction ..................................................... 5:433
5.11.2 Composition of Crude Vegetable Oils .............. 5:434
5.11.3 Crude Vegetable Oil Refining .......................... 5:434
5.11.4 Vegetable Oil Degumming ............................... 5:435
5.11.5 Membrane Processing of Oils and Fats ........... 5:437
5.11.6 Conclusions ..................................................... 5:456
References .................................................................... 5:457
Contents xli
5.12 Margarine Processing Plants and Equipment ..................... 5:459

5.12.1 Crystallization of Oil and Fat Products ............. 5:460
5.12.2 Processing Equipment for Margarine and
Related Fat Products ....................................... 5:469
5.12.3 Refrigerants for the Future ............................... 5:498
5.12.4 Plant Layout and Process Flowsheet ............... 5:499
5.12.5 Processing of Low-fat Spreads, Puff Pastry
Margarine, and Puff Pastry Butter .................... 5:511
5.12.6 Production Control, Quality Control, and
Sanitation ........................................................ 5:523
References .................................................................... 5:528
5.13 Extrusion Processing of Oilseed Meals for Food and
Feed Production ................................................................... 5:533
5.13.1 Introduction ..................................................... 5:533
5.13.2 Types of Extruders .......................................... 5:534
5.13.3 Why Process Oilseed with Extrusion? ............. 5:538
5.13.4 Soybeans Can Be Converted into Full-fat
Soy by Using Dry or Wet Extruders ................. 5:546
5.13.5 Extrusion-expelling of Oilseeds ....................... 5:554
5.13.6 Extrusion-expelling of Soybeans ...................... 5:554
5.13.7 Nutritional Advantages of Extrusion-expelling
of Oilseeds ...................................................... 5:559
5.13.8 Mechanical Crushing with Expanders .............. 5:563
5.13.9 Extrusion of Oilseeds before Extraction ........... 5:564
5.13.10 Extrusion of High Oilseeds .............................. 5:567
References .................................................................... 5:569
Volume 6. Industrial and Nonedible Products from

Oils and Fats ......................................................... 6: I
6.1 Fatty Acids and Derivatives from Coconut Oil ..................... 6:1
6.1.1 The Worlds Fats and Oils Output .................... 6:1
xlii Contents
6.1.2 The Role of Coconut Oil in the Oleochemical

Industry Worldwide .......................................... 6:2
6.1.3 Types of Fatty Acids and Derivatives from
Coconut Oil and Their General Applications .... 6:4
6.1.4 Fatty Acids ...................................................... 6:7
6.1.5 Methyl Esters ................................................... 6:13
6.1.6 Fatty Alcohols .................................................. 6:21
6.1.7 Glycerine ......................................................... 6:29
6.1.8 Monoalkyl Phosphates ..................................... 6:36
6.1.9 Alkanolamides ................................................. 6:39
6.1.10 Surfactants ...................................................... 6:43
6.1.11 Tertiary Amines ............................................... 6:52
References .................................................................... 6:54
6.2 Rendering ............................................................................. 6:57
6.2.1 Introduction ..................................................... 6:57
6.2.2 Modern Day Rendering .................................... 6:58
6.2.3 Byproducts ...................................................... 6:60
6.2.4 Processing ....................................................... 6:66
6.2.5 Product Composition ....................................... 6:75
6.2.6 Proteins ........................................................... 6:78
6.2.7 Fats ................................................................. 6:79
6.2.8 Oleochemistry ................................................. 6:82
6.2.9 Refining ........................................................... 6:85
6.2.10 Bleaching or Color Reduction .......................... 6:86
6.2.11 Uses of Refined and Bleached Fats ................ 6:87
6.2.12 Hydrogenation and Hydrogenated Products .... 6:87
6.2.13 Trans-esterification and Fatty Acid Esters ....... 6:89
6.2.14 Bio-fuel and Bio-diesel Products ...................... 6:93
6.2.15 Governmental Regulations .............................. 6:95
6.2.16 Environmental Issues ...................................... 6:98
6.2.17 Future Outlook ................................................. 6:99
Contents xliii
References .................................................................... 6:100

6.3 Soaps ................................................................................... 6:103
6.3.1 Introduction ..................................................... 6:103
6.3.2 Physical Properties of Surfactants ................... 6:104
6.3.3 Soap Raw Materials and Their Processing ...... 6:105
6.3.4 Soap Solution-phase Properties ...................... 6:108
6.3.5 Soap Solid-phase Properties and
Crystallization .................................................. 6:111
6.3.6 Commercial Processing ................................... 6:113
6.3.7 Bar Soap Manufacturing .................................. 6:122
6.3.8 Formulation of Soaps ...................................... 6:126
6.3.9 Economic Aspects ........................................... 6:132
6.3.10 Analytical Characterization of Soap ................. 6:132
6.3.11 Health, Safety, and Toxicology ........................ 6:133
6.3.12 Additional Uses of Soap .................................. 6:133
References .................................................................... 6:134
6.4 Detergents and Detergency ................................................. 6:137
6.4.1 Introduction ..................................................... 6:137
6.4.2 Components of Detersive Systems .................. 6:140
6.4.3 Formulation ..................................................... 6:142
6.4.4 Surfactants ...................................................... 6:142
6.4.5 Factors Influencing Detergency ....................... 6:149
6.4.6 Mechanisms .................................................... 6:156
6.4.7 Solid-soil Detergency ....................................... 6:158
6.4.8 Oily-soil Detergency ........................................ 6:164
6.4.9 Measurement of Detergency ............................ 6:168
6.4.10 Fabric Detergency ........................................... 6:169
6.4.11 Hard-surface Detergency ................................. 6:170
6.4.12 Detergent Manufacture .................................... 6:173
6.4.13 Analysis ........................................................... 6:176
xliv Contents
6.4.14 Health and Safety Factors ............................... 6:176

6.4.15 Environmental Considerations ......................... 6:179
References .................................................................... 6:181
6.5 Glycerine .............................................................................. 6:191
6.5.1 Introduction ..................................................... 6:191
6.5.2 Processing Principals and Details .................... 6:194
6.5.3 Processing Plants ............................................ 6:212
6.5.4 Properties of Glyceine ..................................... 6:215
6.5.5 Quality and Testing .......................................... 6:216
6.5.6 Processing Losses .......................................... 6:218
6.5.7 Waste Management ........................................ 6:219
6.5.8 Uses, Applications, and Economics ................. 6:219
6.5.9 Future Considerations ..................................... 6:221
References .................................................................... 6:221
6.6 Vegetable Oils as Biodiesel ................................................. 6:223
6.6.1 Introduction ..................................................... 6:223
6.6.2 Biodiesel Quality .............................................. 6:224
6.6.3 Biodiesel and Diesel Emissions ....................... 6:230
6.6.4 Resources for Biodiesel Production ................. 6:233
6.6.5 Production Technology .................................... 6:234
6.6.6 Utilization Technology ..................................... 6:246
6.6.7 Coproduct Use ................................................ 6:251
6.6.8 The Future ....................................................... 6:252
References .................................................................... 6:252
6.7 Vegetable Oils as Lubricants, Hydraulic Fluids, and
Inks ....................................................................................... 6:259
6.7.1 Introduction ..................................................... 6:259
6.7.2 Vegetable Oil Structure and Composition ........ 6:260
6.7.3 Oxidative Stability ............................................ 6:261
6.7.4 Low-temperature Properties ............................ 6:269
Contents xlv
6.7.5 Viscosities, Pour Points, and Oxidative

Degradation Tendencies of Major Lubricant
Basestocks ...................................................... 6:272
6.7.6 Effect of Diluent and Additives on Low-
temperature Properties .................................... 6:274
6.7.7 Conclusions ..................................................... 6:275
References .................................................................... 6:276
6.8 Vegetable Oils in Production of Polymers and Plastics ...... 6:279
6.8.1 Introduction ..................................................... 6:279
6.8.2 Polymers from Renewable Resources ............. 6:280
6.8.3 Exploitation of the Functional Groups on
Triglycerol Molecules for the Production of
Polymers ......................................................... 6:289
6.8.4 Use of Naturally Functionalized
Triacylglycerol Oils in Interpenetrating
Polymer Networks ........................................... 6:299
6.8.5 Conclusions ..................................................... 6:303
References .................................................................... 6:303
6.9 Paints, Varnishes, and Related Products ............................ 6:307
6.9.1 Relationship of Fats and Oils to the Paint-
coating Industry ............................................... 6:307
6.9.2 A Brief Overview of the Coatings
Technology ...................................................... 6:309
6.9.3 Film Drying Process of Oil-based Coating
Materials .......................................................... 6:314
6.9.4 Oleoresinous Varnishes ................................... 6:318
6.9.5 Alkyd Resins .................................................... 6:319
6.9.6 Safety and Environmental Precautions ............ 6:343
6.9.7 Modification of Alkyd Resins by Blending
with Other Polymers ........................................ 6:343
6.9.8 Economic Aspects ........................................... 6:348
6.9.9 Future Prospects ............................................. 6:349
xlvi Contents
References .................................................................... 6:349

6.10 Leather and Textile Uses of Fats and Oils .......................... 6:353
6.10.1 General Use of Fats and Oils in Leather .......... 6:353
6.10.2 Softening of Leather ........................................ 6:354
6.10.3 Softening with Fatliquor ................................... 6:355
6.10.4 Other Softening Materials ................................ 6:358
6.10.5 Evaluating Effects of Fat in Leather ................. 6:359
6.10.6 General Use of Fats and Oil for Textiles .......... 6:359
6.10.7 Common Textile Fibers .................................... 6:360
6.10.8 Processing of Fibers ........................................ 6:360
6.10.9 Physical Effects of Oils and Fats on Fibers
and Yarns ........................................................ 6:362
6.10.10 Oils and Fats in Textile Processing .................. 6:365
6.10.11 Concluding Remarks ....................................... 6:367
References .................................................................... 6:367
6.11 Edible Films and Coatings from Soybean and Other
Protein Sources .................................................................... 6:371
6.11.1 Edible Films and Coatings ............................... 6:371
6.11.2 Protein Films ................................................... 6:374
References .................................................................... 6:388
6.12 Pharmaceutical and Cosmetic Use of Lipids ....................... 6:391
6.12.1 Introduction ..................................................... 6:391
6.12.2 Lipids in Disease Prevention and Treatment .... 6:393
6.12.3 Lipids in Drug Delivery ..................................... 6:396
6.12.4 Lipids in Cosmetic Applications ....................... 6:400
6.12.5 Processing Oils for Pharmaceutical and
Cosmetics Applications ................................... 6:404
References .................................................................... 6:408
Index .................................................................................. I:1

BAILEYS INDUSTRIAL
OIL AND FAT
PRODUCTS
Sixth Edition
Volume 1
Chemistry, Properties, and
Health Effects
Edited by
Fereidoon Shahidi


1
Chemistry of Fatty Acids
Charlie Scrimgeour
Scottish Crop Research Institute
Dundee, Scotland
1. INTRODUCTION
Fatty acids, esterified to glycerol, are the main constituents of oils and fats. The
industrial exploitation of oils and fats, both for food and oleochemical products, is
based on chemical modification of both the carboxyl and unsaturated groups present
in fatty acids. Although the most reactive sites in fatty acids are the carboxyl group
and double bonds, methylenes adjacent to them are activated, increasing their
reactivity. Only rarely do saturated chains show reactivity. Carboxyl groups and
unsaturated centers usually react independently, but when in close proximity, both
may react through neighboring group participation. In enzymatic reactions, the
reactivity of the carboxyl group can be influenced by the presence of a nearby double
bond.
The industrial chemistry of oils and fats is a mature technology, with decades of
experience and refinement behind current practices. It is not, however, static. Envir-
onmental pressures demand cleaner processes, and there is a market for new pro-
ducts. Current developments are in three areas: green chemistry, using cleaner
processes, less energy, and renewable resources; enzyme catalyzed reactions,
used both as environmentally friendly processes and to produce tailor-made
products; and novel chemistry to functionalize the carbon chain, leading to new
Baileys Industrial Oil and Fat Products, Sixth Edition, Six Volume Set.
Edited by Fereidoon Shahidi. Copyright # 2005 John Wiley & Sons, Inc.
1
2 CHEMISTRY OF FATTY ACIDS
compounds. Changing perceptions of what is nutritionally desirable in fat-based

products also drives changing technology; interesterification is more widely
used and may replace partial hydrogenation in the formulation of some modified
fats.
The coverage in this chapter is necessarily selective, focusing on aspects of fatty
acid and lipid chemistry relevant to the analysis and industrial exploitation of oils
and fats. The emphasis is on fatty acids and acylglycerols found in commodity oils
and the reactions used in the food and oleochemical industries. The practical appli-
cation of this chemistry is dealt with in detail in other chapters. Current areas
of research, either to improve existing processes or to develop new ones, are also
covered, a common theme being the use of chemical and enzyme catalysts. Com-
pounds of second-row transition metals rhodium and ruthenium and the oxides of
rhenium and tungsten have attracted particular interest as catalysts for diverse reac-
tions at double bonds. Recent interest in developing novel compounds by functio-
nalizing the fatty acid chain is also mentioned. To date, few of these developments
have found industrial use, but they suggest where future developments are likely.
A number of recent reviews and books cover and expand on topics discussed
here (110).
2. COMPOSITION AND STRUCTURE
2.1. Fatty Acids

Fatty acids are almost entirely straight chain aliphatic carboxylic acids. The broad-
est definition includes all chain lengths, but most natural fatty acids are C4 to C22,
with C18 most common. Naturally occurring fatty acids share a common biosynth-
esis. The chain is built from two carbon units, and cis double bonds are inserted by
desaturase enzymes at specific positions relative to the carboxyl group. This results
in even-chain-length fatty acids with a characteristic pattern of methylene inter-
rupted cis double bonds. A large number of fatty acids varying in chain length
and unsaturation result from this pathway.
Systematic names for fatty acids are too cumbersome for general use, and
shorter alternatives are widely used. Two numbers separated by a colon give,
respectively, the chain length and number of double bonds: octadecenoic acid
with 18 carbons and 1 double bond is therefore 18:1. The position of double bonds
is indicated in a number of ways: explicitly, defining the position and configuration;
or locating double bonds relative to the methyl or carboxyl ends of the chain.
Double-bond position relative to the methyl end is shown as n-x or ox, where x
is the number of carbons from the methyl end. The n-system is now preferred,
but both are widely used. The position of the first double bond from the carboxyl
end is designated x. Common names (Table 1) may be historical, often conveying
no structural information, or abbreviations of systematic names. Alternative repre-
COMPOSITION AND STRUCTURE 3
TABLE 1. Fatty Acids in Commodity Oils and Fats. (a) Nomenclature and Structure.
Fatty acid Common name Formula Chain length
4:0 butyric CH3(CH2)2CO2H short

6:0 caproic CH3(CH2)4CO2H short
8:0 caprylic CH3(CH2)6CO2H short/medium
10:0 capric CH3(CH2)8CO2H medium
12:0 lauric CH3(CH2)10CO2H medium
14:0 myristic CH3(CH2)12CO2H medium
16:0 palmitic CH3(CH2)14CO2H
18:0 stearic CH3(CH2)16CO2H
18:1 9c oleic CH3(CH2)7CHCH(CH2)7CO2H
18:2 9c12c linoleic CH3(CH2)4(CH CHCH2)2(CH2)6CO2H
18:3 9c12c15c a-linolenic CH3CH2(CH CHCH2)3(CH2)6CO2H
22:1 13c erucic CH3(CH2)7CHCH(CH2)11CO2H long
20:5 5c 8c11c14c17c EPA CH3CH2(CH CHCH2)5(CH2)2CO2H long
22:6 4c7c10c13c16c19c DHA CH3CH2(CH CHCH2)6CH2CO2H long

Abbreviations of the systematic names eicosapentaenoic acid and docosahexaenoic acid.
sentations of linoleic acid (1) are 9Z,12Z-octadecadienoic acid; 18:2 9c12c; 18:2
n-6; 18:2 o6; 18:2 9,12; or CH3(CH2)4CH CHCH2CH CH(CH2)7COOH.
18
COOH
12 9 1
1
The terms cis and trans, abbreviated c and t, are used widely for double-bond
geometry; as with only two substituents, there is no ambiguity that requires the sys-
tematic Z/E convention. An expansive discussion of fatty acid and lipid nomencla-
ture and structure appears in Akoh and Min (1).
TABLE 1. (b) Occurrence.
Fatty Acid Significant Sources
4:0 butter, dairy fats

6:0 (coconut, palm kernel)
12:0 coconut, palm kernel
14:0 coconut, palm kernel
16:0 cottonseed, palm
18:0 cocoa butter, tallow
18:1 9c cottonseed, olive, palm, rape
18:2 9c12c corn, sesame, soybean, sunflower
18:3 9c12c15c linseed
20:1 13c high erucic rape
20:5 5c8c11c14c17c fish and animal fats
22:6 4c7c10c13c16c19c fish and animal fats
Over 1000 fatty acids are known, but 20 or less are encountered in significant
amounts in the oils and fats of commercial importance (Table 1). The most common
acids are C16 and C18. Below this range, they are characterized as short or medium
chain and above it as long-chain acids.
Fatty acids with trans or non-methylene-interrupted unsaturation occur naturally
or are formed during processing; for example, vaccenic acid (18:1 11t) and the con-
jugated linoleic acid (CLA) rumenic acid (18:2 9t11c) are found in dairy fats.
Hydroxy, epoxy, cyclopropane, cyclopropene acetylenic, and methyl branched fatty
acids are known, but only ricinoleic acid (12(R)-hydroxy-9Z-octadecenoic acid) (2)
from castor oil is used for oleochemical production. Oils containing vernolic acid
(12(S),13(R)-epoxy-9Z-octadecenoic acid) (3) have potential for industrial use.
OH
COOH
2
H O H
COOH
3
Typical fatty acid composition of the most widely traded commodity oils is
shown in Table 2.
TABLE 2. Fatty Acid Content of the Major Commodity Oils (wt%).
16:0 18:1 18:2 18:3 Other

(wt%) (wt%) (wt%) (wt%) [Fatty Acid (wt%)]
butter 28 14 1 1 4:0 (9); 6:012:0 (18); 14:0 (14) odd chain

and trans
castor 1 3 4 18:1(OH) (90)
coconut 9 6 2 8:0 (8); 10:0 (7); 12:0 (48); 14:0 (18)
corn 13 31 52 1
cottonseed 24 19 53
fish 14 22 1 16:1 n-7 (12); 20:1 n-9 (12); 22:1 n-11 (11);
20:5 n-3 (7); 22:6 n-3 (7)
groundnut 13 37 41 C20C24 (7)
(peanut)
lard 27 44 11 1 14:0 (2) 18:0 (11) long and odd chain
linseed 6 17 14 60
olive 10 78 7
palm 44 40 10
palm kernel 9 15 2 8:0 (3); 10:0 (4); 12:0 (49); 14:0 (16)
rape 4 56 26 10
sesame 9 38 45 18:0 (6)
soybean 11 22 53 8
sunflower 6 18 69 18:0 (6)
tallow 26 31 2 14:0 (6) 18:0 (31) long and odd chain
Typical midrange values shown; the balance are minor components. Data from (9).

Cod liver oil.

Low-erucic-acid rape, e.g., Canola.
Most commodity oils contain fatty acids with chain lengths between C16 and
C22, with C18 fatty acids dominating in most plant oils. Palm kernel and coconut,
sources of medium-chain fatty acids, are referred to as lauric oils. Animal fats have
a wider range of chain length, and high erucic varieties of rape are rich in this
C22 monoene acid. Potential new oil crops with unusual unsaturation or additional
functionality are under development. Compilations of the fatty acid composition of
oils and fats (6, 9, 11, 12) and less-common fatty acids (13) are available.
The basic structure, a hydrophobic hydrocarbon chain with a hydrophilic polar
group at one end, endows fatty acids and their derivatives with distinctive proper-
ties, reflected in both their food and industrial use. Saturated fatty acids have a
straight hydrocarbon chain. A trans-double bond is accommodated with little
change in shape, but a cis bond introduces a pronounced bend in the chain (Fig. 1).
In the solid phase, fatty acids and related compounds pack with the hydrocarbon
chains aligned and, usually, the polar groups together. The details of the packing,
such as the unit cell angles and head-to-tail or head-to-head arrangement depend on
the fatty acid structure (Fig. 2).
The melting point increases with chain length and decreases with increased
unsaturation (Table 3). Among saturated acids, odd chain acids are lower melting
than adjacent even chain acids. The presence of cis-double bonds markedly lowers
the melting point, the bent chains packing less well. Trans-acids have melting
points much closer to those of the corresponding saturates. Polymorphism results
in two or more solid phases with different melting points. Methyl esters are lower
melting than fatty acids but follow similar trends.
Fatty acid salts and many polar derivatives of fatty acids are amphiphilic, pos-
sessing both hydrophobic and hydrophilic areas within the one molecule. These are
surface-active compounds that form monolayers at water/air and water/surface
interfaces and micelles in solution. Their surface-active properties are highly
dependent on the nature of the polar head group and, to a lesser extent, on the
length of the alkyl chain. Most oleochemical processes are modifications of the car-
boxyl group to produce specific surfactants.
TABLE 3. Melting Points of Some Fatty Acids and Methyl Esters

Illustrating the Effect of Chain Length and Unsaturation.
Fatty acid Melting Point ( C) Fatty Acid Melting Point ( C)
16:0 62.9 (30.7)

17:0 61.3 (29.7)
18:0 70.1 (37.8)
18:1 9c 16.3, 13.4 18:1 9t 45
18:2 9c12c 5 18:2 9t12t 29
19:0 69.4 (38.5)
20:0 76.1 (46.4)
Values for methyl esters in parenthesis.

Data from (8) and (9).
Figure 1. Ball and stick models of (a) stearic acid, 18:0; (b) elaidic acid, 18:1 9t; and (c) oleic
acid 18:1 9c. All three lie flat in the plane of the paper. The cis double bond causes a distinct kink
in the alkyl chain of oleic acid.
2.2. Acylglycerols
Fatty acids in oils and fats are found esterified to glycerol. Glycerol (1,2,3-trihy-
droxypropane) is a prochiral molecule. It has a plane of symmetry, but if the pri-
mary hydroxyls are esterified to different groups, the resulting molecule is chiral
and exists as two enantiomers. The stereospecific numbering system is used to
Figure 2. Simplified diagram shows packing patterns of fatty acids in the solid phase. (a) and
(b): Hydrocarbon tails (straight lines) aligned at different angles to the line of the polar head
groups (circles). (c): Head to tail packing. (d): Head to head packing.
CH2OH sn-1 () CH2OOCR P

HO H sn-2 () RCOO H e.g. L
CH2OH sn-3 () or () CH2OOCR O
stereospecific numbering triacylglycerol

of glycerol backbone
CH2OOCR CH2OH
HO H RCOO H
CH2OH CH2OH
1-monoacyl-sn-glycerol 2-monoacyl-sn-glycerol
(1-MAG) (2-MAG)
CH2OOCR CH2OOCR
RCOO H HO H
CH2OH CH2OOCR
1,2-diacyl-sn-glycerol 1,3-diacyl-sn-glycerol
(1,2-DAG) (1,3-DAG)
CH2OOCR
RCOO H O
CH2O P OX
O
phosphatidylcholine X = CH2CH2N+(CH3)3
phosphatidylethanolamine X = CH2CH2N+H3
Figure 3. Structure and stereospecific numbering of acylglycerols.
distinguish between enantiomers. The Fischer projection of glycerol is drawn with

the backbone bonds going into the paper and the hydroxyl on the middle carbon to
the left. The carbons are then numbered 1 to 3 from the top (Figure 3). The prefix
sn- (for stereospecific numbering) denotes a particular enantiomer, rac- an equal
mixture of enantiomers, and x- an unknown stereochemistry. In an asymmetric
environment such as an enzyme binding site, the sn-1 and sn-3 groups are not inter-
changeable and reaction will only occur at one position. Simplified structures are
often used; e.g., 1-palmitoyl-2-linoleoyl-3-oleoyl-sn-glycerol is abbreviated to PLO
or drawn as shown in Figure 3.
Storage fats (seed oils and animal adipose tissue) consist chiefly (98%) of tria-
cylglycerols, with the fatty acids distributed among different molecular species.
With only two fatty acids, a total of eight triacylglycerol isomers are possible,
including enantiomers (Table 4). A full analysis of triacylglycerol molecular spe-
cies is a major undertaking, and for some oils, there are still technical difficulties to
be resolved. More commonly, triacylglycerols are distinguished by carbon number
(the sum of the fatty acid chain lengths) or unsaturation, using GC or HPLC for
analysis. The number of isomers increases as the cube of the number of fatty acids;
TABLE 4. Molecular Species of Triacylglycerols Containing only Palmitic and Oleic Acid.
PPP POP PPO OPP POO OOP OPO OOO
enantiomers * * ** **
carbon number 48 50 50 50 52 52 52 54
double bonds 0 1 1 1 2 2 2 3
Different methods of analysis will give different and often incomplete information about such a mixture. GC
analysis will separate molecular species by carbon number (sum of fatty acid chain lengths). Silver-ion HPLC
will separate by number of double bonds. Stereospecific analysis measures the proportions of fatty acids at
the sn-1, sn-2, and sn-3 positions, but it does not detect individual molecular species.
hence, even in oils with a simple fatty acid composition, many molecular species of
triacylglycerol may be present.
Most natural triacylglycerols do not have a random distribution of fatty acids on
the glycerol backbone. In plant oils, unsaturated acids predominate at the sn-2 posi-
tion, with more saturated acids at sn-1 and sn-3. The distribution of fatty acids at the
sn-1 and sn-3 positions is often similar, although not identical. However, a random
distribution between these two positions is often assumed as full stereospecific ana-
lysis is a time-consuming specialist procedure. In animal fats, the type of fatty acid
predominating at the sn-2 position is more variable; for example, palmitate may be
selectively incorporated as well as unsaturated acids (Table 5).
Only oils that are rich in one fatty acid contain much monoacid triacylglycerol,
for example, olive (Table 5), sunflower, and linseed oils containing OOO, LLL, and
LnLnLn, respectively. Compilations of the triacylglycerol composition of commod-
ity and other oils are available (8, 9).
The melting behavior of triacylglycerols generally reflects that expected from
the fatty acid composition; triacylglycerols rich in long-chain and saturated acids
TABLE 5. Contrasting Triacylglycerol Composition of Some Commodity Oils [Molecular

Species (wt%)].
Cocoa butter Coconut Lard Olive Soybean
POP (18-23) 12,12,8 (12) PPSt (2) OOL (11) LnLL (7)
POSt (36-41) 12,12,10 (6) StPSt (2) OOO (43) LnLO (5)
StOSt (23-31) 12,12,12 (11) PPO (8) POP (3) LLL (15)
12,12,14 (11) StOP (13) POL (4) LLO (16)
unsymmetrical 14,12,8 (9) POO (5) POO (22) LLS (13)
e.g., SSO <1% StOO (6) StOO (5) LOO (8)
OPO (18) LOS (12)
StPL (2) OOS (5)
OOO (12)
OPL (7)
Llinoleic; Lnlinolenic; Ooleic; Ppalmitic; Ssaturate; Ststearic; 88:0; 1010:0; 1212:0

(lauric); 1414:0.
Analysis by methods that do not distinguish all isomers; only major components are listed.
Data from (6).
are high melting, and those rich in polyunsaturated acids are lower melting. How-
ever, the situation is complicated by the possibility that the fatty acids can be dis-
tributed in different molecular species with different melting points. Oils with
similar fatty acid composition may have different solid fat content, polymorphic
forms, and melting behavior as a result of a different triacylglycerol composition.
Mono- and diacylglycerols (Figure 3) are not significant components of good
quality oils, but elevated levels may be found in badly stored seeds, resulting
from the activity of lipolytic enzymes. These compounds are produced industrially
by partial hydrolysis or glycerolysis of triacylglycerols for use as food grade emul-
sifiers. Mono- and diacylglycerols readily isomerize under acid or base catalysis
and are normally produced as an equilibrium mixture in which 1(3)-monoacylgly-
cerols or 1,3-diacylglycerols predominate.
Phospholipids (Figure 3) are constituents of membranes and are only minor
components of oils and fats, sometimes responsible for cloudiness. They are usually
removed during degumming, the residue from soybean oil processing being a
source of phospholipids used as food emulsifiers. The term lecithin is used
very loosely for such material, and it may variously mean phosphatidylcholine,
mixed glycerophospholipids, or crude phospholipid extracts from various sources.
Where possible, more specific nomenclature or the source and purity should be
used (14).
2.3. Bulk Properties

Saponification value and iodine value. Oils and fats are now characterized mainly
by their fatty acid composition determined by gas chromatography, replacing the
titrimetric and gravimetric assays used previously. However, the saponification
value (SV) or equivalent (SE) and iodine value (IV) are still used in specifications
and to monitor processes. SE, expressed as grams of fat saponified by one mole of
potassium hydroxide, is an indication of the average molecular weight and hence
chain length, whereas the IV, expressed as the weight percent of iodine consumed
by the fat in a reaction with iodine monochloride, is an index of unsaturation
(Table 6). Standard analytical methods are available (15), but these parameters
are now often calculated from the fatty acid composition, assuming that the sample
is all triacylglycerol (15). Indirect measurement of IV (16, 17) and SV (17) (as well
as peroxide and trans-content) using FT-NIR spectroscopy have been developed for
real-time process monitoring.
Unsaponifiable matter. Oils and fats contain variable amounts of sterols, hydro-
carbons, tocopherols, carotenoids, and other compounds, collectively referred to
as unsaponifiable matter because they do not produce soaps upon hydrolysis
(Table 6). The sterol and tocopherol composition of commodity oils is discussed
in another chapter. Some of these minor components are removed during refining,
and the resulting concentrates may be useful byproducts, for example, tocopherol
antioxidants. Characteristic fingerprints of minor components, particularly phytos-
terols and tocopherols, are also used to authenticate oils and detect adulteration
(18).
TABLE 6. Saponification Equivalent (SE), Saponification Value (SV), Iodine Value (IV),
and Unsaponifiable Matter of Some Commodity Oils.
SE SV IV Unsaponifiable
(g oil/mol KOH) (mg KOH/g oil) (100 g iodine/g oil) matter (wt%)
butter 242267 210232 2640 <0.5

castor 300319 176187 8191
coconut 212226 248265 611 <1.5
corn 288300 187195 107128 13
cottonseed 283297 189198 100115 <2
fish 292312 180192 142176 <2
groundnut (peanut) 286300 187196 86107 <1
lard 276292 192203 4570 <0.2
linseed 286298 188196 170203 <2
olive 286305 184196 7594 <1.5
palm 268295 190209 5055 <1.3
palm kernel 221244 230254 1421 <1
rape 291308 182193 110126 <0.2
sesame 288300 187195 104120 <2
soybean 288297 189195 124139 <1.5
sunflower 289298 188194 118145 <2
tallow 281295 190200 3347 <0.5

SE 56108/SV.

Cod liver oil.

Low erucic rape (Canola).
Data from (11).
3. HYDROLYSIS, ESTERIFICATION, AND ESTER EXCHANGE
Reactions converting acids to esters or vice versa and the exchange of ester groups
are among the most widely used in fatty acid and lipid chemistry (Figure 4). They
find applications from microscale preparation of methyl esters for GC analysis to
the industrial production of oleochemicals and biodiesel. The exchange of groups
attached to the fatty acid carboxyl is usually an equilibrium process driven to one
product by an excess of one reactant or the removal of one product, and it is usually
RCOOR + H2O RCOOH + ROH (1)
RCOOH + ROH RCOOR + H2O (2)
RCOOR + RCOOH RCOOR + RCOOH (3)
RCOOR + ROH RCOOR + ROH (4)
TAG + glycerol MAG + DAG (5)
Figure 4. Exchange reactions at the carboxyl group (1) hydrolysis (Chapter xx), (2) esterification
(Chapter xx), (3) acidolysis (Chapter xx), (4) alcoholysis (Chapter xx), and (5) glycerolysis
(Chapter xx). The starting ester RCOOR 0 will often be a triacylglycerol. MAGmonoacylglycer-
ol; DAGdiacylglycerol; TAGtriacylglycerol.
HYDROLYSIS, ESTERIFICATION, AND ESTER EXCHANGE 11
carried out with the aid of a catalyst. The catalyst may be an acid, a base, or a lipo-
lytic enzyme. These reactions produce the fatty acids and methyl esters that are the
starting point for most oleochemical production. As the primary feedstocks are oils
and fats, glycerol is produced as a valuable byproduct. Reaction routes and condi-
tions with efficient glycerol recovery are required to maximize the economics of
large-scale production.
There is increasing interest in the use of lipase enzymes for large-scale reactions.
Enzyme reactions require milder conditions, less solvent, and give cleaner pro-
ductsattributes of green chemistry. Enzymes can exert regio- or stereospecific
control over reactions and may also offer a degree of selectivity for particular fatty
acids, not observed with acid or base catalysts. Although the reactions of the car-
boxyl group are normally independent of those of the double bonds in the fatty acid
molecule, the presence of a double bond at the 4, 5, or 6 position often results
in slower reaction when a reaction is catalyzed by a lipase. Lipase catalyzed reac-
tions are considered in detail below, following a brief description of the reactions
involved.
3.1. Hydrolysis
The reaction can be catalyzed by acid, base, or lipase, but it also occurs as an unca-
talyzed reaction between fats and water dissolved in the fat phase at suitable tem-
peratures and pressures.
Base catalyzed hydrolysis. Historically, soaps were produced by alkaline hydrolysis
of oils and fats, and this process is still referred to as saponification. Soaps are now
produced by neutralization of fatty acids produced by fat splitting (see below), but
alkaline hydrolysis may still be preferred for heat-sensitive fatty acids.
On a laboratory scale, alkaline hydrolysis is carried out with only a slight excess
of alkali, typically 1M potassium hydroxide in 95% ethanol, refluxing for one hour,
and the fatty acids recovered after acidification of the reaction mixture. This is a
sufficiently mild procedure that most fatty acids, including polyunsaturates, epox-
ides, and cyclopropenes, are unaltered (19).
Fat splitting. The industrial production of fatty acids uses the direct reaction
between water and fats, which proceeds rapidly at 250 C and 26 MPa (20
60 bar). Under these conditions, water is moderately soluble in the oil phase, and
stepwise hydrolysis of the triacylglycerols proceeds without the aid of a catalyst.
The reaction is carried out with a countercurrent of water that removes the glycerol
formed, resulting in 99% conversion to fatty acids. Glycerol is recovered from the
aqueous phase. Sonntag has reviewed industrial fat splitting in detail (20).
3.2. Esterification
Fatty acids are converted to esters by reaction with an excess of alcohol using an
acid catalyst or a lipase. For the preparation of methyl esters for GC analysis, boron
trifluoride, sulfuric acid, or anhydrous hydrogen chloride in methanol are com-
monly used (19). Reaction is complete in 30 minutes at reflux. Propyl and butyl
esters are prepared in a similar way with the corresponding alcohols. It is not
always possible to use an excess of alcohol, for example, in the synthesis of tria-
cylglycerols using a protected glycerol. A more reactive fatty acid derivative such
as the acid chloride or anhydride is used, or the fatty acid is reacted directly with
the alcohol, using dicyclohexylcarbodiimide (DCC) plus 4-dimethylaminopyridine
(DMAP) as a coupling agent, for example, in the synthesis of acylglycerols (21).
Some groups in more unusual fatty acids are acid sensitive, for example, epoxides,
cyclopropanes, cyclopropenes, and hydroxy compounds, and methods avoiding
acids catalysts are needed. Reaction with diazomethane or the less hazardous
trimethylsilyl-diazomethane are possibilities (19).
3.3. Ester Exchange Reactions

The fatty acid or alcohol groups present in an ester can be exchanged in a number
of ways: by reaction with an excess of other fatty acids (acidolysis), alcohols (alco-
holysis), or other esters (interesterification). Generally, the starting point will be a
triacylglycerol, and these reactions provide routes by which the composition and
properties of oils and fats can be modified.
Acidolysis. This reaction can be acid or enzyme catalyzed and may be used to mod-
ify triacylglycerol composition. Acidolysis of an oil containing only C16 and C18
fatty acids with fatty acids rich in lauric acid (e.g., from palm-kernel oil) results
in a triacylglycerol enriched in medium-chain fatty acids.
Alcoholysis. Methanolysis of triacylglycerols is used to prepare methyl esters for
fatty acid analysis, a process frequently referred to as transesterification. This
can be acid-or base-catalyzed, the method being chosen to avoid modifying
acid-or base-sensitive fatty acids and to minimize reaction times. Sterol esters of
fatty acids react more slowly than triacylglycerols, and samples containing them
require more vigorous reaction conditions. The preparation of methyl esters
from oils and fats for GC and GC-MS analysis has been extensively reviewed
(19, 22, 23).
Biodiesel is produced on the industrial scale by methanolysis of vegetable oils
(usually rape or soybean) or waste fat, particularly using frying oils. Methanolysis
proceeds with modest amounts of base catalyst, provided the levels of free fatty
acid and water in the oil are low (24, 25). The fatty acid content may be reduced
by physical or chemical treatment before methanolysis but for waste fats, alterna-
tive processes that do not use base catalysis may be preferred. Lipase catalyzed
methanolysis is less sensitive to fatty acid and water in the oil and has been tested
in batch (26) and fixed-bed reactor (27) conversion of waste oil and grease to
biodiesel.
Glycerolysis, the treatment of triacylglycerols with glycerol and a basic catalyst
(sodium hydroxide or sodium methoxide), is used to produce mono- and diacylgly-
cerols on an industrial scale. Molecular distillation is used to produce MAG,
which is 9095% pure and is widely used as an emulsifying agent in foods and
other applications.
HYDROLYSIS, ESTERIFICATION, AND ESTER EXCHANGE 13
Interesterification. Interesterification is the intra- and intermolecular exchange of

fatty acids on the glycerol backbone of triacylglycerols, although the term is also
used more loosely to include acidolysis and other ester exchange reactions. It is
applied to either an individual oil or a blend of oils, to produce triacylglycerols
with different properties. The molecular species of natural triacylglycerols is not
a random mixture of all possible isomers, but it shows greater or lesser selectivity
in the distribution of fatty acids between the sn-1 and sn-3 and the sn-2 positions
(Table 5). This, as well as the overall fatty acid mixture, determines many of the
technically important properties of the oil or fat, for example, solid fat content and
melting point. Once subjected to interesterification with a chemical catalyst, the
triacylglycerol becomes a random mixture of molecular species. Lipase catalyzed
interesterification may alter the distribution of molecular species in a more selective
way.
Chemical interesterification (28, 29) is carried out at moderate temperatures
(70100 C), with neat oils and a low concentration (<0.4%) of a base catalyst
such as sodium methoxide or ethoxide or Na/K alloy. As the catalyst is destroyed
by water and free fatty acids, the oil must be carefully refined and dried before
adding the catalyst. Reaction proceeds through sequential fatty acid exchange
reactions, following formation of what is believed to be the true catalyst, the alkali
metal derivative of a diacylglycerol. There is no observed selectivity for fatty acid
or glycerol position, leading to a fully random product. The product composition
can be controlled through directed interesterification at lower temperatures.
Na/K alloy is used as catalyst as it is active at temperatures below 50 C and cooling
the reaction mixture causes high melting trisaturated triacylglycerols to crystallize
out, altering the composition of the liquid phase in which reaction occurs. The
remaining liquid phase is randomized by further reaction and high melting
products continue to crystallize out, eventually leading to solid and liquid
products richer in trisaturated and triunsaturated species than the fully randomized
fat (29).
Interesterification is used to modify fat properties without recourse to partial
hydrogenation. Hardened fats produced by partial hydrogenation contain trans-
isomers, which are now regarded as undesirable by nutritionists and will be increas-
ingly subject to product labeling regulations. Liquid fats can be hardened by inter-
esterification with fully saturated fats (either stearin fractions or fully hydrogenated
oils), raising the solid fat content without isomerizing any of the fatty acids. The use
of interesterification to produce margarine and spreads has increased recently, par-
ticularly in Europe.
3.4. Lipase Catalyzed Reactions

Lipases are enzymes that hydrolyze fatty acids from lipid species (e.g., triacylgly-
cerols or phospholipids) in vivo. A number of lipases, mainly of bacterial origin, are
now available immobilized onto a solid support for use as industrial scale catalysts.
Immobilized lipases catalyze the whole range of ester exchange reactions described
above (alcoholysis, acidolysis, esterification) as well as hydrolysis. There are two
significant differences between lipase and chemically catalyzed reactions. First,
lipase catalyzed reactions take place at a lower temperature and with fewer side
reactions, leading to cleaner products: an environmentally friendly alternative to
some existing processes. Second, enzyme catalyzed reactions are more selective,
offering control over reactions not possible with a chemical catalyst. Selectivity
may be for fatty acids at different positions on the glycerol backbone (sn-1 and
sn-3 rather than sn-2) or for particular fatty acids, discriminating by double-bond
position or chain length (30, 31). The widely studied Lipozyme RM IM (Rhizomu-
cor miehei lipase immobilized onto a weak anion exchange resin) preferentially
hydrolyzes short-chain acids relative to medium and long chains from triacylglycer-
ols. Hydrolysis at the sn-1 position is somewhat faster than at sn-3, and hydrolysis
at sn-2 is very slow (31).
Lipase catalyzed reactions take place in the neat oil or in a nonpolar (usually
hydrocarbon) solvent. The efficiency depends on the amount of water, solvent (if
present), temperature, and ratio of reactants. A factorial approach can be used to
optimize the conditions (32). In interesterification reactions, 1,3-specific enzymes
give control over product composition that is not possible using chemical catalysts.
For example, starting with SOS and OOO, chemical interesterification produces all
eight possible isomers (see Table 5). Enzymatic interesterification does not
exchange fatty acids at the sn-2 position, and it will result in only two additional
molecular species, OOS and SOO. In more realistic situations, chemical and enzy-
matic interesterification may produce the same or a similar number of molecular
species, but in different proportions (31).
Enzymatic interesterification has most potential for high-value products such as
confectionary fats and nutritional products, for example, cocoa butter equivalents
prepared from cheap and readily available starting materials. Acidolysis of palm
mid fraction, rich in POP, with stearic acid gives a cocoa butter equivalent rich
in POSt and StOSt, through exchange at the sn-1 and sn-3 positions while retaining
the oleate at the sn-2 position. Tripalmitin treated similarly with oleic acid gives
products where the palmitate is retained at the sn-2 position, whereas oleate is intro-
duced at sn-1 and sn-3, producing a human milk fat substitute such as Betapol. In
practice, pure starting materials are not used. Feedstocks rich in tripalmitin and
oleic acid are reacted in a two step-process: alcoholysis to sn-2- monoacylglycerols
followed by esterification (33).
Both batch and fixed-bed reactors have been used and tested on the near ton
scale (34) for the production of high-value fats. This technology has now pro-
gressed to pilot production, using a 1-m3 fixed-bed plug-in reactor containing the
immobilized enzyme Lipozyme TL IM (35). Blends of palm oil or stearin with
palm-kernel or coconut oil are interesterified in less than one hour at 70 C, and
no downstream processing is required as the enzyme is retained in the reactor.
This is a practical, lower energy alternative to hydrogenation and chemical interes-
terification, free from the trans-isomer production of the former and more selective
and natural than the latter.
OXIDATION 15
Lipases also discriminate between fatty acids with different double-bond posi-
tions. The reaction of fatty acids with 4, 5, and 6 double bonds is significantly
slower than 9 acids when catalyzed by some enzymes. This is illustrated by
some examples of attempts to concentrate g-linolenic acid (GLA; 18:3 6c9c12c)
from borage oil. Hydrolysis of borage oil with Candida rugosa lipase resulted in
selective hydrolysis of the 9 acids (mainly 18:2) increasing the amount of
GLA in the remaining acylglycerols (36). The efficiency of the enrichment was
influenced by the initial triacylglycerol composition and the extent of hydrolysis.
Starting with a borage oil containing 22% GLA, the upper limit of enrichment
was to 46%, but higher values resulted from repeated hydrolysis of the recovered
acylglycerols. A two-step sequence involving both enzymatic hydrolysis and re-
esterification achieved higher enrichment (37). Nonselective hydrolysis with Pseu-
domonas sp. lipase was optimized for high GLA recovery (93%). Esterification with
lauryl alcohol, using Rhizopus delemar lipase, discriminated strongly against GLA,
resulting in enrichment in the unesterified fatty acids from 22.5% to 70.2% with a
recovery efficiency of 75.1%. A 92.1% GLA concentrate, obtained by low-tempera-
ture crystallization of borage oil fatty acids, was enriched to 99.1% by esterification
with butanol, catalyzed by Lipozyme IM-60 (38).The overall recovery was 72.8%.
The operating parameters (alcohol, concentration, temperature, and solvent) were
systematically investigated.
Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), 5 and 4
acids respectively, are discriminated against during lipase catalyzed reactions and
reaction of DHA may be significantly slower than EPA. Alcoholysis of tuna oil
ethyl esters with lauryl alcohol using Rhizomucor miehei lipase enriches the
DHA in the unreacted ethyl esters, whereas the concentration of EPA is simulta-
neously reduced (39). A concentrate containing 60% DHA and 8.6% EPA was alco-
holyzed with excess lauryl alcohol (1:7 mole ratio). The remaining ethyl esters
contained 93% DHA in 74% recovery, and EPA was reduced to 2.9%. Both nonre-
giospecific and sn-1,3-specific enzymes incorporate GLA into seal blubber and
menhaden oil (3:1 mole ratio of GLA to triacylglycerol) producing an oil rich in
both n-3 and n-6 polyenes (40). The highest incorporation was with the nonspecific
enzyme.
4. OXIDATION
The fatty acid alkyl chain is susceptible to oxidation both at double bonds and adja-
cent allylic carbons. Free-radical and photooxidation at allylic carbons are respon-
sible for deterioration of unsaturated oils and fats, resulting in rancid flavors and
reduced nutritional quality, but they are also used deliberately to polymerize drying
oils. Oxidation of double bonds is used in oleochemical production either to cleave
the alkyl chain or to introduce additional functionality along the chain. Enzyme cat-
alyzed oxidation is the initial step in the production of eicosanoids and jasmonates
(biologically active metabolites in animals and plants respectively) but is not dis-
cussed further here.
4.1. Autoxidation and Photooxidation

Both autoxidation and photooxidation produce allylic hydroperoxides from unsatu-
rated centers.
CH CH CH2 + O2 CH CH CH(OOH)
During this process, the position and geometry of the double bond may change.
The hydroperoxide mixtures produced by autoxidation and photooxidation are not
the same, indicating that different mechanisms are involved. Free radical oxidation
can be promoted or inhibited. Deliberate promotion speeds the polymerization of
drying oils, and strenuous efforts are made to inhibit the onset of rancidity in edible
oils. Frankel has recently reviewed this topic in depth (41); see also (1) for an exten-
sive discussion of oxidation of food lipids.
4.1.1. Autoxidation Autoxidation is a free-radical chain reaction, involving a

complex series of reactions that initiate, propagate, and terminate the chain.
initiation RH ! R
propagation R
O2 ! ROO
fast
ROO
RH ! ROOH R
rate determining
termination R
, ROO
! stable products
The chain reaction is initiated by abstraction of an allylic hydrogen to give an

allylic radical stabilized by delocalization over three or more carbons. The initiator
is a free radical, most probably produced by decomposition of hydroperoxides
already present or produced by photooxidation. The decomposition may be thermal,
but it is more likely promoted by traces of variable redox state metal ions. Auto-
xidation is characterized by an induction period during which the concentration of
free radicals increases until the autocatalytic propagation steps become dominant.
During the induction period, there is little increase in oxidation products.
The first step of the propagation sequence is reaction of the allylic radical with
molecular oxygen, producing a peroxy radical. This step is much faster than the
subsequent abstraction of another allylic hydrogen by the peroxy radical, producing
both an allylic hydroperoxide and a new allylic radical that continues the chain
reaction. Hydrogen abstraction is the rate-determining step and is therefore selec-
tive for the most readily abstracted hydrogen. Methylene-interrupted dienes and
polyenes, where the allylic radical can be delocalized over five carbons, are oxi-
dized faster than monoenes where the radical is delocalized over three carbons
(Figure 5).
The chain reaction is terminated by reactions that remove radicals that would
otherwise produce more allylic radicals by hydrogen abstraction. Examples are
the combination of two hydroperoxy radicals leading to nonradical products and
molecular oxygen or reaction with a free-radical scavenger (antioxidant) generating
a more stable radical.
OXIDATION 17
(a) (b)
Figure 5. Allylic radicals produced during autoxidation. (a) Those from isolated double bonds
are delocalized over three carbons. (b) Those from methylene-interrupted dienes or polyenes are
delocalized over five carbons. The arrows show the site of attachment of O2 giving a peroxy
radical.
The rate of autoxidation generally increases with increasing unsaturation.

Linoleate, as neat methyl or ethyl ester, reacts approximately 40 times faster
than oleate, and for higher polyenes, the rate doubles for each additional double
bond (42). Trilinolein does not follow the same kinetics as the simple esters and
oxidizes somewhat faster. The medium also influences susceptibility to oxidation,
and these generalizations may not hold in emulsified systems (e.g., many food for-
mulations) where oxidation occurs at the interface between aqueous and fat phases
(43). In aqueous micelles, EPA and DHA are unexpectedly stable (44), oxidizing
much more slowly than linoleate. In one experiment, over half the linoleate was
oxidized within 50 hours and 90% of EPA and DHA was still present after
2000 hours. The stability of the higher polyenes is attributed to their tightly coiled
configuration in the aqueous medium, making attack by oxygen or free radicals
more difficult.
Mechanistic studies of autoxidation have concentrated on methylene-interrupted
fatty acids, but many of the observations are valid for other compounds. Conjugated
fatty acids such as CLA also oxidize through an autocatalytic free radical reaction,
with the predominant hydroperoxide determined by the geometry of the conjugated
diene system (45). Other groups with activated methylenes may be susceptible to
oxidation, for example, the ether methylenes of ethoxylated alcohols used as sur-
factants (46).
4.1.2. Photooxidation Light, in the presence of oxygen, promotes oxidation of

unsaturated fatty acids. Ultraviolet radiation decomposes existing hydroperoxides,
peroxides, and carbonyl and other oxygen-containing compounds, producing radi-
cals that initiate autoxidation (42). Photooxidation by longer wavelength near ultra-
violet or visible light requires a sensitizer. Naturally present pigments such as
chlorophyll, hematoporphyrins, and riboflavin act as sensitizers as do dyes, includ-
ing erythrosine and methylene blue. Light excites these sensitizers to the triplet
state that promotes oxidation by type I and type II mechanisms. Unlike autoxida-
tion, there is no induction period.
In type I photosensitized oxidation, the triplet state sensitizer abstracts a hydro-
gen or electron from the unsaturated oil, producing radicals that initiate chain pro-
pagation as in autoxidation. However, chain-breaking antioxidants do not stop this
reaction as new radicals are produced photochemically. In type II photooxidation,
the energy of the triplet sensitizer is transferred to molecular oxygen, converting it
HOO
H
O O
Figure 6. Ene reaction between singlet oxygen and an olefinic bond. The hydroperoxide may be
attached to either of the inital double bond carbons.
to its excited singlet state. Singlet oxygen is highly electrophilic and reacts rapidly
with olefins in an ene reaction, producing allylic hydroperoxides with oxygen
attached to one of the original olefinic carbons and the shifted double bond now
trans (Figure 6).
The ene reaction differs from free-radical oxidation, where oxygen attaches to an
outer carbon of the delocalized allylic radical (Figure 5), resulting in a different
mixture of hydroperoxides. For example, photooxidation of linoleate produces
four isomers: 9-OOH,10t12c, 10-OOH,8t12c, 12-OOH,9c13t, and 13-OOH,9c11t.
The same 9- and 13-hydroperoxides are produces by autoxidation, but the 10-
and 12-hydroperoxides are only produced by photooxidation.
Photooxidation is much faster than autoxidation; the reaction of linoleate with
singlet oxygen is approximately 1500 times faster than that with triplet oxygen
(47). There is less difference in the rate of photooxidation between monoenes
and polyenes than is seen in autoxidation. The relative rates for oleate, linoleate,
linolenate, and arachidonate are 1.0, 1.7, 2.6, and 3.1 (48, 49). This contrasts
with the 40-fold increase in rate of autoxidation between oleate and linoleate.
4.1.3. Decomposition of Hydroperoxides Allylic hydroperoxides are reactive

molecules and decompose readily in a complex series of reactions, the course of
which depends on the medium and other conditions (1, 41). Cleavage between
the oxygens is energetically favored, leading to alkoxy and hydroxyl radicals.
Redox metal ions such as Fe2/Fe3 and Cu/Cu2 are particularly effective cata-
lysts. The resulting radicals can initiate further autoxidation and produce a number
of stable products, many with undesirable nutritional and flavor properties (Fig-
ure 7). Products with the same chain length as the alkoxy radical include epoxides,
ketones, and hydroxy fatty acids. The significant products producing off-flavors are
those resulting from chain scission b to the alkoxy radical, producing shorter chain
aldehydes and hydrocarbons. Alkadienals have particularly low-odor thresholds and
a few parts per billion of nonadienals from n-3 fatty acids are responsible for a
marked fishy taint even when other signs of oxidation are absent (50).
There are a number of analytical measures of oxidative deterioration of oils and
fats. The most widely used are the peroxide value (PV) (15), which measures the
hydroperoxide content by iodine titration and the anisidine value (AV) (15), which
detects aldehydes by a color reaction. As an oil suffers damage because of autoxi-
dation, the hydroperoxide content, and PV rise but do not do so indefinitely. As
the hydroperoxides break down, the concentration of aldehydes and AV increase.
Oxidation is better assessed by a combination of PV and AV, the Totox value
OXIDATION 19
R CH CH CH R
OOH
keto
epoxy
fatty acids
OH + R CH CH CH R hydroxy
dihydroxy
O
oligomers and polymers
A B
A RCHO + RCH CH
OH
RCH CH OH RCH2CHO
B RCH CHCHO + R
RH
Figure 7. Decomposition reactions of allylic hydroperoxides.
( 2 PV AV) being a better index of oxidation than either PV or AV alone.

Volatile products can be removed from oils by deodorization, but aldehydes
attached to the carboxyl end of the chain remain part of the triacylglycerol (some-
times called core aldehydes) and are indicators of previous oxidative damage.
4.1.4. Antioxidants Lipid oxidation is influenced by many factors: the medium,

oxygen concentration, temperature, light, degree of unsaturation, and metal ions
among others. In the presence of oxygen, oxidation cannot be entirely prevented
nor can it be reversed, but it can be inhibited, delaying the buildup of oxidized
products to unacceptable levels. Antioxidants can interact with several steps of
free-radical or photooxidation. Their performance is medium and concentration
dependent and requires care as they can also act as prooxidants under some
conditions (51).
The most widely used antioxidants are free radical scavengers that remove reac-
tive radicals formed in the initiation and propagation steps of autoxidation. A num-
ber of natural or synthetic phenols can compete, even at low concentrations, with
lipid molecules as hydrogen donors to hydroperoxy and alkoxy radicals, producing
hydroperoxides and alcohols and an unreactive radical. b-carotene reacts with per-
oxy radicals, producing a less-reactive radical. These stabilized radicals do not
initiate or propagate the chain reaction.
OH
CH3 HO
HO HOOC
CH3 O
CH3
(a) (b)
OH OH OH
HO OH
HO O
O
OCH3 CH3 COOC3H7
(c) (d) (e) (f)
Figure 8. Natural antioxidants (a) a-tocopherol, (b) carnosic acid, and (c) sesamol. Synthetic
antioxidants (d) butylated hydroxyanisole (BHA), (e) butylated hydroxytoluene (BHT), and (f)
propyl gallate.
Tocopherols are phenolic antioxidants (Figure 8) naturally present in most plant

oils (see Chapter X). They are concentrated in the distillate from physical refining,
which results in a corresponding decrease in the refined oil. Soybean distillate is a
source of tocopherols for antioxidant formulations. Carnosic acid (Figure 8) is iso-
lated from rosemary and other herbs. Sesamol (Figure 8) is a characteristic antiox-
idant of sesame oil, responsible for its high stability (Chapter xx). Synthetic
antioxidants are monocyclic phenols with highly branched substituents (Figure 8).
In all of these compounds, the radicals formed by abstraction of the phenolic
hydrogen are highly delocalized and unreactive. The antioxidant action of free-
radical scavengers is sacrificial, delaying oxidation until the antioxidant is used
up. Oxidized tocopherols may be regenerated by ascorbic acid, extending their
effective life while keeping their concentration below prooxidant levels.
Photooxidation is not inhibited by free-radical scavengers. Natural pigments that
act as sensitizers may be reduced during refining, increasing stability. Singlet oxy-
gen and excited state sensitizers can be deactivated either by competitive reaction
or physical energy transfer, for example, to b-carotene. Tocopherols and some
amines also act as singlet oxygen quenchers through physical energy transfer.
Redox metal ions, particularly iron and copper, react with hydroperoxides, initi-
ating further autoxidation and producing undesirable decomposition products.
Complete removal of these metal ions is not possible, but steps can be taken to
reduce their effect. Chelating agents such as EDTA, citric acid, phosphate, and
polyphosphates may reduce the effective metal ion concentration. Their efficacy
depends on pH, and they may also show prooxidant activity. The role of metal
ions in hydroperoxide decomposition in food emulsions has been reviewed recently
(52).
OXIDATION 21
R H R H O R H H
H O H O
O R O +
O O
R H O R R H O R H O R
Figure 9. Epoxidation mechanism proposed by Bartlett (53). The cis-olefin gives rise to a cis-
epoxide.
4.2. Epoxidation
Epoxides are produced by reaction of double bonds with peracids. This proceeds by
a concerted mechanism, giving cis stereospecific addition (Figure 9) (53). Thus, a
cis olefin leads to a cis epoxide and a trans olefin to a trans epoxide. The order of
reactivity of some peracids is m-chloroperbenzoic > performic > perbenzoic >
peracetic; electron withdrawing groups promote the reaction. The carboxylic acid
produced is a stronger acid than the strongly hydrogen bonded peracid and may
lead to subsequent ring opening reactions especially in the case of formic acid.
Small scale reactions are carried out with m-chloroperbenzoic acid in a halocarbon
or aromatic solvent, in the presence of bicarbonate to neutralize the carboxylic acid
as it is formed (54, 55).
Oils, mainly soybean but also linseed, are epoxidized on an industrial scale
(100,000 tons per year) as stabilizers and plasticizers for PVC. The reactive epoxide
groups scavenge HCl produced by degradation of the polymer. Epoxidation is car-
ried out with performic or peracetic acid produced in situ from formic or acetic acid
and high strength hydrogen peroxide (70% w/w). Peracids are unstable, and the
reaction is exothermic. The concentration of peracid is kept low by using a low con-
centration of the carboxylic acid either in the neat oil or in a hydrocarbon solvent.
The carboxylic acid is regenerated after epoxidation. Complete epoxidation is not
achieved as in the acidic medium ring opening reactions occur producing dihydroxy
and hydroxy carboxylates as byproducts.
Recent studies have attempted to improve the efficiency of epoxidation under
milder conditions that minimize the formation of byproducts. Chemo-enzymatic
epoxidation uses the immobilized lipase from Candida antartica (Novozym 435)
(56) to catalyze conversion of fatty acids to peracids with 60% hydrogen peroxide.
The fatty acid is then self-epoxidized in an intermolecular reaction. The lipase is
remarkably stable under the reaction conditions and can be recovered and reused
15 times without loss of activity. Competitive lipolysis of triacylglycerols is inhib-
ited by small amounts of fatty acid, allowing the reaction to be carried out on intact
oils (57). Rapeseed oil with 5% of rapeseed fatty acids was converted to epoxidized
rapeseed oil in 91% yield with no hydroxy byproducts. Linseed oil was epoxidized
in 80% yield. Methyl esters are also epoxidized without hydrolysis under these
conditions.
Methyltrioxorhenium (MTO) catalyses direct epoxidation by hydrogen peroxide.
The reaction is carried out in pyridine, avoiding acidic conditions detrimental to
high epoxide yield and uses less concentrated hydrogen peroxide (30%) than other
methods (58). This method epoxidized soybean and metathesized (see Section 7.4)
(1)
O O
i
Mn HO OH
O O erythro
ii
iii OAc iv OH
(2)
O HO HO
threo
Figure 10. Stereochemistry of hydroxylation reactions: (1) with dilute alkaline permanganate
and (2) through epoxide ring opening. (i) KMnO4, NaOH; (ii) m-chloroperbenzoic acid, NaHCO3,
CH2Cl2; (iii) CH3COOH; (iv) base catalyzed hydrolysis.
soybean oil in high yield (59). The epoxidized metathesized oil was more stable to
polymerization than that produced using m-chloroperbenzoic acid, presumably
because it was free of acidic impurities. These and other novel approaches to epox-
idation have recently been reviewed (4, 60, 61). None has yet found industrial
application.
Epoxides are reactive and readily ring open in acid, following protonation of the
epoxy oxygen (Figure 10). This is a route to diols (see Section 4.3), polyols used
in polymer production and a range of a-hydroxy compounds. Ring opening of
methylene-interrupted diepoxides leads to 5 and 6 membered ring ethers through
neighboring group participation (7).
4.3. Hydroxylation
Double bonds are converted to monohydroxy derivatives by acid catalyzed addition
of carboxylic acids, followed by hydrolysis. The carbocation intermediate is prone
to rearrangement, leading to a mixture of positional isomers. Hydroboration with
borane:1,4-oxathiane followed by alkaline hydrolysis a regioselective reaction
(62) has been used to prepare hydroxy fatty acids as GC-MS standards in high
yield (63).
Hydroxylation reactions leading to diols have much in common with epoxida-
tion and oxidative cleavage reactions (see Section 4.4), the end product depending
on the strength of the oxidizing agent. Dilute alkaline permanganate or osmium
tetroxide react through cyclic intermediates resulting from cis addition of the
reagent giving an erythro diol. Ring opening epoxides with acid is a trans addition,
leading to a threo product (Figure 10).
An oxygen bridged manganese complex was recently reported to catalyze
double-bond oxidation by hydrogen peroxide leading to a mixture of epoxide,
cis-diol, and hydroxy ketone products (64). This is an interesting model reaction
for the efficient use of hydrogen peroxide as a cheap hydroxylating agent if the
selectivity can be improved. A number of microorganisms are reported to produce
OXIDATION 23
a range of novel di- and trihydroxy fatty acids and are being investigated as poten-
tial biocatalysts (65).
4.4. Oxidative Cleavage

Double bonds are cleaved by a number of oxidizing agents, converting the olefinic
carbons to carboxylic acids, aldehydes, or alcohols. Fatty acids give a monofunc-
tional product from the methyl end and a difunctional product from the carboxyl
end (along with low-molecular-weight products from methylene-interrupted
systems).
Although now largely superceded by GC and GC-MS methods for structure
determination, oxidative cleavage with ozone or permanganate/periodate and iden-
tification of the resulting products is a powerful method for double-bond location,
particularly for monoenes (19). Reaction with alkaline permanganate/periodate pro-
ceeds through the diol resulting from reaction with dilute permanganate (see Sec-
tion 4.3). The diol is split into two aldehydes by reaction with periodate, and the
aldehydes are subsequently oxidized to carboxylic acids by permanganate. Alterna-
tively, diols derived from double bonds are cleaved to aldehydes by lead tetraace-
tate or periodate.
Ozone reacts directly with double bonds under mild conditions and is the pre-
ferred degradative method for double-bond location (19). The reaction occurs in
several steps (64), starting with a 1,3-dipolar cycloaddition (Figure 11). The addi-
tion product decomposes rapidly into an aldehyde and a carbonyl oxide. In the
absence of solvent or in nonparticipating solvents, these recombine forming a rela-
tively stable 1,2,4-trioxolane or ozonide. The separation into aldehyde and carbonyl
oxide during this rearrangement is supported by production of six ozonide species
from unsymmetrical olefins. Ozonides can be converted to a number of stable pro-
ducts; oxidation yields carboxylic acids, mild reduction gives aldehydes, and treat-
ment with nickel and ammonia gives amines providing useful synthetic routes to
difunctional compounds from fatty acids [e.g., Furniss et al. (67)]. In a carboxylic
acid or alcohol solvent, the carbonyl oxide reacts with the solvent producing mainly
H H H
H H
O H O H
R R R R H
R
O O
O O R O O R O O R
O
O
(1) (2)
H H OR
+ ROH
R O O R O OH
(1) (3)
Figure 11. Ozonolysis reaction mechanism. In nonparticipating solvents, the carbonyl oxide (1)
and aldehyde recombine to give the moderately stable ozonide (2). Hydroperoxides (3) are
formed in protic solvents, and R 00 can be alkyl or acyl.
acyloxy or alkoxyhydroperoxides, respectively, along with other more complex

products (68). These hydroperoxides are oxidized or reduced to the same products
as the ozonides.
Ozonolysis is the only oxidative cleavage that is used industrially. Around
10,000 tons per year of azelaic acid (nonane-1,9-dioic acid) are produced along
with pelargonic acid (nonanoic acid) by ozonolysis of oleic acid. Azelaic acid is
used for polymer production and is not readily available from petrochemical
sources. Other dibasic acids potentially available by this route are brassylic (tride-
cane-1,13-dioc) and adipic (hexane-1,6-dioic) acids from erucic (22:1 13c) and
petroselenic (18:1 6c) acids, respectively. High-purity monoenes are required as
feedstock to avoid excessive ozone consumption and byproducts. Ozonolysis is a
clean reaction, carried out at low temperatures without catalyst. However, ozone
is toxic and unstable, as are the intermediates. Industrial scale ozonolysis is carried
out in pelargonic acid run countercurrent to ozone at 2545 C followed by decom-
position at 60100 C in excess oxygen (69). Ozone must be generated continuously
on-site by electrical discharge in air, and ozone production is the limiting factor for
large-scale production (70).
Ruthenium oxide (RuO4) catalyzes oxidative cleavage of oleic acid to pelargonic
and azelaic acids efficiently in the presence of NaOCl as an oxygen donor to regen-
erate Ru(VIII) (71). However, the production of halogen salt byproducts makes this
impractical for large-scale production. Hydrogen peroxide and peracetic acid are
cheaper and more environmentally benign oxidants, the byproduct from reaction
or regeneration of peracid being water, but give very low yields with RuO4. Ruthe-
nium(III) acetylacetonate (Ru(acac)3) with peracetic acid or Re2O7 with hydrogen
peroxide give moderate yields with internal double bonds, but 80% conversion
with terminal olefins. Terminal olefins, produced from fatty acids with an internal
double bond by metathesis with ethylene, are converted to dibasic acids without
COOR
H2C CH2
metathesis
COOR
HO OH
RuO2/NaOCl
+
H2O2
Re2O7 COOR
CH3CO3H/Ru(acac)3
or
HOOC COOR H2O2/Re2O7
+
COOH HOOC COOR

Figure 12. Alternative oxidative cleavage reactions.
REDUCTION 25
concomitant production of monobasic acids. Diols produced by hydroxylation are

cleaved by Re2O7 with hydrogen peroxide to di- and monobasic acids (Figure 12).
These reactions offer an alternative to ozonolysis for the production of dibasic
acids, but they have still to be optimized for industrial application (71, 72).
5. REDUCTION
Both carboncarbon double bonds and the carboxyl group of fatty acids can be
reduced, either together or separately depending on the reaction conditions. Cata-
lytic reduction is an important industrial route to hardened fats, fatty alcohols, and
fatty amines, using well-established technologies.
5.1. Hydrogenation of Double Bonds

Transition metals such as Co, Ni, Cu, Ru, Pd, and Pt catalyze hydrogenation of dou-
ble bonds. Palladium on charcoal or Adams catalyst (platinum oxide) promote
saturation of fatty acids at ambient temperature and hydrogen pressure. Hydrogena-
tion is accompanied by exchange and movement of hydrogen atoms along the chain
in the region of the double bonds, demonstrated by the large number of isotopomers
formed on deuteration. Homogeneous deuteration with Wilkinsons catalyst (tris
(triphenylphosphine)rhodium(I) chloride) proceeds without hydrogen movement
or exchange (73) and in conjunction with GC-MS analysis is used to locate double
bonds. Partial hydrogenation with hydrazine does not isomerize unreacted double
bonds and is useful for structural analysis of polyenes and was recently used to
examine long-chain metabolites of conjugated linoleic acid (CLA) (74).
5.2. Catalytic Partial Hydrogenation

Partial hydrogenation reduces the polyene content of oils while maintaining or
increasing the monoene content. Reduction of double bonds is accompanied by a
variable degree of cis-to trans-isomerization. Brush hydrogenation of soybean or
rape oil reduces linolenic content, improving oxidative stability, whereas more
extensive hydrogenation increases solid fat content, producing hardened fats
for spreads and shortenings. Partial hydrogenation has been used for the past cen-
tury, in margarine production and remains an important process for edible fat mod-
ification (Chapter xx) despite concerns about adverse nutritional properties of trans-
fatty acids. There are recent reviews of the mechanism (75, 76) and technology
(77).
A number of uncertainties remain about the mechanism of the reaction and the
factors controlling selectivity between polyenes and monoenes, and the balance
between hydrogenation and isomerization. Hydrogenation is a three-phase reaction
among liquid oil, gaseous hydrogen, and solid catalysts carried out as a batch pro-
cess in autoclaves to maintain consistent products. Temperature, hydrogen pressure,
amount and formulation of catalyst, and agitation are all carefully controlled.
Supported nickel is invariably used as catalyst. Although other catalysts are equally
or more effective, nickel has widespread acceptance from long use, ease of removal,
and low cost. Unremoved traces of other metals such as copper might also reduce
the oxidative stability of the product.
The reaction mechanism must account for the selectivity of the reaction (poly-
enes reacting faster than monoenes) and the production of trans-monoenes. Hydro-
gen addition is in two steps with a semihydrogenated intermediate. Addition of the
first hydrogen is reversible, regenerating a double bond with potentially altered
position or geometry. Addition of a second hydrogen irreversibly produces a satu-
rated bond (Figure 13). Dijkstra (76) proposed that for dienes, the formation of the
semihydrogenated intermediate is rate determining and hydrogen concentration
dependent, whereas for the conversion of monoene to saturate, the rate-determining
and hydrogen concentration-dependent step is the addition of the second hydrogen.
At low dissolved hydrogen concentrations, isomerization of monoenes is favored
over saturation, allowing control of the product composition by hydrogen pressure,
agitation, and reaction time.
Copper catalysts show different selectivity compared with nickel. Copper only
catalyzes hydrogenation of methylene-interrupted systems, showing high selectivity
for polyenes and no reaction with oleate or other monoenes produced by reduction
of polyenes. The first step is production of conjugated dienes that are the species
hydrogenated. Dijkstra recently reassessed this reaction, suggesting removal of an
allylic hydrogen as the first step in production of the conjugated diene (78).
catalyst H
(1)
H H
D* + H
D + H DH +H
slow
M
M*
M + H MH +H
slow
S
Figure 13. Partial hydrogenation. The partially hydrogenated intermediate (1) may lead to cis or
trans unsaturated or saturated products. Ddiene; Mmonoene; Ssaturate; potentially
isomerized. Formation of M is favored at a low hydrogen concentration.
PRODUCTION OF SURFACE ACTIVE COMPOUNDS AND OLEOCHEMICALS 27
5.3. Production of Fatty Alcohols

Triacylglycerols, fatty acids, and esters can be reduced to aldehydes, alcohols, or
hydrocarbons, the main application being the production of fatty alcohols. On a
small scale, lithium aluminum hydride (in excess of stochiometric requirement)
is a convenient reducing agent for the carboxyl group without affecting polyunsa-
turated chains. Industrially, catalytic hydrogenation is used and has been reviewed
(79, 80).
Long-chain alcohols are produced from both oleochemical and petrochemical
sources. Oils and fats provide straight chain lengths not readily available otherwise
and the possibility of unsaturated chains. The main feed stocks are coconut and
palm-kernel oil for C12C14 alcohols and technical grades of tallow and palm oil
for C16C18 alcohols. The preferred starting material for catalytic hydrogenation
is methyl ester. Fatty acids are corrosive and need harsh reaction conditions, leading
to unwanted byproducts. Reduction of intact oils leads to loss of glycerol, a valu-
able byproduct, through over-reduction to propane diol and propanol, as well as
excessive hydrogen and catalyst consumption. Methyl esters are reduced to satu-
rated alcohols with copper chromite catalyst (2%) at 250300 C and 25
30-MPa (250300 bar) hydrogen in a suspension system or at 200250 C with a
fixed-bed catalyst. The methanol produced is recycled for methyl ester production.
Zinc-based catalysts do not hydrogenate double bonds and are used to produce
unsaturated alcohols such as oleyl alcohol.
6. PRODUCTION OF SURFACE ACTIVE COMPOUNDS

AND OLEOCHEMICALS
The main non-food use of oils and fats is the production of surfactants. The amphi-
philic properties of fatty acids, exploited for centuries in the use of soaps, can be
modified by changing the carboxyl group into other hydrophilic groupings, giving
anionic, cationic, amphoteric, and nonionic surfactants. There is also scope for
functionalizing the aliphatic chain, but this has not been widely used commercially.
The chain length of the feed stock, C12 C14 from lauric oils, C22 from high erucic
rape and fish oils, and C16 C18 from most other sources, can be used to modify
solubility. The main starting materials for surfactant production are fatty acids
and alcohols with a range of N-containing derivatives produced through amides and
amines. Surfactants of oleochemical origin may biodegrade better than petrochem-
ical products, giving an environmental benefit in addition to being derived from
renewable resources. Recently, surfactants have been produced from fully renew-
able resources. Oleochemical surfactant production has been reviewed (8185).
6.1. Nitrogen-Containing Compounds

The presence of nitrogen, either in a neutral or cationic group, gives surfactant
properties that are not easily produced with other compounds. A diverse range of
nitrogen-containing compounds are produced, for which the starting point is an
TABLE 7. Routes to Nitrogen-Containing Surfactants.
Product
RCH2NH2 CH2O ! (reduction) ! RCH2NMe2 tertiary amine

RCH2CONMe2 ! (reduction) ! RCH2NMe2 tertiary amine
RCH2OH Me2NH ! (catalytic hydrogenation) ! RCH2NMe2 tertiary amine
ROH CH2 CHCN ! RO(CH2)2CN ! (reduction) ! RO(CH2)3NH2 etheramine
RNH2 CH2 CHCN ! RNH(CH2)2CN ! (reduction) ! RNH(CH2)3NH2 diamine
RNH(CH2)3NH2 CH2
CHCN ! RNH(CH2)3NH(CH2)2CN !
triamine
(reduction) ! RNH(CH2)3NH(CH2)3NH2
RO(CH2)3NH2 2nCH2(O)CH2 ! RO(CH2)3N((CH2CH2O)nH)2 ethoxylated
etheramine
RNH(CH2)3NH2 2nCH2(O)CH2 ! RNH(CH2)3N(CH2CH2O)nH)2 ethoxylated diamine
RNH2 nCH2(O)CH2 ! H(OCH2CH2)nN(R)(CH2CH2O)nH ethoxylated amine
RN(Me)2 (H2O2) ! RN(Me)2O amine oxide
RN(Me)2 (MeCl or Me2SO4) ! RN(Me)3 X quaternary amine
R3N (benzyl chloride) ! R3NBz X quaternary amine
RCOOH NH2(CH2)2NH(CH2)2NH2 ! 4 imidazoline
2RCOOH (HOCH2CH2)2NCH3 ! (RCOOCH2CH2)2NCH3 H2O ester amine
amide or amine. Amides are formed by direct reaction of the fatty acid and ammo-
nia at 180200 C and 0.30.7 MPa (37 bar), through dehydration of the initially
formed salt. Long-chain amides, e.g., erucamide, are the principle industrial pro-
ducts, used as polythene film additives.
Amines are produced from fatty acids in a reaction sequence in which the nitrile
is an intermediate. Nitriles are produced by reaction of the fatty acid with ammonia,
giving the amide that is dehydrated in situ at 280360 C in the liquid phase on a
zinc oxide, manganese acetate, or alumina catalyst. Lower temperature and longer
reaction times are used with unsaturated fatty acids to avoid polymerization. Hydro-
genation with nickel or cobalt catalyst reduces the nitrile to amines via the aldimine
(RCH NH). Depending on the reaction conditions, the aldimine reacts with hydro-
gen or primary or secondary amines, giving primary, secondary, or tertiary amines,
respectively, as the major product. Primary amines are produced at 120180 C and
24 MPa (2040 bar); higher temperature and lower pressure favors production of
secondary and tertiary amines with a symmetrical substitution at the nitrogen. The
long-chain composition closely reflects the fatty acid composition of the feedstock,
although hydrogenation conditions can be adjusted to hydrogenate the alkyl chains
or induce cistrans-isomerism. The more widely used unsymmetrical tertiary
amines are produced from primary amines, amides, or alcohols (Table 7). Reactions
converting amines to other surface-active derivatives and for the preparation of
other nitrogen-containing compounds are shown in Table 7. These have appeared
in several reviews (2, 82, 84, 86, 87).
N CH2
RC
N CH2
CH2CH2NH2
4
6.2. Ethoxylation
Long-chain molecules with active hydrogen (alcohols, amines, and amides) react as
nucleophiles with ethylene oxide usually with a basic catalyst. The product has a
hydroxyl group that can react with further ethylene oxide, leading to polyoxyethy-
lene products with a range of molecular weights. The average number of ethylene
oxide molecules added depends on the reaction conditions and can be adjusted to
alter the solubility and surfactant properties of the product.
ROH nC2 H4 O ! ROC2 H4 On H
Typical reaction conditions are 120200 C and pressures of 0.20.8 MPa (28 bar)
with potassium hydroxide or sodium alcoholates as catalyst (83). In the reaction
with primary amines, both active hydrogens are replaced before further ethylene
oxide addition leading to dipolyoxyethylene derivatives. Polyoxyethylenes have a
terminal hydroxyl that may be further functionalized under conditions that do not
damage the ether linkages, for example, sulfation.
6.3. Sulfation
Sulfate esters of alcohols or polyoxyethylene alcohols are prepared by reaction with
sulfur trioxide in continuous falling-film plants, immediately followed by neutrali-
zation with sodium hydroxide to give the sodium salt (81).
ROH SO3 ! ROSO3 H

ROSO3 H NaOH ! ROSO3 Na H2 O
Alcohol sulfates are not stable in acid and are used in alkaline formulations.
C12C16 alcohol sulfates have excellent detergency, high foam, and good wetting
properties. Alcohol sulfates are fully biodegradable under aerobic and anaerobic
conditions and compete in performance with petrochemical-derived linear alkyl-
benzene sulfonates (LABS).
Mono- and diacylglycerols are starting materials for sulfate ester surfactants that
can be prepared directly from triacylglycerols without reduction to the fatty alco-
hol. Cocomonoacylglycerol sulfates, used in cosmetic formulations, are produced
in a solvent-free process (88). Glycerolysis of coconut oil (mole ratio of glycerol
to oil of 2:1) gives the raw material for sulfatization, predominantly mono- and dia-
cylglycerols. Membrane filtration is used to desalt the product.
6.4. a-Sulfonates
The methylene adjacent to the carboxyl group is sufficiently activated to react with
sulfur trioxide, giving a-sulfonate products. As allylic methylenes are similarly
activated, the reaction is usually carried out with saturated starting materials. The
complex reaction involves two moles of sulfur trioxide, giving a disulfonate inter-
mediate that reacts with methyl ester to give the a-sulfonate ester, or on treatment
with sodium hydroxide the disodium salt (81). a-Sulfonates have low toxicity and
are fully biodegradable.
RCH2 COOCH3 2SO3 ! RCHSO3 HCOOSO2 OCH3

RCHSO3 HCOOSO2 OCH3 RCH2 COOCH3 ! 2RCHSO3 HCOOCH3
6.5. Carbohydrate-Based Surfactants

Carbohydrates and related polyols (as well as amino acids) have attracted attention
as the hydrophilic component of nonionic surfactants, particularly as a benign alter-
native to manufacture using ethylene oxide. Sucrose, glucose, and sorbitol (from
hydrogenation of glucose) are available in quantity from renewable resources.
Although sorbitol esters have been in use for many years, large-scale synthesis
of sugar esters remains difficult because of the similar reactivity of all the carbohy-
drate hydroxyls, leading to many molecular species in the product. Further difficul-
ties are the insolubility and charring of the carbohydrate in the reaction medium. A
more controllable reaction is that between long-chain alcohols and glucose, giving
alkyl polyglycosides with the fatty alcohol ether linked only to position C-1 on the
glucose ring. Further glucose units are also joined through ether links. Both the
alcohol and glucose can be produced from renewable resources (oils and fats and
starch, respectively), and the reaction can be carried out in a solvent-free system. In
commercial production, glucose is suspended in excess alcohol and reacted at 100
120 C with a sulfonic acid catalyst. The product has an average degree of polymer-
ization of 1.2 to 1.7 glucose units per molecule (Figure 14) and is nonirritant and
fully biodegradable (8891). Alkyl polyglycoside production is currently 100,000
tons per year, which is used in detergent formulations in place of petrochemical-
derived products.
6.6. Dimers and Estolides

A number of different dimers and oligomers are produced from fatty acids and alco-
hols. These are branched-chain compounds with significantly lower melting points
than straight chain structures of similar molecular weight. Fully saturated dimers
OH
O
OH O
HO
OH O
OH O
HO
OH
y
Figure 14. Alkyl polyglycoside. Degree of polymerization y 1.

have excellent oxidative stability. This and their extended liquid range are exploited
in their use as lubricants and cosmetic additives. Polyfunctional dimers are used in
polymer formulations.
Dimer acids. Dimer acids are produced by heating monoene or diene fatty acids
(e.g., tall oil acids, a byproduct of wood pulping) with a cationic clay catalyst
(92). Typical conditions are 4% montmorillonite at 230 C for 48 hours. After dis-
tillation, the product is a complex mixture of acyclic, cyclic, and bicyclic dimers
along with some trimer. Dimer acids are dibasic and react with diamines and tria-
mines to give polyamides. Imidazole derivatives are used as corrosion inhibitors
and esters as lubricants.
Guerbet compounds. Guerbet alcohols have been known for over a century and are
produced by the alkali catalyzed dimerization of aliphatic alcohols with accompa-
nying loss of water. Typical reaction conditions are heating at 200300 C with
potassium hydroxide in the presence of transition metal compounds to catalyze
the intermediate reduction step. Dehydrogenation of the alcohol to the aldehyde
is followed by aldol condensation and rehydrogenation to give the branched-chain
alcohol (Figure 15a).
The alcohols can be oxidized to the corresponding acids. Guerbet alcohols,
acids, their esters, sulfates, and ether sulfates are used as lubricants, cosmetic addi-
tives, and surfactants. Their synthesis, characterization, and applications have been
reviewed (93).
Estolides. Estolides are ester-linked branched-chain compounds. They are normally
produced under harsh conditions similar to those used to produce dimer acids, but
with the addition of around 10% water. Mono- and polyestolides are used as lubri-
cants, greases, and surfactants, and in cosmetic, ink, and plastic formulations. Esto-
lides biodegrade rapidly and completely, at rates comparable with the vegetable oils
and fatty acids from which they are derived (94), making them environmentally
benign products. The 5 monoene acids in meadowfoam oil form estolides under
OH
(a)
13
O O
13
O O
n
OH
13
(b)
Figure 15. (a) Guerbet alcohol from lauryl alcohol (12:0). (b) Estolide from meadowfoam acids
(20:1 5c).
mild acid catalysis, neighboring group participation by the carboxyl group facilitat-
ing the reaction (Figure 15b) (95). The product from meadowfoam acids shows
higher regioselectivity than that from acids with mid-chain olefins where the double
bond is further from the carboxyl group. Estolides from mid-chain olefins have sig-
nificantly lower pour points than the corresponding fatty acids or triacylglycerols,
but those from meadowfoam acids show little difference.
7. MODIFYING FATTY ACID STRUCTURE
Isomerization and conjugation change the properties of natural methylene-inter-

rupted fatty acids, leading to new applications and potential added value. Chain
shortening or extension produces fatty acids not readily isolated from natural
sources and is also used to introduce radioactive or stable isotope labels. Metathesis
provides a flexible method for modifying the alkyl chain.
7.1. Isomerization
Trans-isomers of fatty acids are more stable thermodynamically than cis-isomers,
because of reduced steric crowding; the equilibrium ratio is approximately 4:1
trans:cis. There is a considerable energy barrier to interconversion (125 kJ/
mole). Before the attached groups can rotate about the double bond, it has to be
weakened by coordination to a catalyst, high temperature, or temporary conversion
to a single bond through addition and elimination reactions. Chemical isomeriza-
tion agents leading to an equilibrium mixture include selenium (through a p-com-
plex) and nitrogen oxides or thiols (through free-radical addition/elimination).
Cis-to trans-isomerization accompanies partial hydrogenation (see Section 5.2)
and may be exploited to raise the melting point. Unwanted isomerization occurs
during physical refining at temperatures above 250 C. More unsaturated acids iso-
merize faster, making linolenic containing seed oils (e.g., soybean and canola) par-
ticularly vulnerable. Conditions for deodorizing rape oil without isomerization have
been optimized following a detailed study and development of a model of the iso-
merization kinetics (96).
7.2. Conjugation
Heating with alkali has long been used to produce conjugated drying oils for paints
and varnishes. The anion resulting from removal of a bis-allylic methylene rear-
ranges through migration and isomerization, giving a cis,trans-conjugated system
(Figure 16). Thus, linoleic acid (18:2 9c12c) gives both 9c11t and 10t12c isomers,
whereas trienes give a mixture of partially and fully conjugated isomers depending
on whether the middle or an outer double bond migrates first. Under the harsh con-
ditions used to prepare drying oils (aqueous alkali at 230 C), a complex mixture
of isomers is eventually formed, but under controlled conditions (e.g., KOH in
MODIFYING FATTY ACID STRUCTURE 33
OH
H2O
Figure 16. Alkali-induced conjugation of methylene-interrupted olefins.
propylene glycol at 150 C), a mixture containing only the 9c11t and 10t12c CLA
isomers is produced (97). This product and individual isomers prepared from the
mixture are used as nutritional supplements.
Thermal isomerization of linoleic acid produces a conjugated isomer mixture
that does not contain all possible cis-and trans-isomers. The absence of the 8c10t
and 11t13c isomers suggests a concerted pericyclic mechanism that limits the geo-
metrical possibilities for the rearranged double bonds (98). [RhCl(C8H14)2]2 in the
presence of (p-CH3C6H4)3P and SnCl2.2H2O is an efficient homogeneous catalyst
for the conjugation of linoleic acid, producing conjugated soybean oil with excep-
tional drying properties and high solvent resistance in high yield (99).
7.3 Chain Shortening and Extension

Fatty acids can be labeled at the carboxyl carbon with 13C or 14C by chain short-
ening followed by chain extension with labeled carbon. Chain shortening to the nor-
halide using the Hunsdieker reaction (decarboxylation of fatty acid silver salts in
the presence of halogens) is only suitable for saturated acids, but unsaturation is
not altered using the alternative developed by Barton employing N-hydroxy-pyri-
dine-2-thione in a halocarbon solvent (100). Chain extension with labeled cyanide
followed by hydrolysis or reaction of the derived Grignard reagent with labeled car-
bon dioxide gives the labeled fatty acid. The Barton decarboxylation was recently
used to prepare gram quantities of 1-[13C ]-linoleic and 1-[13C ]-linolenic acids for
metabolic studies (101).
Two-carbon chain extension at the carboxyl end, mimicking biosynthesis, uses
the malonic ester route (102). After reduction of the carboxyl to an alcohol, the
readily displaced mesylate is prepared and reacted with sodium diethylmalonate.
Saponification and decarboxylation gives the chain extended product in high yield.
H H catalyst H H H H
+
R R R R R R
R R R
LnM C LnM LnM
+
R R R R R R
PR3
Ph Mes N N Mes
Cl Cl
Ru Ph Ru
Cl Cl
PR3 PCy Ph 3
(a) (b)
Figure 17. Olefin metathesis reaction and mechanism. (a) and (b) Grubb catalysts.
This is an efficient route to C20 polyenes, not easily isolated from natural sources,
starting from readily available C18 sources.
Metathesis (see Section 7.4) provides a flexible route to longer and shorter
chains after reaction at a (usually monoene) double bond.
7.4 Olefin Metathesis

Olefin metathesis is the catalytic exchange of groups attached to a double bond. It
presents a number of interesting possibilities for modifying the alkyl chain of fatty
acids (Figure 17).
The mechanism involves a [2,2] cycloaddition between a transition metal alky-
lidene complex and the olefin, resulting in an intermediate metallocyclobutane
(103). The metallocycle breaks in the opposite way to give a new alkylidene and
a new olefin. Repeated exchange at the metal results in an equilibrium mixture
of olefins, usually as an equilibrium mixture of cis-and trans-isomers. The reaction
is used in the petrochemical industry to modify hydrocarbon structure, using cata-
lysts such as WCl6/SnMe4 or Re2O7/Al2O3. These catalysts are less active when
other functional groups compete for the active site, and the application of metath-
esis in oleochemistry has paralleled development of novel catalysts, such as Grubb
catalysts, containing sterically hindered metal alkylidenes (Figure 17a,b).
Self-metathesis describes the reaction of an unsaturated fatty acid with itself.
For example, methyl oleate gives a mixture of starting material (50%), unsaturated
hydrocarbon (25%), and long-chain unsaturated diester (25%), all as a mixture
of cis-and trans-isomers. (Figure 18). The diester can be converted to the musk
component civetone, but a more efficient route is through self metathesis of
the ketone oleon derived from methyl oleate by Claisen condensation (104)
(Figure 18).
MODIFYING FATTY ACID STRUCTURE 35
COOMe 2
+
MeOOC
COOMe
oleon
Re2O7/SiO2, Al 2O3/Bu4Sn
civetone
Figure 18. Self-metathesis reactions.
Cross-metathesis of an unsaturated fatty ester with a normal alkene is a versatile

way of producing chain-shortened or chain-extended homologues leading to oleo-
chemicals with chain lengths outside the C16 C22 range of most commodity oils.
Methyl oleate reacts with hex-3-ene, in large excess to suppress self-metathesis and
push the reaction toward the C12 ester and hydrocarbon products. o-Olefins may be
chain extended similarly, the ethene produced being removed to drive the reaction
to completion. Cross metathesis provides a route to compounds otherwise difficult
to obtain, for example, triacontanol from reduction of the product from methyl eru-
cate and 1-octadecene. Ethenolysis (cross-metathesis with ethene) produces shorter
chain o-olefins with a wide range of applications. A high pressure of ethene is used
to force the reaction to the desired products. o-Olefins produced either by metath-
esis or from pyrolysis of castor oil can be coupled to give long-chain dibasic acids
(105).
Metathesis of intact oils produces polymeric products resulting from intra- and
intermolecular bond formation, and they can be used to produce high-viscosity
stand oils from drying oils without the loss of double bonds that occurs on thermal
polymerization. Vegetable oils can be metathesized efficiently at low temperature
and pressure using Grubbs ruthenium catalyst (Cy3P)2Cl2Ru
CHPh, without the

rigorous exclusion of water and oxygen required with WCl6/SnMe4 (106). Pretreat-
ment of the oil with silica gel may be required.
As a reaction with 100% atom efficiency achieved at moderate temperature
(<100 C) using renewable resources, metathesis has potential in a sustainable
chemical industry. A recently developed catalyst (Figure 17b) has an efficiency that
justifies industrial application in the production of fine chemicals (106). The hydro-
carbon byproducts of metathesis, for example, a-olefins, are also valuable starting
materials. Metathesis in oleochemistry, in the context of green chemistry, has
recently been reviewed (107).
8. NOVEL CHEMISTRY FOR FUNCTIONALIZING

THE ALKYL CHAIN
Oils and fats are renewable resources for the chemical industry. Increasing the
range of oleochemicals that can be produced could add value to existing crops
and provide a market for new crops, driving research into novel fatty acid deriva-
tives. Most current oleochemical production involves reaction at the carboxyl
group, with the chain length and unsaturation of the alkyl chain chosen to give
the desired melting behavior or hydrophobicity. Introducing functionality to the
alkyl chain through radical, electrophilic, nucleophilic, pericyclic, and transition
metal catalyzed addition to carboncarbon double bonds leads to novel compounds
with commercial potential. Only a small selection of recent research is illustrated
here, focusing on three promising approaches: neighboring group participation,
Friedel Crafts acylation, and free-radical addition reactions.
Functionalizing the alkyl chain places more emphasis on the structure of the
fatty acids used as feedstock. Model reactions use single fatty acids, often mono-
enes with particular double-bond positions. Large-scale use of these reactions needs
oils rich in single fatty acids to maintain the purity of the product and minimize
wasteful side reactions. Suitable feedstocks may be current crops such as high oleic
or high erucic varieties or new crops with unusual fatty acids (Chapter xx). Petro-
selenic acid (18:1 6c) from umbelliferae oils and 5-eicosenoic acid (20:1 5c) from
meadowfoam oil are of particular interest as distinctive products can result from
neighboring group participation. Breeding to increase the monoene content of
some oils may be desirable. o-Olefins are useful starting materials; 10-undecenoic
acid is available from pyrolysis of castor oil, and others may be produced by
metathesis (see Section 7.4). Recent, wide-ranging reviews of this area are available
(4, 5, 108)
8.1. Neighboring Group Participation

Neighboring group participation is the involvement of a nearby functional group in
the reaction of another functional group. It may influence the regioselectivity of the
reaction or lead to specific products, often as a result of cyclization to five-and six-
membered rings. Neighboring group participation reactions of fatty acids were
reviewed recently (109) and can be used to introduce mid-chain functionality,
including heterocyclic groups. Double bonds and the carboxyl group usually react
independently of each other, but 4 and 5 bonds may interact with the carboxyl
through neighboring group participation leading to g- and d-lactones (five- and
NOVEL CHEMISTRY FOR FUNCTIONALIZING THE ALKYL CHAIN 37
OH
HClO4
CH2Cl2
O O
HX ROH
H+ Lewis acid
OH O OR O
X OR
14 14
Figure 19. Neighboring group participation leading to lactones and other products from 5
acids. X OH, RO, or RNH.
six-membered rings, respectively). The 5 acids from meadowfoam oil readily

form lactones when refluxed with perchloric acid. The proportion of d- and
g-lactones depends on the solvent: 6:1 in hexane and 40:1 in dichloromethane.
The d-lactone is formed faster, but the g-lactone is the more thermodynamically
stable isomer. High dilution and a nonparticipating polar solvent that stabilizes
the intermediate cation favor kinetic control of the reaction (110). The lactones
can be ring opened by treatment with water, alcohols, and amines in acid, giving
4- and 5-hydroxy acids, esters, and amides (111); alternatively, treatment with an
alcohol and a Lewis acid catalyst under more vigorous conditions results in an alkyl
group ether linked to the chain (112) (Figure 19).
8.2. Friedel Crafts Acylation

Friedel Crafts acylation with an acyl chloride and Lewis acid catalyst is more often
associated with aromatic compounds. Ethylaluminium dichloride (EtAlCl2) is an
effective catalyst for the acylation of aliphatic olefins, including fatty acids and
alcohols, giving b,g-unsaturated ketones (113). The reaction occurs with both
terminal and internal double bonds, with the acyl group becoming attached to
one of the double-bond carbons while the double bond migrates one carbon. Reac-
tion at terminal olefins is regiospecific with addition to the terminal carbon giving a
linear product and a predominantly trans-double bond. Internal double bonds give
an approximately equal mixture of trans-regioisomers (Figure 19). a,b-Unsaturated
acid chlorides give allyl vinyl ketones that undergo Nazarov cyclization to prosta-
glandin- and jasmonate-like molecules (Figure 20) (114). Neighboring group parti-
cipation in petroselenic acid (18:1 6c) leads to intramolecular cyclization (115).
Friedel Crafts acylation is a flexible route to new and highly functionalized oleo-
chemicals containing reactive allyl keto functions (115).
O O
11
+
OH R Cl
10
EtAlCl2
O
9
R
OH
10
O
COOH
10 9
Cl
EtAlCl2
O
O
11
COOH
10
(+ regioisomer)
H3PO4/HCOOH
O
COOH
(+ regioisomer)
Figure 20. Friedel Crafts acylation reactions.
8.3. Free Radical Addition Reactions

Double bonds participate in free radical addition reactions, and these can be of syn-
thetic use in introducing functional groups (116). A particularly simple reaction is
the preparation of g-lactones by solvent-free addition of 2-halocarboxylates to fatty
esters, catalyzed by commercial copper powder at 100130 C (117). Iodides are
most reactive and can be prepared in situ from more readily available bromides
and sodium iodide (Figure 21).
Perfluoro alkyl iodides add to both terminal and internal double bonds when the
reaction is initiated by electron transfer from metals such as finely divided silver,
copper powder, and lead with copper acetate. Using an o-olefin and a perfluoro-
alkyl-a,o-diiodide, a perfluoro group can be inserted into a long-chain compound
(118) (Figure 21). Deiodination of the product by catalytic reduction results in
highly hydrophobic alkyl chains with interesting surfactant properties.
REFERENCES 39
O
11
+ COOMe
OH
10
I
Cu
OH
O
O
O
11
2 OMe + I(CF2)nI
10
Pb/Cu(OAc) 2
I I
(CF2)n
MeOOC COOMe
7 7
H2
Pd/C
(CF2)n
MeOOC COOMe
Figure 21. Radical addition reactions.
REFERENCES
1. C. C. Akoh and D. B. Min, eds., Food Lipids: Chemistry, Nutrition, and Biotechnology,
2nd ed., Marcel Dekker, Inc., New York, 2002.
2. F. D. Gunstone, Eur. J. Lipid Sci. Technol., 103, 307314 (2001).
3. F. D. Gunstone and R. J. Hamilton, eds., Oleochemical Manufacture and Applications,
Sheffield Academic Press, Sheffield, U.K., 2001.
4. U. Biermann, W. Friedt, S. Lang, W. Luhs, G. Machmuller, J. O. Metzger, M. Rusch gen.
Klaas, H. J. Schafer, and M. P. Schneider, Angew. Chem. Int. Ed., 39, 22062224 (2000).
5. G. Knothe and J. T. P. Derksen, eds., Recent Developments in the Synthesis of Fatty Acid
Derivatives, AOCS Press, Champaign, Illinois, 1999.
6. F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and Applications, Marcel
Dekker, Inc., New York, 1997.
7. A. Karleskind, ed., Oils and Fats Manual, Vols. 1 and 2, Intercept Ltd., Andover, U.K.,
1996.
8. F. D. Gunstone, Fatty Acid and Lipid Chemistry, Blackie Academic and Professional,
London, U.K., 1996.
9. F. D. Gunstone, J. L. Harwood, and F. B. Padley, eds., The Lipid Handbook, 2nd ed.,
Chapman and Hall, London, U.K., 1994.
10. H. Baumann, M. Buhler, H. Fochem, F Hirsinger, H. Zoebelein, and J. Falbe, Angew.
Chem. Int. Ed., 27, 4162 (1988).
11. Official Methods and Recommended Practices of the American Oil Chemists Society,
Section I, American Oil Chemists Society, Champaign, Illinois, 1997.
12. E. Ucciani, Dictionnaire des Huiles Vegetales, Lavoisier Publishing, Paris, 1995.
13. R. O. Adlof and F. D. Gunstone. (2001). Available: http://www.aocs.org/member/
division/analytic/fanames.htm.
14. M. J. Parnham, Inform 7, 11681175 (1996).
Fifth Edition, Sections Cd 1 and Cd 3, American Oil Chemists Society, Champaign,
Illinois, 1998.
16. H. Li, F. R. van de Voort, A. A. Ismail, J. Sedman, R. Cox, C. Simard, and H. Buijs, J.
Amer. Oil Chem. Soc., 77, 2936 (2000).
17. H. Li, F. R. van de Voort, J. Sedman, and A. A. Ismail, J. Amer. Oil Chem. Soc., 76, 491
497 (1999).
18. M. H. Gordon, in M. Jee, ed., Oils and Fats Authentication, Blackwell, Oxford, 2002,
pp. 143155.
19. W. W. Christie Lipid Analysis, The Oily Press, Bridgwater, UK, 3rd Edition, in press
2002.
20. N. O. V. Sonntag, in R. W. Johnson and E. Fritz, eds., Fatty Acids in Industry, Marcel
Dekker, Inc., New York, 1988. pp. 2372.
21. D. R. Kodali, J. Lipid Res., 28, 464469 (1987).
22. F. D. Gunstone, in F. D. Gunstone, ed., Lipid Synthesis and Manufacture, Sheffield
Academic Press, Sheffield, U.K., 1999, pp. 321346.
23. W. W. Christie. (2002, April 29). The Lipid Library. Available: http://www.lipid.co.uk/
infores/select.html.
24. Y. M. Choo, A. N. Ma, and A. S. H. Ong, in F. D. Gunstone and F. B. Padley, eds., Lipid
Technologies and Applications, Marcel Dekker, Inc., New York, 1997, pp. 771773.
25. G. Knothe and R. O. Dunn, in F. D. Gunstone and R. J. Hamilton, eds., Oleochemical
Manufacture and Applications, Sheffield Academic Press, Sheffield, U.K., 2001,
pp. 128132.
26. K.-L. Lee, T. A. Foglia, and Y.-S. Chang, J. Amer. Oil Chem. Soc., 79, 191195 (2002).
27. Y. Watanabe, Y. Shimada, A. Sugihara, and Y. Tominga, J. Amer. Oil Chem. Soc., 78,
703707 (2001).
28. A. Rozendaal and A. R. Macrae, in F. D. Gunstone and F. B. Padley, eds., Lipid
Technologies and Applications, Marcel Dekker, Inc., New York, 1997, pp. 223245.
29. D. Rousseau and A. G. Marangoni, in C. C. Akoh and D. B. Min, eds., Food Lipids:
Chemistry, Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, Inc., New York, 2002,
pp. 301333.
30. F. D. Gunstone, J. Sci. Food Agric., 79, 15351549 (1999).
31. X. Xu, Eur. J. Lipid Sci.Technol., 102, 287303 (2000).
32. C. F. Torres, E. Barrios, and C. G. Hill, J. Amer. Oil Chem. Soc., 79, 457466 (2002).
REFERENCES 41
33. C. C. Akoh and X. Xu, in T. M. Kuo and H. W. Gardner, eds., Lipid Biotechnology,
Marcel Dekker, Inc., New York, 2002, pp. 461469.
34. M. W. Christensen, L. Andersen, O. Kirk, and H. C. Holm, Lipid Technol. Newslett., 7,
3337 (2001).
35. P. M. Nielsen, Oils and Fats International, July, 1819 (2002).
36. Y. Shimada, N. Fukushima, H. Fujita, Y. Honda, A. Sugihara, and Y. Tominga, J. Amer.
Oil Chem. Soc., 75, 15811586 (1998).
37. Y. Shimada, A. Sugihara, M. Shibahiraki, H. Fujita, H. Nakano, T. Nagao, T. Terai, and
Y. Tominaga, J. Amer. Oil Chem. Soc., 74, 14651470 (1997).
38. Y.-H. Ju and T.-C. Chen, J. Amer. Oil Chem. Soc., 79, 2932 (2002).
39. Y. Shimada, K. Maruyama, A. Sugihara, T. Baba, S. Komemushi, S. Moriyama, and
Y. Tominaga, J. Amer. Oil Chem. Soc., 75, 15651571 (1998).
40. S. A. Spurvey, S. P. J. N. Senanayake, and F. Shahidi, J. Amer. Oil Chem. Soc., 78, 1105
1112 (2001).
41. E. N. Frankel, Lipid Oxidation, The Oily Press, Dundee, U.K., 1998.
42. E. N. Frankel, Lipid Oxidation, The Oily Press, Dundee, U.K., 1998, pp. 19 and 4354.
43. E. N. Frankel, J. Oleo Sci., 50, 387391 (2001).
44. K. Miyashita, Lipid Technol. Newslett., 8, 3541 (2002).
45. T. I. Hamalainen, S. Sundberg, T. Hase, and A. Hopia, Lipids, 37, 533540 (2002).
46. A. Bodin, M. Linnerborg, J. L. G. Nilsson, and A. T. Karlberg, J. Surfactants Deterg., 5,
107110 (2002).
47. H. R. Rawls and P. J. Van Santen, J. Amer. Oil Chem. Soc., 47, 121125 (1970).
48. F. H. Doleiden, S. R. Farenholtz, A. A. Lamola, and A. M. Trozzolo, Photochem.
Photobiol., 20, 519521 (1974).
49. D. B. Min, in C. C. Akoh and D. B. Min eds., Food Lipids: Chemistry, Nutrition and
Biotechnology, Marcel Dekker, Inc., New York, 1998, pp. 289292.
50. N. Macfarlane, J. Salt, R. Birkin, and A. Kendrick, Inform, 12, 244249 (2001).
51. P. Schuler, in B. J. F. Hudson ed., Food Antioxidants, Elsevier, London, U.K., 1999,
pp. 138141.
52. E. A. Decker and D. J. McClements, Inform, 12, 251256 (2001).
53. P. D. Bartlett, Rec. Chem. Prog., 11, 4751 (1950).
54. S.-P. Chang, J. Amer. Oil Chem. Soc., 56, 855856 (1979).
55. K. D. Karlson, R. Kleiman, and M. O. Bagby, J. Amer. Oil Chem. Soc., 71, 175182 (1994).
56. S. Warwel and M. Rusch gen. Klass, J. Mol. Catal. B, 1, 2935 (1995).
57. M. Rusch gen. Klass and S. Warwel, J. Amer. Oil Chem. Soc., 73, 14531457 (1996).
58. J. Rudolph, K. L. Reddy, J. P. Chiang, and K. B. Sharpless, J. Amer. Chem. Soc., 119,
61896190 (1997).
59. M. D. Refvik and R. C. Larock, J. Amer. Oil Chem. Soc., 76, 99102 (1999).
60. M. Rusch gen. Klass and S. Warwel in G. Knothe and J. T. P. Derksen, eds., Recent
Developments in the Synthesis of Fatty Acid Derivatives, AOCS Press, Champaign,
Illinois, 1999, pp. 157167.
61. G. J. Piazza, in G. Knothe and J. T. P. Derksen, eds., Recent Developments in the
Synthesis of Fatty Acid Derivatives, AOCS Press, Champaign, Illinois, 1999, pp. 182
195.
62. H. C. Brown and A. K. Mandal, Synthesis, 153155 (1980).

63. R. Wilson, R. Smith, P. Wilson, M. J. Shepherd, and R. A. Riemersma, Anal. Biochem.,
248, 7685 (1997).
64. J. Brinksma, L. Schmieder, G. van Vliet, R. Boaron, R. Hage, D. E. DeVos, P. L. Alsters,
and B. L. Feringa, Tetrahedron Lett., 43, 26192622 (2002).
65. C. T. Hou, T. M. Kuo, and A. C. Lanser, Inform, 13, 307316 (2002).
66. R. Criegee, Angew. Chem. Int. Ed., 14, 745752 (1975).
67. B. S. Furniss, A. J. Hannaford, P. W. G. Smith, and A. R. Tatchell, Vogels Textbook of
Practical Organic Chemistry, Longman, Harlow, U.K., 1989, pp. 492494.
68. L. Rebrovic, J. Amer. Oil Chem. Soc., 69, 159165 (1992).
69. L. Rebrovic and F. D. Gunstone, Lipid Technol., 8, 135137 (1996).
70. H.-P. Klein, J. Amer. Oil Chem. Soc., 61, 306312 (1984).
71. S. Warwel and M. Rusch gen. Klass, Lipid Technol., 9, 1014 (1997).
72. M. Rusch gen. Klass and S. Warwel, in G. Knothe and J. T. P. Derksen, eds., Recent
Illinois, 1999, pp. 167173.
73. B. F. Dickens, C. S. Ramesha, and G. A. Thompson, Anal. Biochem., 127, 3748
(1982).
74. J. L. Sebedio, P. Juaneda, S. Gregoire, J. M. Chardigny, J. C. Martin, and C. Ginies,
Lipids, 34, 13191325 (1999).
75. A. J. Dijkstra, in K. K. Rajah, ed., Fats in Food Technology, Sheffield Academic Press,
Sheffield, U.K., 2002, pp. 123141.
76. A. J. Dijkstra, Inform, 8, 11501158 (1997).
77. W. T. Koetsier, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and
Applications, Marcel Dekker, Inc., New York, 1997, pp. 265303.
78. A. J. Dijkstra, Eur. J. Lipid Sci. Technol., 104, 2935 (2002).
79. U. R. Kreutzer, J. Amer. Oil Chem. Soc., 61, 343348 (1984).
80. R. W. Johnson, in R. W. Johnson and E. Fritz, eds., Fatty Acids in Industry, Marcel
Dekker, Inc., New York, 1988, pp. 217231.
81. M. R. Porter, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and
82. A. D. James, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and
83. G. Bogndo, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and Applica-
tions, Marcel Dekker, Inc., New York, 1997, pp. 633694.
84. R. Franklin, M. D. Hoey, and J. Zachwieja, in F. D. Gunstone and R. J. Hamilton, eds.,
Oleochemical Manufacture and Applications, Sheffield Academic Press, Sheffield,
U.K., 2001, pp. 2354.
85. D. W. Roberts, in F. D. Gunstone and R. J. Hamilton, eds., Oleochemical Manufacture
and Applications, Sheffield Academic Press, Sheffield, U.K., 2001, pp. 5573.
86. S. Billenstein and G. Blaschke, J. Amer. Oil Chem. Soc., 61, 353357 (1984).
87. L. Dahlgren, Inform, 11, 508512 (2000).
88. K. Hill, Agro-Food Industry Hi-Tech., September/October, 915 (1998).
89. K. Hill and O. Rhode, Fette/Lipid, 101, 2533 (1999).
REFERENCES 43
90. K. Hill, W. von Rybinski, and G. Stoll, eds., Alkyl Polyglycosides, VCH, Weinheim,
FRG, 1997.
91. W. von Rybinski and K. Hill, Angew. Chem. Int. Ed., 37, 13281345.
92. R. W. Johnson, in R. W. Johnson, and E. Fritz, eds., Fatty Acids in Industry, Marcel
93. G. Knothe, Lipid Technol., 14, 101104 (2002).
94. S. M. Erhan and R. Kleiman, J. Amer. Oil Chem. Soc., 74, 605607 (1997).
95. T. A. Isbell and R. Kleiman, J. Amer. Oil Chem. Soc., 73, 10971107 (1996).
96. G. Henon, Z. Kemeny, K. Recseg, F. Zwobada, and K. Kovari, J. Amer. Oil Chem. Soc.,
76, 7381 (1999).
97. A. Sb, Lipid Technol. Newslett., 7, 913 (2001).
98. F. Destaillats and P. Angers, Lipids, 37, 435438 (2002).
99. R. C. Larock, X. Dong, S. Chung, C. K. Reddy, and L. E. Ehlers, J. Amer. Oil Chem. Soc.,
78, 447453 (2001).
100. D. H. R. Barton, D. Critch, and W. B. Motherwell, Tetrahedron Lett., 24, 49794982
(1983).
101. O. Loreau, A. Maret, D. Poullain, J. M. Chardigny, J.L. Sebedio, B. Beaufrere, and J. P.
Noel, Chem. Phys. Lipids, 106, 6578 (2000).
102. F. Spener and H. K. Mangold, Chem. Phys. Lipids, 11, 215218 (1973).
103. R. Toreki. (2002, July 3). Organometallic HyperTextBook. Available http://www.ilpi.
com/organomet/olmetathesis.html.
104. M. F. C. Plugge and J. C. Mol, Synlett, 507508 (1991).
105. S. Warwel, H. G. Jagers, and S. Thomas, Fat Sci. Technol., 94, 323328 (1992).
106. M. D. Refvik, R. C. Larock, and Q. Tian, J. Amer. Oil Chem. Soc., 76, 9398 (1999).
107. J. C. Mol, Green Chem., 4, 513 (2002).
108. U. Biermann, S. Furmeier, and J. O. Metzger, in F. D. Gunstone and R. J. Hamilton, eds.,
Oleochemical Manufacture and Applications, Sheffield Academic Press, Sheffield,
U.K., 2001, pp. 266299.
109. F. D. Gunstone, in G. Knothe and J. T. P. Derksen, eds., Recent Developments in the
Synthesis of Fatty Acid Derivatives, AOCS Press, Champaign, Illinois, 1999, pp. 119.
110. T. A. Isbell and B. A. Plattner, J. Amer. Oil Chem. Soc., 74, 153158 (1997).
111. T. A. Isbell and B. A. Steiner, J. Amer. Oil Chem. Soc., 75, 6366 (1998).
112. T. A. Isbell and M. S. Mund, J. Amer. Oil Chem. Soc., 75, 10211029 (1998).
113. J. O. Metzger and U. Biermann, Liebigs Ann. Chem., 645650 (1993).
114. J. O. Metzger and U. Biermann, Fette/Lipid, 100, 26 (1998).
115. U. Biermann and J. O. Metzger, in G. Knothe and J. T. P. Derksen, eds., Recent
Illinois, 1999, pp. 8089.
116. J. O. Metzger, R. Mahler, G. Francke, and A. Hayen, in G. Knothe and J. T. P. Derksen,
eds., Recent Developments in the Synthesis of Fatty Acid Derivatives, AOCS Press,
Champaign, Illinois, 1999, pp. 9099.
117. J. O. Metzger, R. Mahler, and G. Francke, Liebigs Ann./Recueil, 23032313 (1997).
118. J. O. Metzger, R. Mahler, and A. Schmidt, Liebigs Ann. Chem., 693696 (1996).
2
Crystallization of
Fats and Oils
Serpil Metin1 and Richard W. Hartel2
1
Cargill Inc.
Minneapolis, Minnesota
2
University of Wisconsin
Madison, Wisconsin
1. INTRODUCTION
1.1. Control of Lipid Crystallization

In many food products and even some processing operations, it is important to be
able to control lipid crystallization to obtain the desired number, size distribution,
polymorph, and dispersion of the crystalline phase. In most foods, it is crystalliza-
tion of triacylglycerols (TAG) that is most important, although, at times, crystalli-
zation of other lipids (i.e., monoacylglycerols, diacylglycerols, phospholipids, etc.)
may also be important to product quality.
Proper control of the crystalline microstructure leads to products with the
desired textural properties and physical characteristics. For example, tempering
of chocolate prior to molding or enrobing is designed to control crystallization of
the cocoa butter into a large number of very small crystals that are all in the desired
polymorphic form. When controlled properly, the cocoa butter crystals in chocolate
contribute to the desired appearance (shine or gloss), snap, flavor release, melt-
down rate upon consumption, and stability during shelf life (fat bloom). Similar
45
46 CRYSTALLIZATION OF FATS AND OILS
arguments can be made for other products such as butter, margarine, whipped
cream, ice cream, shortening, peanut butter, and a host of others.
During processing of fats, crystallization is often used to modify the properties
of the fat. For example, winterization of vegetable oils is needed to ensure that the
oil remains a clear liquid even when stored at low temperatures for extended time
periods. The process of fractionation of fats to produce components of natural fats
with different melting properties also requires control of crystallization to optimize
the separation process. Many fats, including palm oil, palm-kernel oil, milk fat, and
tallow, are fractionated by crystallization to produce different functional fats.
1.2. Crystallization of Natural Fats

There are several aspects of lipid crystallization that make it unique from crystal-
lization of other components in foods (like water, sugars, salts, etc.). These are
related to the complex molecular composition of natural fats and the orientation
of the triacylglycerol molecules.
Fats are made up primarily of TAGs, approximately 98%, with the remainder of
the fat being more polar lipids like diacylglycerols (DAGs), monoacylglycerols
(MAGs), free fatty acids (FFAs), phospholipids, glycolipids, sterols, and other minor
components. In refined fats, these minor lipids are much lower in concentration than
in unrefined fats. Although the TAGs form the main crystalline phase, the minor
components, or impurities, can often play a large role in how crystallization
occurs and crystallization may be substantially different in a refined oil than in
the unrefined starting material.
Natural fats also contain a wide range of TAG species with fatty acids of differ-
ent chain length and degree of unsaturation. Milkfat, for example, contains hun-
dreds of different TAG species with no single species present at greater than
about 5%. TAGs are composed of three fatty acids arranged on a glycerol molecule,
and with variations in chain length and degree of saturation of the fatty acids, a
wide range of components is possible. This range of composition leads to interest-
ing complexities in crystallization.
The nature of the TAG molecule is such that it can often take multiple forms in a
crystal lattice. That is, the same molecule can crystallize into different crystalline
forms dependent on processing conditions. The phenomenon is called polymorph-
ism. Although there are numerous molecules that exhibit polymorphism in nature
(many in the pharmaceutical field), polymorphism is somewhat unique to lipids in
the food industry (although some sugar alcohols also form polymorphs).
In this chapter, the complex nature of lipid crystallization, primarily related to
TAG, will be discussed.
2. LIPID PHASE BEHAVIOR
2.1. Nature of the Liquid Phase

It is important to understand the nature of the liquid phase prior to crystallization to
understand how crystals form. It is widely recognized that lipids retain some degree
LIPID PHASE BEHAVIOR 47
of ordering in the liquid phase, with temperatures well above the melting point
needed to fully dissociate this ordering. When melting fats, this liquid ordering
is termed a crystalline memory effect, where subsequent recooling leads to forma-
tion of a different (usually more stable) phase than would occur if the fat was heated
to higher temperatures to destroy the liquid memory (13).
In nucleation, or the formation of the crystalline phase from the liquid, some
organization of molecules is expected. In lipids, the natural ordering of the liquid
phase leads to crystal formation. In fact, rapid cooling of liquid lipids results in the
formation of a diffuse crystalline phase (low-energy polymorph) because of the
ordering structure in the liquid phase. Such rapid cooling of other systems, most
Figure 1. Proposed mechanism (highly schematic) for nucleation of triacylglycerols (TAGs).

Straight chains indicate crystallized TAGs, whereas bent chains indicate fluid TAGs (4).
notably sugars and starches, often results in the formation of a glassy state consist-
ing of molecules that are randomly organized together with no long-term ordering.
Upon slower cooling from the liquid, the lipid molecules have time to organize
into lamellae (1) and eventually can form coherent, three-dimensional crystals
(shown schematically in Figure 1). The arrangement of the molecules into the crys-
talline state depends on such factors as the cooling rate, the temperature at which
crystallization occurs, the agitation rate, and the composition of the lipid phase.
2.2. Polymorphism
Polymorphism is the ability of a molecule to take more than one crystalline form
depending on its arrangement within the crystal lattice. In lipids, differences in
hydrocarbon chain packing and variations in the angle of tilt of the hydrocarbon
chain packing differentiate polymorphic forms. The crystallization behavior of
TAG, including crystallization rate, crystal size, morphology, and total crystallinity,
are affected by polymorphism. The molecular structure of the TAG and several
external factors like temperature, pressure, rate of crystallization, impurities, and
shear rate influence polymorphism (5).
TAGs are oriented in a chair or tuning fork configuration in the crystalline lat-
tice. The TAG can take either a double or triple chain-length structure as seen in
Figure 2. The fatty acids of TAG pairs overlap in a double chain-length structure
whereas in triple chain packing, the fatty acids do not overlap. The height of these
chair structures and the distance between the molecules in the chair structures
are found by using the X-ray spectra as the long and short spacings, respectively.
Figure 2. Packing arrangements of triacylglycerol molecules in the crystal lattice (4).

TABLE 1. Identification of Polymorphic Forms of Fats Based on X-ray Analysis of Short

Spacings (6).
Polymorphic Form Unit Cell Lines and Short Spacings (A )
a Hexagonal A single strong and very broad @ 4.15

b0 Orthorhombic Two strong lines @ 4.2 and 3.8
b Triclinic A strong line @ 4.6
The polymorphic forms of fats are often simply classified into three categories,
a, b0 , and b, in increasing order of stability. The a form is the least stable poly-
morph with the lowest melting point and latent heat of fusion. The b form is the
most stable, with the highest melting point and latent heat. Each polymorphic
form has distinct short spacings (the distances between parallel acyl groups on the
TAG) that are used to distinguish the polymorphic forms based on their X-ray dif-
fraction patterns, as summarized in Table 1. Based on the unique configuration of
the molecules within the crystal lattice, each polymorph has a different crystallo-
graphic unit cell, also shown in Table 1.
In general, TAGs with three saturated fatty acids crystallize in double chain-
length packing, whereas triple chain-length packing is obtained if the TAG contains
fatty acids with different structures (chain length and unsaturation). Lutton (7) sta-
ted that if the fatty acids of a TAG differ in length by more than four carbons, it
forms a triple chain-length structure. Triple chain-length packing is also observed
in TAG containing a cis-unsaturated fatty acid because this causes a kink in the
structure, as seen in Figure 2. Cis-unsaturated fatty acids do not mix in one layer
with saturated fatty acids, and triple chain-length crystals are formed (8). It should
be noted that trans-unsaturated fatty acids incorporate into a crystal structure in the
same way as the saturated fatty acids (8). The chain-length structure influences
the mixing-phase behavior of different types of TAGs in solid phases (5). The triple
chain-length structure has greater long spacings than does the double chain-length
structure.
Lipids exhibit monotropic polymorphism, where unstable forms are the first to
crystallize in a subcooled fat because of their lower energy state, according to the
Gibbs free energy (5). Subsequent transformation of unstable polymorphs into
more stable forms occurs over time until, eventually, the most stable polymorph
for a given lipid is reached. Transformation of unstable to stable polymorphs can
be achieved by a slight increase in temperature above the melting point of the less-
stable forms. This increase in temperature first causes the melting of the unstable
forms and then solidification in a more stable form. Transformation to a more stable
form can also take place without melting as seen in Figure 3. The difference in
Gibbs, free energy between polymorphs is the driving force for this transformation,
as the molecules become more tightly arranged in the crystal lattice. It is assumed
that the chair structure is maintained during polymorphic transformations (9). The
layer arrangement of the a polymorph does not change when it is transformed to the
b0 polymorph, although its lateral chain packing and angle of tilt changes during
polymorphic transformation.
Figure 3. Monotropic polymorphism of lipids where (Tm)a, (Tm)b0 , and (Tm)b are the melting
temperatures of the a, b0 , and b polymorphs, respectively.
The hydrocarbon chain packing of the b polymorph is denser than that of the a
polymorph. The denser chain packing in the b polymorph gives increased stability
compared with the a polymorph. In addition, stable polymorphs have higher melt-
ing point and higher heat of fusion than the less-stable forms. The different poly-
morphic forms typically crystallize at rates in order of their stability (a < b0 < b).
Thus, the least-stable polymorphic form typically crystallizes first in a strongly
subcooled molten fat because of the lower surface energy (10).
The rate of polymorphic transformation depends on the length of the fatty acid
chain and is the greatest for TAGs with short-chain fatty acids (10). Natural fats
usually contain a large number of TAGs; thus, the transformation of unstable to
stable forms is often very slow. As mentioned previously, the a form is generally
formed first in a rapidly cooled liquid fat, but it is usually very unstable and rapidly
transforms to the b0 form. The b0 form may remain for an extended time (hours to
days), although in many fats, it eventually transforms into the b polymorph, which
is usually the most stable form. However, in many natural fats, the b0 polymorph
can exist for long periods of time because of compound or solid solution formation
(11). That is, in some mixed-acid TAGs, no b polymorph may form and b0 is the
most stable. In other cases, two b forms may be present (5). For example, SOS, a
mixed-acid TAG, has five polymorphic forms in which two b forms are present. The
molecular structures of the five polymorphic forms have been identified using XRD,
differential scanning colorimetry (DSC), and Fourier-transformed infrared spectro-
scopy (FT-IR) techniques (5). In addition, two liquid crystalline phases called LC1
and LC2 were found for SOS using time-resolved synchrotron radiation X-ray
TABLE 2. Polymorphic Forms of Cocoa Butter.
Melting Temperature ( C)
Form Wille and Lutton (13) Davis and Dimick (13)
I g 17.3 13.1
II a 23.3 17.7
III b0 2 25.5 22.4
IV b0 1 27.5 26.4
V b2 33.8 30.7
VI b 36.3 33.8
diffraction (SR-XRD) analysis (12). The researchers stated that the crystallization
properties of SOS polymorphic forms were somehow influenced by the presence of
the two liquid crystal phases.
Additionally, more than one subtype within the main polymorphic grouping has
been identified in some fats. For example, six different polymorphic forms have
been identified in cocoa butter, although there is still some debate whether they
are all truly unique polymorphs (Table 2). Two b0 and two b forms have been iden-
tified for cocoa butter. These polymorphs have slightly different melting points, but
they have X-ray spectra that fit within the definition of that polymorph.
Different nomenclatures have been used for denoting polymorphic forms, as
seen in Table 2 for cocoa butter. In the Greek nomenclature, where polymorphs
are given a Greek letter, the most stable form within a polymorph type is given
the subscript 1, and other polymorphs within that form are ordered in decreasing
stability or melting temperature. For example, cocoa butter has two b0 forms,
with the b0 1 form having the highest melting point (most stable). It is also common
to see a hyphenated number following the Greek letter, usually 2 or 3, stating
the chain-packing arrangement (double or triple chain packing, respectively). Wille
and Lutton (13) denoted the different polymorphs of cocoa butter with Roman
numerals, ordered in increasing melting point.
The time-temperature relationships governing the polymorphic behavior of
cocoa butter (in the temperature range of 20 C to 40 C and a time range of
10 days) were investigated by using real-time XRD (15). The g, a, and b0 poly-
morphs crystallized directly from the melt, and formation of b0 is much quicker
when it transforms from a compared with its formation from the melt. The least-
stable polymorph g stayed unchanged at solidification temperatures (Tp) below
10 C for 10 days. At higher Tp, the g polymporph transformed to a within a short
time. The g always transformed to a and a transformed to b0 . The a phase trans-
formed into b0 phase within 1 hour or less at temperatures above 6 C. They noted
that two b phases (polymorphs V and VI) were obtained via direct transformation
from the b0 phase only, not from the melt. Direct b phase formation from melt is
only viable if the melt has a memory effect. Their observation of two different b
phases from the b0 phase is contradictory to the results of the work of Schlichter-
Aronhime and Garti (16) who stated that b-V can be directly formed from the melt
and that b-VI can be formed only from the transformation of b-V.
2.3. Phase Behavior

In order to understand and control lipid crystallization, one should know the ther-
modynamic driving force for crystallization. In a pure system, like a single TAG,
the melting point, Tm, defines the driving force and a temperature below Tm is
required to induce crystallization. That is, the subcooling or the melting temp-
erature minus the actual temperature (Tm T) defines the driving force for crystal-
lization.
When two TAGs are mixed together, each species can influence the melting
properties of the other and a phase diagram is needed to define the crystallization
driving force at any condition. Rossell (17) summarized the phase behavior of bin-
ary mixtures of various TAGs. Depending on molecular differences (chain length
and degree of unsaturation), most binary TAG mixtures had either monotectic,
eutectic, or peritectic behavior (Figure 4), where the melting temperatures (liquidus
Figure 4. Phase behavior in binary systems: (a) monotectic, continuous solid solution;
(b) eutectic; (c) monotectic, partial solution; and (d) peritectic (18).
lines) of the species with the higher melting point decreased with increasing
addition of the species with the lower melting point. Wesdorp (19) used a thermo-
dynamic approach to predict phase behavior of each of the polymorphs for different
binary mixtures of TAGs. At the liquidus line on the phase diagram, the chemical
potential of the crystallizing species in the liquid state is equal to the chemical
potential of that species in the crystalline state (the definition of equilibrium).
If one of the species in a binary mixture is a liquid (oil or solvent), the other
species (higher melting point) will dissolve to some extent into the solvent (the
liquid oil can be considered a solvent in this case too). For example, a certain
amount of trisaturated TAGs (SSS) dissolves in solvent (either organic solvent
like acetone or hexane or a liquid oil), with the solubility concentration increasing
with temperature in the normal fashion (as shown schematically in Figure 5). In this
case, a binary mixture of SSS and solvent can be supersaturated with SSS once its
concentration exceeds the saturation concentration at any temperature, as indicated
by line AC in Figure 5. Thermodynamically, the driving force for crystallization is
the difference in chemical potential of SSS at point A and the chemical potential at
saturation (point C). Often, this crystallization driving force is approximated as
the difference in concentrations between points A and C.
When more than two TAG species are mixed together, the phase behavior is sig-
nificantly more complicated. For mixtures of three TAGs, a ternary phase diagram
(sometimes called a triangle diagram) can be used to denote phase behavior at any
temperature. The effects of temperature on phase behavior, however, must be taken
into account in yet another dimension, and thus, characterizing phase behavior in
ternary systems gets very difficult very quickly. The situation is even more complex
when there are greater than three TAG components, as occurs when a natural fat is
crystallized. Natural fats are mixtures of numerous TAGs, containing perhaps 10 to
12 different TAGs (as in cocoa butter) to well over 100 (as in milkfat). In natural
(g SSS/g solvent)
A
SSS Content
Temperature (C)
Figure 5. Schematic of a solubility diagram for a high-melting fat (SSS) in a liquid oil or solvent.
Line AC represents supersaturation for system at point A.
fats, the complex interactions among mixtures of various TAGs with different fatty
acids (chain length and degree of unsaturation) and having different melting points
result in melting over a range of temperatures. This range of temperatures may be
fairly narrow (as for cocoa butter) or may be broad (as for milkfat).
At a temperature above the melting point of the highest melting component, the
entire lipid is melted and the natural fat is in a liquid state. This highest melting
point, often characterized as the clear point (the temperature at which the last
crystal melts under carefully controlled heating conditions), is actually the melting
temperature of the TAG with highest melting point in the specific mixture of the
other TAG. Some researchers use this highest melting point, or some measure of
melting point like the Mettler dropping point, to define the driving force for crystal-
lization when the fat is cooled (20, 21). However, when the natural fat contains a
wide range of TAGs with different melting points, cocrystallization of different
TAGs into compound crystals is dependent on the temperature of crystallization.
Thus, the highest melting point does not necessarily represent the true driving force
for crystallization of the TAG species that are cocrystallizing.
If the fat is cooled to some point below the melting point of the highest melting
component and allowed to fully equilibrate (crystallize to the maximum extent in
the most stable polymorph), there will be some ratio of solid to liquid fat dependent
on the nature of the TAG mixture in the natural fat. This solid fat content (SFC) is
often measured by a pulsed nuclear magnetic resonance (NMR) technique. A plot of
the maximum amount of fat crystallized (SFC) at sequentially higher temperatures
100
90 100% cocoa butter

20% milk fat
80 50% milk fat
70 75% milk fat

Solid Fat Content (%)
100% milk fat

60
50
40
30
20
10
0
0 10 20 30 40 50
Temperature (C)
Figure 6. Solid fat curves for milkfat, cocoa butter, and their mixtures (4).
gives a melting profile that represents a type of phase equilibrium for a natural fat.
Some fats, like cocoa butter, have a very high SFC at low temperatures (about 90%
at 0 C) and then melt very sharply over a narrow temperature range (2535 C).
Other natural fats, like milkfat, have lower SFC at low temperatures (about 50%
at 0 C) and melt gradually with increased temperature. These SFC melting curves
are dependent on the specific molecular composition of the natural fat, as seen in
Figure 6 for cocoa butter and milkfat. Although SFC melting curves denote a cer-
tain aspect of phase behavior, they are not true phase diagrams because the compo-
sition of the crystalline phase changes as temperature increases. Nevertheless,
melting profiles are useful tools for understanding the crystallization behavior of
natural fats.
In mixtures of two or more natural fats, as often occurs in processed foods (e.g.,
milkfat and cocoa butter in chocolate), it is even more difficult to characterize the
true phase behavior for crystallization of fats. One approach that has been used to
characterize compatibility of fat mixtures is the isosolids diagram (22). SFC melt-
ing curves are obtained (by NMR) for various mixtures of the two fats, as seen for
cocoa butter and milkfat in Figure 6. Lines of constant SFC for different tempera-
ture and composition are calculated and plotted on an isosolids diagram (Figure 7).
40
% SOLID
30
10
Temperature (C)
20
20
60 30
40
70
50
10
80
0
100 % ICB 50:50 100 % AMF
0% AMF 0% ICB
Composition
Figure 7. Isosolids diagram for mixtures of anhydrous milkfat (AMF) and cocoa butter (ICB) (4).
Eutectic behavior is seen where the SFC of a mixture falls below the SFC for either
of the two individual components, as seen between 30% and 70% milkfat in
Figure 7. Isosolids diagrams allow phase compatibility to be studied (4), but they
do not provide a thermodynamic measure of driving force for crystallization.
Again, because the crystal phase composition may be different at different temp-
eratures (and mixture ratios), isosolids diagrams do not represent true phase
diagrams.
Recently, attempts have been made to characterize the driving force for crystal-
lization of natural fats by considering classes of TAG (high-melting, low-melting,
etc.). For example, milkfat contains three primary fractions that crystallize nearly
independently. The effective solubility of the high-melting fraction (HMF) in the
low-melting fraction (LMF) was found by using a turbidimetry technique (23).
Through chemical analysis of the major TAG constituents of HMF, an effective
solubility curve in terms of chemical composition of HMF in LMF was developed
and used to characterize the driving force for crystallization, as shown in Figure 8.
Such an effective solubility takes into account the intersolubility of different TAGs
as well as the melt behavior of individual TAGs. Although this approach is still
somewhat empirical, it provides a reasonable approximation of the crystallization
driving force in complex lipids. Further work is needed in this area to truly define
the driving force for crystallization in natural fats.
0.80
Critical
Concentration solubility
0.75 metastability
L+S
Long-chain TAG Content
(g(C46-C52)/g(<=C40))
Labile
0.70
AMF
0.65
L + S(conditional)
supersaturation at 30C
0.60
0.55
Liquid
Critical temperature for AMF
0.50
20 25 30 35 40 45
Temperature (C)
Figure 8. Operational phase diagram for high-melting components of milk fat dissolved
in low-melting components of milk fat based on triacylglycerol composition (acyl carbon
number) (4).
CRYSTALLIZATION BEHAVIOR 57
3. CRYSTALLIZATION BEHAVIOR
3.1. Nucleation
Nucleation, or the formation of a crystalline phase from the liquid state, is probably
the most important factor in controlling crystallization. The nucleation rate is the
major determining factor in the number and size of crystals formed, their poly-
morphic form, and the ultimate distribution of crystalline solids. Crystallization
cannot occur until the phase is supersaturated or subcooled. However, attaining
the supersaturated or subcooled state is not necessarily sufficient to promote crystal-
lization because a certain energy barrier exists to formation of nuclei.
A nucleus is the smallest crystal that can exist in a solution at a certain tempera-
ture. The formation of a nucleus from the liquid phase, or the nucleation process,
requires the molecules to organize into a crystal lattice. There is a free-energy bar-
rier opposing this transition, but when nucleation does occur, there is a release of
energy (latent heat of fusion) as the molecules assume the lower energy state in the
crystal lattice. Based on these energy considerations, a free-energy maximum exists
that must be overcome for nucleation to occur (24). At this maximum free energy,
there is a critical size for a stable nucleus. Above this critical size, a stable nucleus
is formed that continues to grow, whereas clusters smaller than the critical size can
potentially disperse into the liquid state (4, 24, 25).
3.1.1. Nucleation Theories Nucleation is generally classified according to pri-

mary nucleation, which may occur either homogeneously or heterogeneously, and
secondary nucleation mechanisms. The presence of foreign nucleating sites
catalyzes the formation of heterogeneous nucleation, whereas homogeneous
nucleation occurs without the assistance of outside surfaces. Secondary nucleation
occurs when crystals in a subcooled system spawn new nuclei, generally because of
contacts between two crystals, or between a crystal and a surface such as a stirrer or
a solid wall (4).
3.1.1.1. Homogeneous Nucleation Homogeneous nucleation is based on accre-

tion of molecules in the liquid phase. Single species (molecules or ions) come
together and form dimers. Dimers become trimers by addition of a molecule,
and this accumulation process continues until eventually a stable nucleus forms
depending on temperature and supersaturation.
According to the classic nucleation theory, a free-energy barrier must be over-
come to form a stable nucleus. The energy needed to form a crystal is proportional
to the interfacial tension, g, and the surface area. However, once a nucleus is
formed, there is a release of energy (latent heat) associated with the phase change.
The free-energy change for the formation of the crystal surface is positive and
proportional to the surface area (r2) and interfacial tension (g) between the crystal
and the surrounding fluid. The free-energy change for formation of the bulk of the
crystal is negative because energy is released because of latent heat of fusion and
proportional to volume (r3). The total free-energy change during nucleation is the
sum of these free-energy terms for the formation of the crystal surface and the crys-
tal volume. Thus, a maximum in free energy occurs during nucleation at some
critical nucleus size, rc. The critical nucleus size is the minimum size for a stable
nucleus. Above this critical size, a stable nucleus is formed, whereas clusters of
molecules smaller than this critical size can potentially redisperse into the liquid
phase (4, 2426).
Homogeneous nucleation, however, rarely occurs under commercially important
conditions. In practice, nucleation is usually dominated by a heterogeneous mechan-
ism, where a foreign surface serves to reduce the energy barrier to nucleation.
3.1.1.2. Heterogeneous Nucleation Typically, nucleation of fats (as well as most

other substances) occurs by a heterogeneous process catalyzed by foreign nucleat-
ing sites. The presence of these foreign nucleating sites, like dust particles, vessel
walls, and other foreign particles in the system, reduces the free energy required for
nucleation. Even though the exact mechanisms of heterogeneous nucleation are not
clearly understood, it most likely results from the interactions at the interface
between the solid particle and the supersaturated fluid. These interactions result
in a local ordering of molecules of the crystallizing species; thus, the free energy
of formation of a critical size for a stable nucleus is decreased. For example, nuclea-
tion on a surface irregularity at a wall results in a decrease in the surface energy
required to form a stable nucleus. In general, the capability of a foreign surface
to catalyze nucleation is thought to depend on the degree of lattice matching
between the solid surface and the crystals of the nucleating species (26), although
this trend is not always observed (24). In general, a closer lattice match indicates a
greater likelihood that a surface will catalyze heterogeneous nucleation. Because
the foreign surface provides some of the energy needed to overcome the formation
of the crystal surface, heterogeneous nucleation occurs at lower crystallization driv-
ing force (supersaturation or subcooling) than homogeneous nucleation (24, 27).
Interestingly, there is an aging effect on the ability of a heterogeneous nucleation
site to catalyze nuclei formation (26). That is, the same material nucleated multiple
times under identical conditions results in a spread of nucleation capabilities. This
variability in heterogeneous nucleation leads to difficulties in controlling lipid
crystallization.
3.1.1.3. Secondary Nucleation The formation of new nuclei in the presence of

existing crystals is called secondary nucleation. Secondary nucleation may occur
whenever microscopic crystalline elements are separated from an existing crystal
surface (24), although contact secondary nucleation is probably the main mechan-
ism in commercial fat crystallization processes. As a crystal slurry is agitated in a
vessel, crystals contacting with other crystals, vessel walls, or stirrer may lead to
attrition or fracture of the existing crystal structure, and consequently, secondary
nuclei are formed (Figure 9).
Contact secondary nucleation has been explained by two possible mechanisms,
namely, the adsorption layer theory and microattrition (4). The adsorption layer the-
ory involves displacement of a surface layer of organized molecules (precrystalline)
Crystal - Wall
Crystal - Crystal
Crystal - Stirrer
Figure 9. Potential sources of contact nuclei in a stirred crystallizer (4).
as a result of crystal interactions or collisions. Thus, precrystalline embryos are dis-

persed into the crystallizing medium, where under conditions of secondary nuclea-
tion they survive and develop into stable nuclei (24). Microattrition theory involves
the dispersion of broken pieces of a crystal into the fluid, which remain as stable
nuclei (28). Production of secondary nuclei may also result from growing crystals
containing dislocations, inclusions, or defects (24, 28). As expected, secondary
nucleation is also dependent on the crystallization driving force (supersaturation
or subcooling), with more stable nuclei being formed at higher supersaturation
(25, 27).
Secondary nucleation may also occur in static conditions under certain circum-
stances (11). In lipid systems, needle-like or dendritic crystals that form under cer-
tain conditions may lead to the formation of secondary nuclei. Heat dissipation and/
or concentration of noncrystallizing species in certain regions may lead to melting/
dissolution at the base of the branches of dendritic crystals and result in the forma-
tion of numerous nuclei centers. Although the exact mechanisms for this type of
secondary nucleation are not fully understood, it is undoubtedly important for
nucleation in emulsions and in certain cases during seeding of bulk solutions (as in
tempering of chocolates).
During fractionation of fats, secondary nucleation is undesired because the small
crystals, formed in the presence of larger ones means that subsequent separation is
not efficient. Thus, stirring or agitation during fractionation is usually kept to the
minimum needed to facilitate heat transfer.
Secondary nucleation is influenced by numerous parameters, including the driv-
ing force for crystallization, temperature, additives, impurities, agitator, agitation
rate, the number and size of existing crystals, and roughness of the crystallizer sur-
face. The parameters affecting nucleation and nucleation rate will be reviewed in a
subsequent section.
3.1.2. Nucleation Kinetics Nucleation rate is generally measured as the rate of

formation of nuclei (numbers formed per unit volume per unit time). Sometimes the
induction time, or the time necessary for the onset of nucleation once the subcooled
state has been attained, is used for calculation of nucleation rate because the actual
rate is often very difficult to measure. Induction time for nucleation will be
reviewed later in this section.
In some cases, as in crystallization of viscous materials from the melt, the
FisherTurnbull equation (29) is often used to describe nucleation of lipids (20, 30)
( )
NkT Gd 16pg3 Tf2
J exp exp : 1
h kT 3kTHf2 Tf T2
Here, N is the number of molecules (monomers) per mole, k is the Boltzman con-
stant, T is absolute temperature, h is Plancks constant, Gd is a term denoting the
mobility of the lipid molecules, g is interfacial tension, Tf is melting temperature,
and Hf is latent heat of fusion. The first exponential term in Equation 1 has been
related to the ability of a lipid molecule to attain the necessary conformation to
become attached to the crystal lattice, and it is often given as (20)
Gd aS
; 2
kT R
where a is the fraction of molecules with the correct configuration to be incorpo-

rated into the crystal lattice, S is the decrease of entropy associated with incorpora-
tion of one mole of lipid, and R is the ideal gas constant. Kloek (31) determined that
80% of TAG molecules were in the correct conformation for incorporation into a
nucleus.
According to the classic theory, nucleation is a very strong function of crystal-
lization driving force. At low driving forces (low supersaturation or high tempera-
tures), nucleation rate is essentially zero. After some critical driving force is
attained, nucleation becomes spontaneous and occurs almost instantaneously
once the critical driving force has been attained. In natural fats, cooling below a
certain temperature results in massive nucleation with numerous nuclei being

formed. For fats, the nucleation rate also depends on the type of polymorph formed,
because each of the polymorphs has a different melting point and interfacial
tension.
The a polymorph, the least stable of the common polymorphic forms of fats, has
the lowest interfacial tension, heat of crystallization, and melting point temperature.
The b0 and b polymorphs have increasing interfacial tensions, heats of crystallization,
and melting point temperatures. Thus, as a liquid fat is cooled, the polymorph that
forms first depends on the properties of the different polymorphs. For example,
Hernqvist (2) showed that the first polymorph to appear as trisaturated triacylgly-
cerols (with fatty acids from lauric to stearic acid) were cooled was either the a or
b0 polymorph, depending on the chain length, even though the nucleation tempera-
ture was well below the melting point of the b polymorph (Figure 10). In the case of
tristearin, formation of the a polymorph occurred even though both the b0 and b
polymorphs were subcooled to a greater extent (higher driving force).
The formation of a less-stable polymorph under conditions where a more stable
polymorph is subcooled to a great extent has been explained by the difference in
interfacial tensions of the different polymorphs (28). A small difference in interfa-
cial tension can result in a large difference in nucleation rate (25), and this effect
80
mp
mp
60
mp
Temperature (C)
liq -
liq - 1
40
liq - 1 liq - 1
20 1 -
0
12 14 16 18 nc
Figure 10. Onset temperature of nucleation and polymorphic form of monoacid triacylglycerols
with different chain lengths (nc) at slow cooling rate (0.4 C/min). amp, b0 mp and bmp represent
the melting temperatures of the different polymorphs (2).

Nucleation Rate
Driving Force
Figure 11. Nucleation rate (highly schematic) of lipid polymorphs (4).
generally is greater than the effect of temperature driving force. Thus, nucleation
rate of lipid polymorphs is often considered to follow the general trend shown in
Figure 11.
Kellens et al. (32) studied the nucleation rate of the b0 polymorph of tripalmitin
(PPP) by using a microscope counting technique. An increase in temperature from
45 C to 52 C led to a decrease in nucleation rate, as expected. A semilogarithmic
plot of nucleation rate versus the inverse of the square of the subcooling, according
to the general form of Equation 1, gave a straight line over the range from 45 C to
50 C. Above 50 C, a different straight line was obtained indicative of formation of
a different polymorph (confirmed from the change in crystal habit observed micro-
scopically).
Another important kinetic aspect of nucleation is the induction time, defined as
the time required for a system to nucleate once a certain subcooling has been
attained. That is, induction time for the onset of nucleation is the time required
for detection of the first nuclei in a supersaturated or subcooled system. In reality,
induction time includes the true time required for nucleation plus the time required
for detection of crystallization by the experimental technique. Techniques that have
been used for studying lipid nucleation include microscopy, refractive index, light
scattering, calorimetry, viscosity, turbidimetry, laser polarized-light turbidimetry,
and NMR (4). Each method has its advantages and limitations for studying lipid
nucleation (33). Herrera et al. (34) showed that light microscopy could detect a
crystal with a minimum size of 0.2 mm, whereas laser polarized-light turbidimetry
detected a smaller size of nuclei. Thus, the laser polarized-light turbidimetry
technique was more accurate and suitable when size of nuclei is very small. Any
method of studying induction time for nucleation must be used with caution (35).
120
melt
100
80
melt melt
Induction Time (s)
60
40

20
0
35 40 45 50 55 60 65
Temperature (C)
Figure 12. Induction time kinetics for onset of nucleation of different polymorphis forms of
tripalmitin. Melting temperatures of each polymorph indicated by straight line (4).
The induction time, t, is a function of subcooling and reflects the time necessary
for a critical size of nucleus to be developed in the liquid. The induction time is also
dependent on the size at which nuclei are detectable and the growth rate at this early
stage. Despite this limitation in measurement methods, induction times are often
considered to be inversely proportional to nucleation rate (4)
ta J 1 : 3
Induction times for nucleation of a tripalmitin melt at different temperatures are

shown in Figure 12 (36). The tripalmitin melt was cooled quickly from 80 C to
the different crystallization temperatures indicated on the figure and induction
time measured as the first point of detection of crystals on a polarized light micro-
scope. The relative time scales for the onset of nucleation are clearly shown, with
the less-stable a form taking significantly less time to nucleate than the b0 poly-
morph. The induction time for the most stable b polymorph was substantially
longer than for either of the less-stable polymorphs.
3.1.3. Nucleation in Lipid Emulsions In many foods, the lipid phase appears
in emulsion form, or small droplets of fat dispersed in a continuous aqueous phase,
as for example found in cream (37). The nature of the fat crystals in cream plays an
important role in determining the physical properties and quality characteristics
of butter. Thus, nucleation of fats in emulsion form is an important commercial
phenomenon.
When a fat is emulsified, nucleation is substantially altered compared with the
same fat in bulk liquid form. This is primarily because of the distribution of hetero-
geneous nucleation sites among the emulsion droplets. If there are more droplets
than heterogeneous nucleation sites, then some of the droplets will nucleate by a
homogeneous nucleation mechanism. That is, as a finely dispersed emulsified sys-
tem is cooled, one population of droplets nucleates at relatively higher temperatures
because of heterogeneous nucleation, whereas another population nucleates at
substantially lower temperature because of homogeneous nucleation.
It is widely recognized that the size of the emulsion droplets is an important
factor in the extent of subcooling (11). Smaller droplet size leads to nucleation at
a lower temperature (greater degree of subcooling). Thus, the probability of nuclea-
tion within an emulsion droplet is lower than in the bulk fat (38). The dispersity of
droplet sizes, however, did not change the critical subcooling required for onset of
nucleation (39).
Crystallization from the emulsified state may lead to different nucleation pro-
cesses than observed for the same fat in bulk liquid form. It has been suggested
that nucleation often occurs at the interface of the droplet where surface-active
agents are located. The general similarity of the lipophilic components of surfac-
tants oriented at the surface may provide some ordering and structure for the lipid
molecules within the droplet and enhance nucleation, as found for example by
Kaneko et al. (40) for a hydrocarbon emulsion. Walstra (11) also suggests that for-
mation of compound crystals from emulsions of natural fats may be different than
the same fat crystallized from bulk liquid. The initial polymorph formed may also
be different, with more stable polymorphs more likely to form in the emulsion (38).
3.2. Crystal Growth

Once nuclei have formed, they grow by the incorporation of other TAG molecules
from the liquid phase. The incorporation of a new TAG molecule into an existing
crystal lattice depends on the probability of it having the correct configuration at the
correct site on a crystal surface. When a molecule diffusing from the liquid phase
reaches the crystal surface, it may bind into the crystal lattice or return to the super-
saturated system, depending on its configuration. Growth continues as long as there
is a driving force for crystallization. Eventually crystal growth ceases when the
system attains phase equilibrium or the entire system is crystallized (4).
For growth to occur, molecules from the liquid phase must migrate to the surface
of the crystal, where rearrangement and orientation takes place. A growth unit
(either an individual molecule or a cluster of molecules) then migrates across the
crystal surface until it finds an appropriate site for incorporation into the lattice.
Once a growth unit has become incorporated, there is a release of latent heat and
this energy must be diffused away from the growing surface or else the temperature
will increase to the point where no further growth can occur. General theories of
crystal growth have been developed for crystallization of pure substances (4, 24).
These theories are based on one or more of the steps in crystal growth being the
rate-limiting step. Further details of these theories can be found in the references
by Mullin (24) and Hartel (4).
In natural fats, the different TAG species come together to form mixed or com-
pound crystals. The likelihood of two TAG crystallizing together depends on the
similarities or differences in molecular configuration (chain length, degree of unsa-
turation, nature of any double bonds, and arrangement of the fatty acids on the gly-
cerol backbone). TAG species that are similar tend to cocrystallize, but under
certain conditions (e.g., very rapid growth), even different TAG species can cocrys-
tallize in a loosely organized crystal lattice (g or a polymorphs). In fact, it is this
molecular diversity that results in some natural fats remaining in the metastable b0
polymorph for extended periods of time.
Growth of TAG crystals is typically very slow (41). There may be several rea-
sons for slow growth rate of TAG crystals:
The incorporation of a TAG molecule into a crystal lattice requires a very

large loss in conformational entropy, and thus, a long time is needed for the
TAG molecule to fit into the crystalline lattice. In addition, the TAG molecule
may be detached before the crystalline lattice before it is fully incorporated
into the crystalline lattice. For example, for growth of tristearin (SSS) in
triolein (OOO), linear growth rates of the order of 108 to 107 m/s have been
observed (41).
In a multicomponent fat, there is a vigorous competition between similar
molecules for a vacant site in a crystal lattice. Multicomponent fats crystallize
more slowly than pure TAG at the same crystallization driving force.
However, crystal growth in multicomponent fats may be enhanced by the
formation of compound crystals. Compound crystals usually occur in the a or
b0 forms and rarely in the b form.
According to Timms (25), more stable polymorphs grow faster than unstable
ones at any given temperature. This is because of the higher melting point of the
more stable polymorphs, which means that the more stable polymorph has a higher
degree of subcooling at any given temperature.
3.3. Modeling of Crystallization Kinetics of Fats

Crystallization data have typically been treated theoretically using either the
FisherTurnbull model or the Avrami equation. These analyses not only allow lipid
crystallization to be modeled but may also shed some light on the mechanisms of
nucleation and growth. However, there is some recent debate about the validity of
such models, especially the application of the Avrami equation (42) to accurately
depict crystallization of lipids.
Recently, Foubert et al. (43) developed a new, empirical model (Foubert model)
to predict the kinetics of fat crystallization. Other authors have used a reparameter-
ized Gompertz equation (Gompertz model) to empirically describe crystallization
kinetics of fats (44, 45).
3.3.1. Avrami Analysis The Avrami equation, a general approach for descrip-
tion of isothermal phase transformation kinetics originally developed for polymers
(46), is often used for describing nucleation and crystal growth in fats. The Avrami
equation is given as
1 X expfktn g; 4
where X is fraction of crystal transformed at time t during crystallization, k is crys-

tallization rate constant that depends primarily on crystallization temperature, and
n, the Avrami exponent, is a constant relating to the dimensionality of the transfor-
mation. The values of n and k are calculated from the linear form of the Avrami
equation (Equation 5) as the slope and intercept at ln t 0, respectively
ln ln1 X lnk nlnt: 5
The Avrami exponent (n) is a function of the number of dimensions in which

growth takes place, and it reflects the details of nucleation and growth mechanisms.
For most transformations, the n is found to be constant over a substantial tempera-
ture range (47). Christian (48) tabulated some values of n expected for various crys-
tallization mechanisms. For example, an n of 4 indicates heterogeneous nucleation
and spherulitic growth from sporadic nuclei, whereas an n of 2 indicates high
nucleation rate and plate-like growth (i.e., two-dimensional growth).
Metin and Hartel (49) applied the Avrami equation to the isothermal crystalliza-
tion of binary mixtures of cocoa butter with milk fat or milk fat fractions at 15 C.
Avrami analysis indicated an n value of 4 for cocoa butter crystallization, so the
suggested mechanism was heterogeneous nucleation with spherulitic growth from
sporadic nuclei. For milk fat, the value of n was 3, suggesting that the crystalliza-
tion mechanism was instantaneous heterogeneous nucleation with spherulitic
growth. For milk fat fractions, the n value was 2, which suggested that the mechan-
ism was high nucleation rate at the beginning of crystallization decreasing with
time, and plate-like growth.
The crystallization rate constant (k) is a combination of nucleation and growth
rate constants, and is a strong function of temperature (47). The numerical value of
k is directly related to the half time of crystallization, t1/2, and therefore, the overall
rate of crystallization (50). For example, Herrera et al. (21) analyzed crystallization
of milkfat, pure TAG fraction of milkfat, and blends of high- and low-melting milk-
fat fractions at temperatures from 10 C to 30 C using the Avrami equation. The n
values were found to fall between 2.8 and 3. 0 regardless of the temperature and
type of fat used. For temperatures above 25 C, a finite induction time for crystal-
lization was observed, whereas for temperatures below 25 C, no induction time was
found (crystallization was instantaneous)). Calculation of crystallization rate con-

stant, k, and half time for crystallization based on the Avrami analysis were in line
with the two different behaviors observed in SFC values of the fats.
Even though the Avrami model has been the most frequently used model to
describe the isothermal kinetics of fat crystallization, there are some concerns about
the use of the model in fat crystallization. Theoretically, integer values should
be obtained for the Avrami exponent, n. However, generally fractional values of
n were obtained in crystallization of fats and oils. Additionally, the linear format
of the Avrami equation should give a single slope associated with the value of
the Avrami exponent. However, in some studies, two regions of different slopes
were obtained. Moreover, secondary nucleation during crystal growth is not consid-
ered in the Avrami model, which may in part explain the noninteger values of the
Avrami exponent.
3.3.2. Fisher-Turnbull Analysis The activation free energy for nucleation, Gc,
may be found from the FisherTurnbull equation given in Equation 1. The term in
the second exponential of Equation 1 is often given as Gc =kT. Combination of
Equations (1) and (3) allows development of the following equation:
( )
h aS 16pg3 Tf2
tT exp exp : 6
Nk k 3kTHf2 Tf T2
Based on Equation 11, a plot of tT versus {1/T(T)2} leads to a straight line for
nucleation of a given polymorph. The critical free energy for nucleation, Gc , is
then found from the slope of that straight line, s, as
sk
Gc : 7
Tf T2
For a given fat system, although the slope is constant, Gc varies with crystallization
temperature.
The FisherTurnbull approach has been used to compare nucleation of various
lipid systems. Ng (51) and Herrera et al. (34), for example, have used this approach
to characterize crystallization of palm oil and hydrogenated sunflower oil, respec-
tively. The use of the FisherTurnbull approach to characterize nucleation leads to a
better understanding of the energy changes needed for onset of nucleation and can
be used to compare nucleation in different systems. However, this approach is based
on a crystallization driving force defined by a single melting point, which may only
occur in cases where a single TAG component (or a TAG grouping with narrow
range of melting temperature) crystallizes from a liquid oil. It also applies only
when the subcooling is low (typically less than 10 C). In cases where massive
cocrystallization and compound crystal formation occurs, this approach does not
work.
3.4. Crystalline Microstructure

The dispersion of the crystalline fat phase in a material determines the physical
and textural properties of a lipid-based product. For example, the hardness, snap,
and glossy appearance of chocolate is caused by crystallization of cocoa butter
in the form of numerous, very small (1 mm or less) crystals of the most stable poly-
morph (b form). The size distribution (mean size and range of sizes), polymorphic
form, and shape of the fat crystals, as well as the network formed among the
crystals, all play important roles in determining physical attributes of lipid-based
products.
In the case of lipid fractionation, however, a different crystal size distribution is
desired. As the fat crystals are to be separated from the liquid phase, uniform crys-
tals of distinct size and shape are needed for the most efficient separation. For the
most efficient separation by filtration, reasonably large (200 to 300 mm) crystals of
fairly uniform size (narrow distribution of sizes) are needed. Fractionation techno-
logies carefully control nucleation and growth to produce this uniform distribution
of crystals to enhance filtration and separation of the high-melting stearin phase
from the low-melting olein phase.
In crystallization of most natural fats, the first crystals formed are often observed
as thin and fairly long platelets (41). For example, cooling of melted milkfat leads
to initial formation of small b0 crystals in needle or platelet shape. As these initial
crystals grow, they aggregate into spherulites (52) consisting of the needles
arranged radially and ranging in size from a few microns up to about 300 mm. If
crystallization is very slow (slow cooling), very large spherulitic crystals form. In
contrast, rapid cooling to a low temperature results in the formation of numerous
small crystals, often found in a random orientation (53). Thus, cooling rate is one of
the most important factors influencing crystalline microstructure. Further details on
lipid crystalline microstructure are given in Chapter 4.
4. CONTROLLING CRYSTALLIZATION
4.1. General Principles of Controlling Crystallization

To truly control crystallization to give the desired crystalline microstructure
requires an advanced knowledge of both the equilibrium phase behavior and the
kinetics of nucleation and growth. The phase behavior of the particular mixture
of TAG in a lipid system controls both the driving force for crystallization and
the ultimate phase volume (solid fat content) of the solidified fat. The crystalliza-
tion kinetics determines the number, size, polymorph, and shape of crystals that are
formed as well as the network interactions among the various crystalline elements.
There are numerous factors that influence both the phase behavior and the crystal-
lization kinetics, and the effects of these parameters must be understood to control
lipid crystallization.
CONTROLLING CRYSTALLIZATION 69
4.2. Parameters Affecting Crystallization

Parameters that affect crystallization may influence either the thermodynamic beha-
vior or the crystallization kinetics (or both). Parameters that influence lipid crystal-
lization include chemical composition, subcooling, cooling rate, agitation, minor
components of fats (mono- and diacylglycerols, polar lipids, etc.), and scale of
operation. The effects of these parameters on lipid crystallization will be reviewed
briefly in this section. More detailed information about the effects of these
parameters on lipid nucleation and crystal growth may be found elsewhere (4, 24,
28, 54).
4.2.1. Compositional Parameters
4.2.1.1. TAG Composition Natural fats are composed of a wide range of TAG
that contain fatty acids of differing chain length, degree of unsaturation, and posi-
tional arrangement on the glycerol backbone. The fatty acid composition of fats
may be broad, as in milkfat, or may be limited, as in cocoa butter. It might be
expected that a faster nucleation rate occurs in molecularly similar fats compared
with the ones with complex structure (wide range of fatty acid species), but this is
not necessarily true. Metin and Hartel (55) observed that the induction times for
nucleation of milkfat were significantly faster than that for cocoa butter at the
same isothermal temperatures (and approximately the same melting point). The
faster induction time for milkfat may be a result of a higher driving force (even
though the difference between crystallization temperature and final melting point
is about the same), or it may be because the TAGs in milkfat more readily come
together into mixed crystals. As both are likely to form in a mixture of a and b0
polymorphs, the differences in nucleation rate cannot be attributed to the formation
of different polymorphs.
Furthermore, when two fats added together are crystallized from the liquid state,
the nucleation rate of the mixture often decreases. For example, the addition of
milkfat or milkfat fractions to cocoa butter is widely known to retard crystallization
of cocoa butter, with higher addition levels having a greater effect. This effect is
commercially important because milk chocolate must be processed at lower tem-
peratures to generate the same level of crystallization as dark chocolate. Metin
and Hartel (55) documented the inhibitory effects of milkfat and milkfat fractions
on induction time for nucleation of cocoa butter. Martini et al. (56) measured the
induction time for nucleation for addition of sunflower oil to a high-melting milkfat
fraction. As the level of sunflower oil increased to 40%, the melting point decreased
only by a few degrees, but induction time increased by more than a factor of two.
This suggests that the effect of sunflower oil on inhibiting nucleation of the milkfat
was primarily caused by a true inhibition rather than to a decrease in the driving
force for crystallization.
4.2.1.2. Minor Constituents Minor constituents in fats that can influence crystal-
lization of TAG include the more polar lipids like DAG, MAG, free fatty acids,
phospholipids, and sterols, although there may be trace amounts of other compo-
nents that can influence crystallization as well. These constituents have long
been considered as active agents for affecting crystallization. In some cases, the
presence of these components may enhance crystallization, whereas in other sys-
tems, an inhibition is observed.
Nucleation of fats may either be enhanced or inhibited by the presence of these
minor components. Dimick (57) has argued that the phospholipids in cocoa butter,
with higher melting point than the cocoa butter TAG, crystallize first and subse-
quently catalyze formation of cocoa butter TAG. The appearance and chemical
composition of cocoa butter crystals formed from refined cocoa butter (phospholi-
pids removed) was different from that of the initial crystals formed in nonrefined
cocoa butter. Recent studies where these minor components have been separated
and then added back to the purified TAG have shown that they invariably inhibit
nucleation (21).
There are three potential mechanisms by which addition of minor lipids might
affect crystallization. They may limit mass transfer rates of crystallizing TAG to the
appropriate site for incorporation into the lattice, they may adsorb on the surface of
the growing crystal or cluster and inhibit further incorporation of the crystallizing
TAG, or they may actually be incorporated into the crystal lattice as a crystal forms
and grows (4). Through any of these mechanisms, the minor constituents in a fat
may affect the polymorphic form that is crystallized and often affects the crystal
microstructure through preferential inhibition on certain crystal faces (28).
However, in some cases, increased crystallization rate may be observed in the
presence of minor constituents. If a macrocrystallizing substance and an additive
have a similar structure or form similarly structured complexes to the lattice of
the crystallizing substance, then new growth sites on the crystal lattice can be
formed by the adsorbed addition. These active sites may be energetically more
favorable for incorporating further substances, resulting in an increased crystalliza-
tion rate (58). For example, Smith et al. (59) found that addition of monolaurin
and lauric acid enhanced the crystal growth rate of trilaurin by decreasing facet
and crystal size. However, addition of dilaurin decreased the crystal growth rate and
altered crystal morphology. They postulated that the varying effects were observed
because of the varying sizes and shapes of the additives.
4.2.1.3. Seeding At times, crystallization of natural fats may be promoted by the

addition of a solid seed material, either of the desired crystallizing species or a for-
eign particle with nucleating properties. If seeds of the desired crystallizing species
are added, they can promote further nucleation and/or provide a surface area for
additional crystal growth. Smith (60) reported that addition of b0 or b seed crystals
to cooled palm oil initiated crystallization at lower degrees of subcooling (higher
temperatures) than in the absence of these seeds.
In a sense, tempering of chocolate is done to create a small (<3%) population of
seed crystals in the melted chocolate, which catalyze further crystallization of the
cocoa butter when the chocolate is subsequently cooled. Through the tempering
process, seed crystals in the b polymorph are formed. These stable crystals then
CONTROLLING CRYSTALLIZATION 71
promote formation of numerous small cocoa butter crystals, also in the stable b
polymorphic form, as the chocolate is cooled. In this case, the existing seed crystals
are thought to spawn additional nuclei through secondary nucleation, although the
exact mechanism for this process is not clearly understood. A similar effect is
observed upon addition of the high-melting TAG, behenic-oleic-behenic (BOB),
to chocolate (61). In this case, the BOB molecules, with very high melting point
(53 C), catalyze formation of the b polymorph of cocoa butter crystals, eliminating
the need for tempering of chocolate.
4.2.2 Operating Parameters
4.2.2.1. Subcooling or Crystallization Temperature Arguably, the most impor-

tant parameter that influences lipid crystallization is subcooling, or the temperature
to which the lipid is cooled below the equilibrium point. As subcooling increases,
nucleation rate increases and induction time for crystallization decreases. In many
natural fats in bulk liquid form (as opposed to emulsified form), only a few degrees
of subcooling are necessary to induce crystallization because of the presence of
nucleation sites. These sites catalyze nucleation by lowering the energy required
for the formation of nuclei.
If subcooling is small, molecules only with the correct configuration (spatial
orientation, fatty acid composition, positional arrangement of fatty acids, etc.)
are incorporated into a crystal because molecules have sufficient time to orient
themselves perfectly. At low subcoolings (crystallization at temperatures within a
few degrees of the melting temperature), crystallization rate is slow and only the
more stable polymorphs form. When the subcooling is large, incorporation of mole-
cules to the crystal surface is faster, resulting in imperfect attachment of TAG
molecules to the surface. Different TAGs can cocrystallize if their chain length
and melting points are reasonably close to each other. Consequently, TAGs of dif-
ferent configuration are more easily incorporated into the crystal. The result is more
rapid crystallization, but at the cost of formation of compound crystals and lower
stability polymorphs.
4.2.2.2. Cooling Rate Fat crystallization is greatly influenced by the cooling rate
(62). Rapid cooling generally leads to nucleation occurring at a lower temperature
than for slow cooling. That is, during slow cooling, the temperature is higher for a
longer time and the TAGs have more opportunity to rearrange into a crystal lattice.
Cooling rate also affects nucleation rate, which governs crystal size. Rapid cooling
to a low temperature promotes a higher nucleation rate, which leads to formation of
numerous small crystals (62). When a fat is cooled very slowly, large crystals form.
Cooling rate also influences crystalline microstructure. Marangoni and Hartel (53)
used confocal microscopy to show that slowly cooled milkfat formed spherulitic
crystals, whereas rapidly cooled milkfat formed random crystalline strands.
4.2.2.3. Agitation (Shear) The speed of mixing is generally thought to promote

both nucleation and crystal growth (4). However, the effects of agitation rate may
be complex because it is sometimes difficult to separate the effects of mixing and

cooling rate on crystallization (higher agitation often results in faster cooling rate).
Thus, higher agitation rate may influence crystallization time and crystal size with-
out necessarily influencing nucleation and growth (41).
Agitation may promote nucleation because of the mechanical disturbance that
supplies energy to overcome the energy barrier for nucleation (24). Agitation
aids cooling, crystallization, and formation of small crystals. Slow cooling rate
and slow agitation of fats may result in increased number of mixed crystals;
thus, melting range is increased. Higher agitation rate results in a higher crystalli-
zation rate and formation of small crystals. Agitation also promotes secondary
nucleation, primarily by detachment of small particles from crystal structures.
Thus, Herrera and Hartel (62) found that higher agitation rates led to the formation
of smaller fat crystals in a milkfat model system.
The structure of the crystal network in fats and oils is strongly influenced by
cooling and shear rates, the degree of subcooling, and annealing time. For example,
crystalline orientation and acceleration of phase transitions induced by shear in dif-
ferent fats (cocoa butter, milkfat, stripped milkfat, and palm oil) were demonstrated
using synchrotron XRD (63). The fats were crystallized under static conditions and
under shear (90 s1 and 1400 s1) from the melt (50 C) to 18 C at a rate of 3 C/min.
During static crystallization (20 C after 1 day), the initial nucleation was character-
ized by the appearance of platelet-like nuclei far apart from each other. As they
grew, the system became a dispersed suspension of rapidly growing crystals. Even-
tually, clusters of crystals were formed. The introduction of a moderate shear field
to the fat system seemed to prevent the formation of these clusters. The presence of
shear field resulted in the formation of small asymmetric crystals. Weak or no orien-
tation of the crystals was observed at low shear rates either because of a random
distribution of anisotropic crystals or the formation of spherical particles upon
aggregation. They also stated that the shear forces accelerated solid-state phase
transformations.
The effects of agitation rate on crystallization kinetics of butter fat were studied
by Grall and Hartel (64). In a 2 L batch crystallizer, increased agitation rate caused
an increase in nucleation rate (more crystals generated per unit time) and an
increase in total crystallization (mass deposition) rate. However, the effects of agi-
tation on growth rates of individual crystals were dependent on temperature of
operation. At 30 C, increased agitation led to a decrease in growth rate, whereas
for crystallization at either 15 C or 20 C, increased agitation caused an increase
in growth rate. These results may be related to the different composition effects
at the different temperatures (different TAGs cocrystallize).
Garbolino et al. (65) studied the effects of shear rate on crystallization of a con-
fectionery coating fat (hydrogenated and fractionated mixture of soybean and cot-
tonseed oils) using ultrasonic sensors. They hypothesized that primary nucleation is
less likely to be affected by shear and suggested that crystal nuclei probably form
from heterogeneous nucleation sites (dust particles or other suspended insoluble
materials and imperfections in the container walls). They also suggested that
REFERENCES 73
growth of crystals and their interactions are more likely to be affected by stirring
because of the occurrence of frequent interparticle collisions.
Thus, from the contradictory results available in the literature, it is clear that our
understanding of the effects of heat and mass transfer on crystallization processes is
still not complete.
4.2.2.4. Scale of Operation The size of the batch being crystallized may influ-
ence rate of crystallization. For example, crystallization from an emulsion generally
occurs at a lower temperature than for the bulk fat based on the separation of cat-
alyzing nucleation sites. In an emulsion, the catalyzing nucleation sites are more
dispersed (spread through the number of droplets) and this leads to nucleation at
a lower temperature than the same fat in bulk phase.
Grall and Hartel (64) studied crystallization of milkfat at different scales of
operation (2 L and 20 L) and found induction times for nucleation were lower
but individual crystal growth rates were higher in the larger scale crystallizer. Other
crystallization parameters (total crystal number, mean size, yield, and nucleation
rate) were not significantly influenced by this difference in crystallizer size. As
scale of operation changes, mixing rates and heat transfer rates change as well,
which can influence crystallization processes. Scale up of fat crystallization
processes is still somewhat of a trial and error process because of the lack of
fundamental understanding of the effects of heat and mass transfer on lipid
crystallization.
5. SUMMARY
Controlling lipid crystallization in foods has proven to be a technical challenge over

the years. Despite a considerable amount of study, controlling the complex interac-
tions between the various lipid components during crystallization remains essen-
tially an empirical process of studying the effects of various operating
parameters on crystal formation. Further work on the fundamental principles of
lipid nucleation, growth, and polymorphic transformation is needed to truly control
crystallization of lipids in foods.
REFERENCES
1. K. Larsson, Fette Seifen Anstrichmittel., 74, 136142 (1972).

2. L. Hernqvist, Fette Seifen Anstrichm. 86, 297300 (1984).
3. L. Hernqvist, in N. Garti and K. Sato, eds. Crystallization and Polymorphism of Fats and
Fatty Acids, Marcel Dekker, New York, 1988, pp. 97137.
4. R. W. Hartel, Crystallization in Foods. Aspen Publishers, Inc., Gaithersburg, MD, 2001.
5. K. Sato, Chem. Eng. Sci., 22552265 (2001).
6. K. Larsson, in S. Friberg and K. Larsson, eds. Food Emulsions, 3rd ed. Marcel Dekker,
New York, 1997, pp. 111140.
7. E. S. Lutton, J. Am. Oil Chem. Soc., 27, 276281 (1950).
8. K. Larsson and P. Dejmek, in S. Friberg and K. Larsson, eds. Food Emulsions. 2nd ed.
9. K. Larsson, Acta Chem. Scand., 20, 22252260 (1966).
10. A. E. Bailey, in Melting and Solidification of Fats, Interscience Publishers, New York,
1950, pp. 2129.
11. P. Walstra, in J. M. Blanshard and P. Lillford, eds. Food Structure and Behaviour,
Academic Press, Orlando, Florida, 1987, pp. 6785.
12. S. Ueno, A. Minato, H. Seto, Y. Amemiya, and K. Sato, J. Phys. Chem. B, 101, 68476854
(1997).
13. R. L. Wille and E. S. Lutton, J. Am. Oil Chem. Soc., 43, 491496 (1966).
14. T. R. Davis and P. S. Dimick, in Proceedings of the Pennsylvania Manufacturing
Confectioners Association (PMCA) Production Conference, Center Valley, PA, 1986,
pp. 104108.
15. K. Van Malssen, A. van Langevelde, R. Peschar, and H. Schenk, J. Am. Oil Chem. Soc., 76,
669676 (1999).
16. J. Schlichter-Aronhime and N. Garti, in N. Garti and K. Sato, eds. Crystallization and
Polymorphism of Fats and Fatty Acids, Marcel Dekker, New York, 1988, pp. 363393.
17. B. Rossell, Prog. Lipid Res., 23, 138 (1967).
18. R. E. Timms, Progress Lipid Res., 23, 138 (1984).
19. L. Wesdorp, Ph.D. Dissertation, Technical University of Delft, The Netherlands,
(1990).
20. W. Kloek, P. Walstra, and T. Van Vliet, J. Am. Oil Chem. Soc., 77, 643652 (2000).
21. M. L. Herrera, M. de Leon Gatti, and R. W. Hartel, Food Res. Int., 32, 289298 (1999).
22. G. Bigalli, Manufacturing Confectioner, 68, 6670, 7980 (1988).
23. R. W. Hartel and K. E. Kaylegian. in N. Garti and K. Sato, eds. Crystallization Processes in
Fats and Lipid Systems, Marcel Dekker, New York, 2001, pp. 329355.
24. J. W. Mullin, Crystallization. 3rd ed., Butterworth-Heineman Ltd., Oxford, U.K., 1993,
pp. 172260.
25. R. E. Timms, in R. J. Hamilton, ed. Developments in Oils and Fats, Chapman & Hall,
Glasgow, U.K., 1995, pp. 204223.
26. A. G. Walton, in A. C. Zettlemoyer, ed. Nucleation, Marcel Dekker, New York, 1969,
pp. 225308.
27. J. Garside, in J. M. Blanshard and P. Lillford, eds. Food Structure and Behaviours.
Academic Press, Inc., Orlando, FL, 1987, pp. 3545.
28. R. Boistelle, in N. Garti and K. Sato eds., Crystallization and Polymorphism of Fats and
Fatty Acids, Marcel Dekker, New York, 1988, pp. 189226.
29. D. Turnbull and J. C. Fisher, J. Chem. Phys. 17, 7173 (1949).
30. R. F. Strickland-Constable, in R. F. Strickland-Constable, ed. Kinetics and Mechanisms of
Crystallization, Academic Press, London, U.K., 1968, pp. 74129.
31. W. Kloek, Ph.D. Dissertation. University of Wageningen, The Netherlands, 1998.
32. M. Kellens, W. Meeussen, and H. Reynaers, J. Am. Oil Chem. Soc., 69, 906911 (1992).
REFERENCES 75
33. A. J. Wright, S. S. Narine, and A. G. Marangoni, J. Am. Oil Chem. Soc., 77, 12391242
(2000).
34. M. L. Herrera, C. Falabella, M. Melgarejo, and M. C. Anon, J. Am. Oil Chem. Soc., 75,
12731280 (1998).
35. S. Martini and M. L. Herrera, J. Am. Oil Chem. Soc., 79, 411412 (2002).
36. K. Sato and T. Kuroda, J. Amer. Oil Chem. Soc., 64, 124130 (1987).
37. D. Precht, in N. Garti and K. Sato, eds. Crystallization and Polymorphism of Fats and Fatty
Acids, Marcel Dekker, New York, 1988, pp. 305362.
38. P. Walstra and E. C. H. Beresteyn, Neth. Milk Dairy J., 29, 3565 (1975).
39. D. N. Kaschiev, Kaneko, and K. Sato, J. Colloid Interface Sci., 208, 167177 (1998).
40. N. Kaneko, T. Horie, S. Ueno, and K. Sato, J. Crystal Growth, 197, 263270 (1999).
41. P. Walstra, W. Kloek, and T. van Vliet, in N. Garti and K. Sato, eds. Crystallization
Processes in Fats and Lipid Systems, Marcel Dekker, New York, 2001, pp. 289328.
42. A. G. Marangoni, J. Am. Oil Chem. Soc., 75, 14651467 (1998).
43. I. Foubert, P. A. Vanrolleghem, B. Vanhoutte, and K. Dewettinck, Food Res. Int., 35, 945
956 (2002).
44. W. Kloek, P. Walstra, and T. Van Vliet, J. Am. Oil Chem. Soc., 77, 389398 (2000).
45. B. Vanhoutte, K. Dewettinck, I. Foubert, B. Vanlerberghe, and A. Huyghebaert, Eur.
J. Lipid Sci. Tech., 104, 490495 (2002).
46. M. Avrami, J. Chem. Phys., 8, 212224 (1940).
47. J. W. Graydon, S. J. Thorpe, and D. W. Kirk, J. Non-Crystalline Solids, 175, 3143 (1994).
48. J. W. Christian, The Theory of Transformation in Metals and Alloys, 2nd ed., Pergamon
Press, London, U.K., 1975.
49. S. Metin and R. W. Hartel, J. Am. Oil Chem. Soc., 75, 16171624 (1998).
50. I. Arvanitoyannis and J. M. V. Blanshard, J. Food Sci., 59, 197205 (1994).
51. W. L. Ng, J. Am. Oil Chem. Soc., 67, 879882 (1990).
52. H. Mulder and P. Walstra, Products and Comparable Foods, Commonwealth Agric.
Bureaux, Bucks, England, 1974, pp. 3352.
53. A. G. Marangoni and R. W. Hartel, Food Technol., 52, 4651 (2000).
54. D. Aquilano and G. Sgualdino, in N. Garti and K. Sato, eds., Crystallization Processes in
55. S. Metin and R. W. Hartel, J. Thermal Anal., 47, 15271544 (1996).
56. S. Martini, M. L. Herrera, and R. W. Hartel, J. Agric Food Chem., 49, 32233229 (2001).
57. P. S. Dimick, in N. Widlak, ed. Proceedings of International Conference on Physical
Properties of Fats, Oils and Emulsifiers with Applications to Foods, Am. Oil Chem. Soc.
Press, Champaign, IL, 2000, pp. 138161.
58. J. Nyvlt, O. Sohnel, M. Matuchova, and M. Broul, The Kinetics of Industrial Crystal-
lization, Elsevier Science Publishers, New York, 1985, pp. 189191.
59. P. R. Smith, D. J. Cebula, and M. J. W. Povey, J. Am. Oil Chem. Soc., 71, 13671372
(1994).
60. K. W. Smith, in N. Garti and K. Sato, eds. Crystallization Processes in Fats and Lipid
Systems, Marcel Dekker, New York, 2001, pp. 357380.
61. T. Koyano, I. Hachiya, and K. Sato, Food Structure, 9, 231240 (1990).
62. M. L. Herrera and R. W. Hartel, J. Am. Oil Chem. Soc., 77, 11771187 (2000).
63. G. Mazzanti, S. E. Guthrie, E. B. Sirota, A. G. Marangoni, and S. H. J. Idziak, Crys.
Growth Design, 3, 721725 (2003).
64. D. S. Grall, and R. W. Hartel, J. Am. Oil Chem. Soc., 69(8), 741747 (1992).
65. C. Garbolino, G. R. Ziegler, and J. N. Coupland, J. Am. Oil Chem. Soc., 77, 157162
(2000).
3
Polymorphism
in Fats and Oils
Kiyotaka Sato and Satoru Ueno
Graduate School of Biosphere Science, Hiroshima University
Higashi-Hiroshima, Japan
1. INTRODUCTION
Triacylglycerols (TAGs) are the major components of fats and oils and biologically
important organic molecules along with proteins and carbohydrates. In industrial
applications, TAGs are the main components in cream, margarine, and confection-
ery fats in foods and as matrix materials in pharmaceuticals and cosmetics. The
physical behavior of TAGs influences the physical properties of fat-based products,
such as appearance, texture, plasticity, morphology, and rheology. Most fat-based
products are multicomponent TAG mixtures, containing different kinds of fatty
acid moieties. Their complex physical properties are ascribed to polymorphism of
individual TAG components and their mixing behavior. Therefore, research into
the physical properties of the fat-based products usually starts with an understand-
ing of individual TAG molecules and subsequently moves on to an understanding of
the mixed systems, while combining this microscopic information with the macro-
scopic properties of texture, crystal morphology, and rheology. The macroscopic
properties of fats and oils will be discussed in other chapters of this volume.
77
78 POLYMORPHISM IN FATS AND OILS
This chapter describes the polymorphism of the principal TAGs with saturated and
unsaturated fatty acid moieties and their binary mixtures.
2. BASIC CONCEPTS OF POLYMORPHISM OF FATS
TAGs are three-fold esters of glycerol and fatty acids, having the general formula
shown in Figure 1. There is a number of fatty acid moiety, as indicated in Figure 1.
According to Figure 1, TAGs can be divided into two classes depending on the fatty
acid composition. TAGs having only one type of fatty acid are called monoacid
TAGs, and those having two and three types of fatty acids are, respectively, called
diacid and triacid TAGs, and both are categorized as mixed-acid TAGs. Almost all
natural fats and oils are mixed-acid TAGs. In addition, the diacid TAGs can be
divided into two types: symmetric and asymmetric TAGs. In the asymmetric diacid
TAGs, chiral properties are revealed: For example, sn-R1R1R2 and sn-R2R1R1 are
stereochemically different from each other, in which sn means a stereospecific
number. The same chiral properties occur in the triacid TAGs. It is noteworthy
that polymorphism of the symmetric TAGs is largely different from that of the
asymmetric TAGs.
The physical properties of TAGs are determined by the types of fatty acids that
compose them; for example, the number of saturated and unsaturated chains, cis-
and trans-double bonds, short and long chains, chains with even and odd numbers
of carbon atoms, and esterified positions of fatty acids with glycerol carbon atoms.
Fats are modified by hydrogenation, interesterification, and fractionation to produce
desirable physical properties for fat-based products.
2.1. Polymorphism of Triacylglycerols

Multiple melting points of fats had already been discovered in the nineteenth cen-
tury. Clarkson and Malkin showed that this melting behavior resulted from the
polymorphism of TAGs (1). In the crystalline state, TAG molecules adopt the ideal
sn-1 CH2O CO R1
sn-2 CH O CO R2
sn-3 CH2 O CO R3
Mono-acid TAG (R 1 = R2 = R3)

* saturated: length, even-odd
* unsaturated: length, even-odd, number-position-conformation
of double bond
Mixed-acid TAG (R 1 R2 R3)
* different length
* saturated and unsaturated
* different sn-positioned
Figure 1. A triacylglycerol molecule (R: fatty acid moiety, sn: stereospecific number).
BASIC CONCEPTS OF POLYMORPHISM OF FATS 79
conformation and arrangement in relation to their neighbors to optimize intramole-

cular and intermolecular interactions and accomplish an efficient close-packing. On
the basis of the structural studies by Larsson (2), the three fundamental polymorphs
are called a, b0 , and b. The significance of the definition of polymorphism of the
TAGs lies in unification of otherwise confused nomenclature of the polymorphic
forms of the fats differently named by researchers, such as sub-a form, vitreous
phase, and so on. In addition, the polymorphic nomenclature makes it convenient
to characterize the crystalline properties of fats employed in many applications. For
example, the structure and texture of ice cream is caused by a network of partially
coalesced a-form crystals and ice crystals that surround air bubbles to form discon-
tinuous foams (3). The small needle-like b0 crystals impart good plasticity that is
desirable in products such as margarine, shortening, and baking fats (4). Cocoa
butter replacers (CBR) and cocoa butter substitutes (CBS) can crystallize without
tempering into their stable b0 polymorph upon simple cooling. Tempering is
required for b form, which is used for chocolate, cocoa butter, and cocoa butter
equivalents (CBE) (5).
One may characterize the polymorphic forms of TAGs by thermal stability,
subcell packing, and chain-length structure as described below.
2.1.1. Thermal Stability Among the three main polymorphic forms of TAGs and
their mixtures, generally, b is the most stable, b0 is less stable, and a is the least
stable form (6, 7). A diagram of the Gibbs free energy (G H-TS, in which H, S,
and T are enthalpy, entropy, and temperature) versus T for TAG polymorphs is
shown in Figure 2. The G-T relationship determines the transformation pathways
among the polymorphs and liquid (8). The polymorphism of TAGs is monotropic,
and the G values are largest for a, intermediate for b0, and smallest for b in the solid
liquid
Temperature
Figure 2. A schematic diagram of Gibbs energy (G) and temperature of three polymorphs of a
triacylglycerol.
phase domain at low temperature. Each polymorph has its own melting temperature
(Tm) that is defined as the temperature where the G value of crystal becomes lower
than that of liquid. These thermodynamic conditions influence the kinetic aspects of
crystallization and transformation of TAGs.
The three basic polymorphic forms shown in Figure 2, which may apply to the
saturated monoacid TAGs, are largely modified when the shape of a TAG molecule
becomes more heterogeneous. For example, TAGs containing unsaturated fatty acid
moieties or saturated diacid moieties exhibit two b0 or b forms. In other cases,
b does not occur and b0 becomes most stable with the highest Tm instead. These
properties will be discussed in Section 3.
A primary concern is polymorphic crystallization in which the Ostwald step rule
is very useful (9). This rule predicts that phase changes occur step by step by way of
successively more stable phases. For the relative rate of nucleation of polymorphic
crystals shown in Figure 2, it follows that nucleation of the metastable forms such
as a and b0 occurs first before the most stable b form, when nucleation occurs under
a large supercooling or high supersaturation. When the amount of supercooling or
supersaturation is decreased, the law is broken and the most stable form tends to
nucleate at a relatively slow rate.
Because of its monotropic nature, the polymorphic transformation occurs irre-
versibly from the least stable a form to the most stable b form. The rate of trans-
formation is both time- and temperature-dependent. There are two modes of
polymorphic transformation processes: solid-solid and melt-mediated transforma-
tions. Solid-solid transformations occur below the melting points of all the poly-
morphs involved. In contrast, melt-mediated crystallization occurs when the
temperature is above the melting points of the less stable forms. Melt-mediated
crystallization involves the following processes:
1. Melting of the less stable form

2. Nucleation and growth of the more stable forms
3. Mass transfer in the liquid formed by melting of the less stable form
It has been observed in some TAGs that the rate of melt-mediated crystallization
is often much higher than that of solid-solid transformation (1014).
2.1.2. Subcell Structure Subcell structure defines a lateral packing mode of

the hydrocarbon chains (2, 15, 16). Three typical subcell structures are shown in
Figure 3. The a, b0 , and b forms have hexagonal (H), orthorhombic perpendicular
O? , and triclinic parallel (T//) subcell structures, respectively (2).
In the hexagonal subcell structure, the two-dimensional lattice is hexagonal and
gives rise to a 0.41-nm wide-angle X-ray diffraction (XRD) pattern. The chain
packing is loose, and the specific chain-chain interactions are lost because of the
ability of the carbon atoms to rotate several degrees and form disordered conforma-
tions of hydrocarbon chains. The two-dimensional lattice of an orthorhombic
perpendicular O? subcell structure is rectangular, and this represents a tightly
Figure 3. Typical subcell structures of TAG polymorphs. The a, b0 , and b forms have hexagonal
(H), orthorhombic perpendicular O ? , and triclinic parallel(T//), respectively.
packed lattice with specific chain-chain interactions. The subcell parameters of O?

are typically shown in two wide-angle XRD patterns at 0.37 nm and 0.41 nm. Tricli-
nic parallel subcell structure (T//) has an oblique two-dimensional lattice and repre-
sents tightly packed chains, in which there are specific chain-chain interactions.
This subcell structure of T// is characterized by a strong wide-angle XRD pattern
at 0.46 nm and week patterns at 0.39 nm and 0.38 nm. The values given for these
wide-angle XRD patterns of the three polymorphs are typical for the saturated
monoacid TAGs; they vary when the fatty acid moieties change from saturated to
unsaturated acids.
2.1.3. Chain Length Structure The TAG crystals form chain-length structures,
in which a repetitive sequence of the hydrocarbon chains is involved in a unit lamel-
lar along the c-axis (Figure 4) (17). One unit layer made up of one hydrocarbon
chain is called a leaflet. Several types of chain-length structures can form as shown
in Figure 4. The TAGs with the same or very similar fatty acids might form a
double chain-length structure. A triple chain-length structure is formed when the
chemical natures of one or two of the fatty acids are much different from the others.
A quarto-chain-length structure consists of two double chain-length structures,
which are combined end-to-end. A hexa-chain-length structure consists of two
Figure 4. Typical variations in the chain-length structures of triacylglycerol crystals. An arrow

means a leaflet.
triple chain-length structures. The quarto- and hexa-chain-length structures were

observed in asymmetric saturated diacid TAGs, as discussed below.
2.2. Phase Equilibria

Figure 5 shows three cases that are generally applicable to many materials, and
observed in TAG binary mixtures: solid-solution mixture, eutectic mixture, and
molecular compound forming mixture (18). Here we summarize basic properties
of the three mixture phases. Various binary mixture systems of TAGs will be
discussed in Section 4.
2.2.1. Solid-Solution Mixture In this system, a binary mixture is cooled but

neither component solidifies without containing some of the other component:
Both components are deposited simultaneously, and the deposited solid phase is
a solid-solution. Only two phases can exist in such a system: a homogeneous
liquid-solution and a solid-solution. The equilibrium phase diagram is shown in
Figure 5. Three typical phase diagrams of binary TAG mixtures represented by A and B
fractions.
Figure 5a. All mixtures of the two components have melting points intermediate
between the melting points of the pure components.
TAGs that have similar physical and chemical properties, for example, similar
melting points, chain-length, polymorphic form, and molecular volume, form
solid-solution mixtures.
2.2.2. Eutectic Mixture A general example for a eutectic mixture system is

shown in Figure 5b, where curves PQ and RQ represent the temperatures at which
homogeneous liquid-solutions begin to crystallize. Above the curves P-Q-R, the
two components are liquid. Line TQU represents the temperature at which solid
mixtures of A-molecule and B-molecule begin to melt, and the two components
are completely solid below the line T-Q-U. The small and large areas of PQT
and RQU represent mixtures of A-molecule crystals in liquid A/B-molecule and
solid B-molecule crystals in liquid A/B-molecule, respectively. It is important to
note that the eutectic is a physical mixture, not a molecular compound. Below
the eutectic temperature, all mixtures are solid.
Among binary TAG mixtures, the eutectic system is most common. Eutectic sys-
tems tend to occur when the TAGs differ in chain-length, molecular volume, shape,
or polymorphic form, but they have similar melting points.
2.2.3. Molecular-Compound-Forming Mixture A-molecule and B-molecule

of a binary system sometimes combine to form a molecular compound. If a mole-
cular compound can coexist in equilibrium with a liquid of the same composition,
the compound has a congruent melting point shown by point R (Figure 5c). The
phase diagram for this system can be split into two subdiagrams: A-molecule/
molecular compound and molecular compound/B-molecule, and each of these
subdiagrams may be considered as a eutectic mixture as shown in Figure 5c. Points
P and Q are the other two eutectic points of the subdiagrams that are placed in a
juxtaposition manner.
A molecular compound-forming mixture can occur in particular combinations of
TAGs, being based on the possibility of achieving specific interactions between
both molecules in the crystalline state. Even now, there is no general explanation
why and how specific combinations of TAGs can result in molecular compound-
forming mixtures, but several examples will be provided in Section 4.
2.3. Basic Methods for Studying the Polymorphism of Fats

Methodology for studying the polymorphism of fats, among which thermal analy-
sis, most typically, differential scanning calorimetry (DSC), X-ray diffraction
(XRD), neutron diffraction, infrared absorption spectroscopy, and nuclear magnetic
resonance (NMR), are briefly mentioned here.
DSC analysis provides the data of temperatures, enthalpy and entropy values of
melting, crystallization, and polymorphic transitions, which are prerequisites for
isolation of individual polymorphic forms and their thermal stability.
Molecular structural information, lamellar distance (long spacing), and subcell
structure (the short spacing) are calculated by small-angle and wide-angle diffraction
patterns from a powder XRD study using polycrystalline powder sample. Atomic-
level crystal structure is revealed by XRD using a high-quality single crystal.
One of the most exciting methodologies that has recently been applied to fat
polymorphism is synchrotron radiation XRD (SR-XRD). It has made it possible
to perform real-time (in situ) observations of polymorphic transformations at rapid
POLYMORPHISM OF MONOACID TRIACYLGLYCEROLS 85
rates of temperature variation as high as 5 C/min under external stimuli of shear

(19) and ultrasonication (20). Furthermore, a combined study of SR-XRD small-
angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS) with
DSC is now one of the most powerful methods for clarifying the kinetics of the
polymorphic transitions of TAGs in single-component and mixture systems.
Observing the correspondence of the DSC thermopeaks and variations in the
SAXS-WAXS patterns during rapid temperature change has clarified the mechan-
isms of complicated polymorphic transformations of binary mixtures of TAGs or
liquid-crystal to polymorphic crystal conversion, which had been overlooked
with conventional laboratory-scale XRD apparatus. Sections 3 and 4 will address
these issues.
The use of neutron diffraction provides structural information about fats in liquid
and crystalline states through interactions of neutrons with atomic nucleus that is
different from the information provided by X-ray diffraction. Neutron diffraction
studies with selective deuteration of glycerine and fatty acid chains of a TAG
indicated nematic-type liquid crystal organization of the TAG molecules in the
liquid phase (21, 22).
For molecular properties of the TAG polymorphs, local molecular structural
information such as methyl-end group, olefinic conformation, and chain-chain
interaction are unveiled by infrared (IR) spectroscopy, especially Fourier-trans-
formed infrared spectroscopy (FT-IR) (23, 24). Compared with a pioneering
work by Chapman (25), great progress has been achieved by using various FT-IR
techniques, such as polarized transmission FT-IR, reflection absorption spectro-
scopy (RAS), and attenuated total reflection (ATR) (2628).
NMR, especially cross-polarization and magic-angle spinning NMR (CP/MAS
NMR), is also a powerful tool for studying the molecular conformations of the
TAGs in a crystalline state, because CP/MAS NMR spectra give detailed informa-
tion about the local environment and mobility of specific carbon sites (2933).
3. POLYMORPHISM OF MONOACID TRIACYLGLYCEROLS
3.1. Saturated Monoacid Triacylglycerol

Saturated monoacid TAGs are of the simplest chemical shape and, therefore, have
been examined as model substances for the study of the complex fats.
The atomic-level crystal structures of the b form of saturated monoacid TAGs
were first clarified almost four decades ago (3436). Based on these structural
data, Lutton postulated the b form structures of saturated monoacid and diacid
TAGs (37).
Quite recently, van Langevelde et al. examined the crystal structures of the
b form of tripalmitin (PPP) in comparison with the results of tricaprin (CCC)
(38, 39), trilaurin (LLL) (36), and predicted the b form structure of trimyristin
(MMM) (38). As shown in Table 1, the unit cell parameters, double-chain length
structure, and T// subcell structure of the three TAGs are almost the same, except
TABLE 1. Unit Cell Parameters of b Form of Tricaprin

(CCC), Trilaurin (LLL), and Tripalmitin (PPP) (V; Unit
Cell Volume, D; Density).
Parameters CCC LLL PPP
Space group P1

P1
P1
a axis (nm) 1.218 1.208 1.195
b axis (nm) 3.156 3.661 4.684
c axis (nm) 0.549 0.547 0.545
a (deg.) 73.4 73.4 73.8
b (deg.) 100.7 100.5 100.2
g (deg.) 119.2 118.7 118.1
V (nm3) 1.7613 2.0292 2.5811
D (g cm3) 1.04 1.04 1.04
O O
O O
1 2 3
3 1
O 2
O O O O
O O O
(a) (b)
Figure 6. Two types of triacylglycerol conformations in crystal: (a) tuning folk conformation and
(b) chair conformation. Numbers correspond to carbon atoms of a glycerine group.
for the chain length parameter (b-axis). Therefore, van Langevelde et al. concluded
that, as long as the series of CnCnCn, in which n means even-number of carbon
atoms, all structure models can be predicted by extrapolation of the cell parameters
and copying of the atomic coordinates.
There are two types of molecular conformation of TAG molecules in the crystal
(39): tuning fork and chair, as shown in Figure 6. In a tuning fork conformation, the
two outer acyl chains (sn-1 and sn-3) point in one direction and the middle acyl
chain (sn-2) in the opposite direction. In contrast, a chair conformation has the
two neighboring acyl chains (sn-1 and sn-2) pointing in one direction and the third
acyl chain (sn-3) in the opposite direction. In the b form of CCC, LLL, and PPP,
asymmetric tuning fork conformation was revealed.
For b0 forms of saturated monoacid TAGs, no crystal structure has so far been
determined, because of difficulty in growing single crystals suitable for atomic-
level structure determination. However, information about the unit cell can be
obtained or calculated. For example, Figure 7 shows the density of liquid, b0 and
b forms of saturated monoacid and diacid TAGs as a function of the number of car-
bon atoms (20). The densities of the b0 forms of CCC, LLL, MMM (trimyristin),
and PPP could be determined by applying a least-square fitting procedure based on
the density of the b0 form of tristearin (SSS) (40) and crystal structure data of satu-
rated diacid TAGs (41, 42).
1.1
1.05
(CLC)
(LML)
(PSP)
density (g/cm3 )
1 (MPM)
(PPM)
0.95
0.9
0.85
30 36 42 48 54
(CCC) (LLL) (MMM) (PPP) (SSS)
Number of carbons in acyl-chains/molecule
Figure 7. Relationship between the density of a TAG and the number of carbon atoms present
in the acyl chain of its molecule. Closed circle: liquid state; open triangle: b0 form; and open
square: b form. Solid and dotted lines are the least-squares fitting curves for each state.
3.2. Unsaturated Monoacid Triacylglycerol

Natural fats and oils having Tm below ambient temperature contain the TAGs with
unsaturated fatty acid moieties. Crystallization properties of unsaturated monoacid
TAGs were examined six decades ago (43, 44). Compared with the saturated mono-
acid TAGs, polymorphism of unsaturated monoacid TAGs is more complicated,
because of diverse variations in the number and the position of double bonds of
their acyl-chain moieties, as clarified by Hagemann et al. (45). The polymorphism
of the series of positional isomers of the TAGs having cis- and trans-octadecenoic
acids (C18, with one double bond) is summarized as follows:
1. a form occurred except for cis 12, 13, 15, and trans 10, where
means the position expressed as the number of the carbon atom counted
from the glycerol back bone at which a double bond is placed,
2. For cis-type TAGs, three b0 forms were observed for cis 7, 9, 11, and
13, but not for 5 and 15,
3. For trans-type TAGs, two b0 forms were observed for trans 11 and 14,
one b0 form for 13, whereas b0 did not appear for 4, 5, 6, 7, 8, 9,
10, 12, and 15,
4. b form was observed in all of the TAGs.
The complexity of polymorphism, especially with regard to the occurrence

of multiple b0 form, may be ascribed to the variety of positional isomers and
cis-trans-conformation.
Triolein (OOO) has been studied over many years. Some inconsistency, however,
still remains among several reports; for example, Wheeler et al. (43) and Ferguson
and Lutton (44) observed an intermediate form, whereas three b0 forms, b01 , b02 , b03 ,
were isolated by Hagemann et al. (45). This kind of inconsistency for the presence
of b0 form may be caused by inconsistency in thermal treatment and purity of the
samples employed in the experiments. We have worked on the polymorphism of
high-purity OOO (>99 %, supplied from Nippon Oil and Fats Co.) using DSC,
X-ray diffraction, and FT-IR (Ueno and Sato, unpublished work). Six polymorphs:
a, b03 , b02 , b01 , b2 , and b1 were isolated, and thermal and structural properties of the
six forms are shown in Table 2. Figure 8 shows the polymorphic transition path-
ways among the six polymorphs and melt of OOO. Two types of transitions
were observed: liquid ! a ! b03 ! b02 ! b2 and liquid ! b01 ! b1 . The former
transition occurred after rapid cooling (20 C/min) to about 80 C and subsequent
TABLE 2. Polymorphism of Triolein (Ueno and Sato, Unpublished).
Polymorph a b03 b02 b01 b2 b1
Tm (?C) 37.5 24.9 15.5 5.8 4.7 5.9

Subcell H O? O? O? T== T==
H (kJ/mol) 110 120
Melting point: Tm , enthalpy of fusion: H.

Figure 8. Polymorphic transition pathways of triolein.
heating, in which crystallization of a and successive transformations to b2 through

intermediate two b0 occurred. By contrast, with slow cooling, the liquid crystallized
in b01 form, which transformed to b1 by subsequent heating. It was interesting to
observe that no direct transformation occurred from b02 to b01 and from b2 to b1 .
Figure 9. Synchrotron-radiation X-ray diffraction patterns of polymorphic transformation of

trielaidin taken during temperature variation shown in an inserted figure (unit, nm). Q: wave
vector.
It is not easy to understand how such a kind of individual transformation pathway

occurs.
As a typical trans-monounsaturated TAG, trielaidin (EEE) with trans-o9- octa-
decenoic acid has been examined. Carter and Malkin (46) reported three poly-
morphs, a, b0 , b. Since then, an argument was made to whether the intermediate
b0 is present or not. In 1990, Desmedt et al. (47) re-examined the existence of
the b0 form by using DSC and the powder XRD pattern, showing that the b0
form was indeed crystallized after the melting of a form, but soon transformed
to b. Therefore, it is difficult to obtain the XRD patterns or FT-IR data of the b0
form of EEE at some fixed temperature.
To further investigate the existence of b0form of EEE, we have carried out time-
resolved synchrotron radiation X-ray diffraction (SR-XRD) (Ueno and Sato, unpub-
lished). Figure 9 shows the time-resolved small-angle SR-XRD patterns, which
reveals the occurrence of a polymorph having the long spacing value of 5.1 nm
on cooling, its conversion to the polymorph with the long spacing value of
4.6 nm on heating, and disappearance of the second polymorph during further
heating. According to previous reports (46, 47), it was evident that the former
polymorph corresponds to the a form that crystallized when temperature was
quenched from 50 C to 13.6 C, and the second form is b. When temperature
was jumped to 22 C, a melted and the b form crystallized, which then melted
when temperature increased to 48 C. During this temperature variation, b0 form was
not observed by the in situ SR-XRD study. Therefore, our conclusion is that
trielaidin has the b0 form, which is very unstable.
4. POLYMORPHISM OF MIXED-ACID TRIACYLGLYCEROLS
Understanding the polymorphism of mixed-acid TAGs is of significance because

the fatty acid compositions of natural fats are generally heterogeneous; namely,
combinations of fatty acids in TAGs are diverse with respect to carbon number,
mixing of saturated-unsaturated chains, the position and the number of double
bond of the unsaturated fatty acids, and so on. This group of mixed-acid TAGs
is classified to two types: (1) saturated mixed acid type and (2) saturated and
unsaturated mixed acid type.
4.1. Saturated Mixed Acid Triacylglycerols

4.1.1. Polymorphic Behavior In this section, the polymorphism of the C16-C16-
Cn, CnCn 2Cn, and CnC2Cn TAG series is discussed in terms of their diversity in
occurrence of polymorphic structures and their thermodynamic stability and mole-
cular structures.
C16-C16-Cn represents a series of homologous TAGs, in which n, the carbon
number of even-numbered carbon atoms of the sn-3 fatty acid chain, varies from
0 to 16. Systematic research on C16-C16-Cn by using XRD, DSC, and FT-IR
POLYMORPHISM OF MIXED-ACID TRIACYLGLYCEROLS 91
techniques (42, 4854) showed remarkable diversity in their polymorphism. The a

form was present in C16-C16-Cn. For the b0 form, however, there were many varia-
tions. One b0 form was present in C16-C16-C2 through C16-C16-C8, but three b0
forms (b03 , b02 , b01 ) were observed in C16-C16-C10, whereas two b0 forms (b02 , b01 )
were isolated in C16-C16-C12 and C16-C16-C14. As for b form, one b form was pre-
sent in C16-C16-C2 through C16-C16-C12. There is, however, no b form in C16-C16-
C14, which revealed the most stable form as b01 . Moreover, chain-length structures
of the polymorphic forms in C16-C16-Cn were complicated. For example, single-
chain-length structure appeared in the a form of C16-C16-C4, C16-C16-C6, and
C16-C16-C8, and hexa- and quarto-chain-length structures were observed in b01 of
C16-C16-C10 and b02 of C16-C16-C14. Correspondingly, the melting point decreased
with increasing carbon number of n in C16-C16-Cn varying from 0 to 6, whereas it
increased with n from 8 to 16. The unique properties of the series of C16-C16-Cn, in
particular for C16-C16-C14 have been discussed elsewhere (51, 54). Here we discuss
the polymorphic structures and thermal transformation pathways of C16-C16-C10
(50), as a typical example of C16-C16-Cn.
Figure 10 shows the polymorphic transformation pathways of five polymorphs of
C16-C16-C10 together with melt. On quenching of melt, the isotropic liquid pro-
duced a form with a double chain-length structure, which melts at 22 C. There
are two transformation pathways starting from the a form. The first transformation
pathway is a ! b03 ! b02 ! b, in which the subcell structure changed from H to
two O? , keeping the double chain length structure in a ! b03 ! b02 . Then, the chain
length structure changed from double to triple, and the subcell structure changed
from O? to T// when b02 transformed to b. In the second transformation pathway
of a ! melt ! b01 , the subcell structure and chain-length structure are changed
Figure 10. Polymorphic transformation b pathways of C16-C16-C10. Chain-length structure is

shown in parenthesis for each polymorph.
drastically. In both cases, many details of the polymorphic transfrmation mechan-

isms of C16-C16-C10 remain uncertain. Similarly, interesting phenomena revealed in
the polymorphic behavior of C16-C16-Cn are not understood, e.g., the occurrence of
the a form (single-chain-length structure) and the b0 form (tripe-chain-length) in
C16-C16-C4, C16-C16-C6 and C16-C16-C8, and disappearance of b in C16-C16-C6
and C16-C16-C8. All of these properties indicate that intramolecular and intermole-
cular interactions are largely affected by changing the molecular shape of the
fatty acid moieties placed at the sn-3 position. Understanding the polymorphism
of asymmetrical mixed-acid TAGs such as C16-C16-Cn must be valuable for the for-
mation of structured fats.
In the CnCn 2Cn-type TAGs in which n is even-numbered and varying from 10
to 16, the most stable form is the b0 form, not the b form. The long spacing values
and melting points of the b0 forms of these TAGs linearly increase with increasing
number of n (41, 55).
The CnC2Cn-type TAGs, in which n is even-numbered varying from 10 to 18,
were investigated by Zacharis et al (56). Three polymorphs, the lowest melting,
the intermediate, and the highest melting forms, were observed by thermal analysis.
As for the highest melting form, the following physical properties were found: (1)
The subcell is T//, (2) enthalpy of fusion, enthalpy of resolidification, and lamellar
spacing increased linearly with increasing length of the acyl chains; and (3) the
hexa-chain-length structure was formed in C14C2C14 and C16C2C16. Little informa-
tion has been obtained for the lowest and intermediated melting forms.
4.1.2. Atomic-Level Crystal Structures of b0 Form We discuss crystal struc-

tures of the b0 forms of two saturated diacid TAGs, C10C12C10 (41) and C16C16C14
(42). Both are the first b0 polymorphs of TAG that have been analyzed at the atomic
level by using single crystals. Two different types of the b0 structures were found as
revealed in the unit cell parameters shown in Table 3.
The unit cell structure of C16C16C14 b02 is shown in Figure 11 (42). The unit cell
is stacked in the quarto-chain-length structure, which is constructed by two double
layers (I and II in Figure 11) in such a way that the methyl end groups of one double
layer are faced with those of another double layer at the center of the unit cell in
the a-b plane. The chain axes of the two double layers in the unit cell are alternately
TABLE 3. Unit Cell Parameters of b0 Forms of C10 C12 C10

and C16 C16 C14
C10 C12 C10 C16 C16 C14
Space group Iba2 C2

a axis (nm) 5.57368 1.6534
b axis (nm) 2.22783 0.7537
c axis (nm) 0.56945 8.1626
b (deg.) 90 90.28
density (g cm3) 1.04 1.018
Figure 11. Unit cell structure of C16C16C14 b02 .
inclined against the lamellar interface in the b-c plane. The methyl end groups
make zigzag arrangements, but the zigzag angles of the outer interface
y1 9:6 and inner interface y2 38 are different. A hybrid-type orthorhom-
bic perpendicular subcell is formed, because of the presence of two asymmetric
units in a unit cell. The molecules of two aymmetric units of C16C16C14 have the
chair conformation defined in Figure 6.
C10C12C10 b0 crystallizes in a chair conformation with the O? subcell, having a
bend at the glycerol moiety as shown in Figure 12 (41). There is no zigzag methyl
end stacking with a flat lamellar interface. Based on the crystal structure of
C10C12C10 b0 , van Langevelde et al. (39) determined the structure of C14C16C14
b0 , using the powder XRD patterns of polycrystalline samples. They concluded
that the two b0 forms of C10C12C10 and C14C16C14 are identical, except for the
chain-length distance.
The above results of the two b0 forms indicate that there is diversity in the
crystal structures of b0 that are affected by chain-chain interactions of the diacid
TAGs.
Figure 12. Unit cell structure of C10C12C10 b0 .
4.2. Saturated-Unsaturated Mixed Acid Triacylglycerols

Saturated-unsaturated mixed acid TAGs are the main components of vegetable fats
and fish oils. The basic polymorphism of saturated-unsaturated mixed acid TAGs is
more complicated than that of the saturated monoacid TAGs (28, 33, 5765). The
chain-chain interactions between the saturated and unsaturated fatty acid moieties
are the essential determining factors of this complexity (23, 63).
4.2.1. 1,3-Disaturated-2-Unsaturated Mixed Acid Triacylglycerols In this

section, we consider the polymorphic behavior of a series of Sat.Unsat.Sat.TAGs,
in which the sn-2 acid moieties are oleic, ricinoleic, and linoleic acids and the even-
numbered saturated acids (palmitic, stearic, arachidic and behenic acids) are placed
at the sn-1 and sn-3 positions.
Figure 13 shows a model of the polymorphic forms of SOS (1,3-distearoyl-2-
oleoyl-sn-glycerol) (58). Polymorphic transformations occur from a to b1 through
Figure 13. A schematic model of polymorphic transformation of SOS.
g, b0 , and b2 . Compared with the saturated monoacid-type and diacid-type TAGs,

the occurrence of the g form is the unique feature of this group of TAGs. In the case
of SOS, the chain-length structure converts from double to triple, and the subcell
structures change in different manners between oleic and stearic acid chains. These
changes are the result of steric hindrance of the stearic and oleic acid moieties,
making the polymorphic transformation of SOS complicated. The main structural
properties of SOS are briefly described below;
1. a form. The double-chain-length structure determined by the SAXS spectra

assumes the coexistence of the stearoyl and oleoyl moieties in the same
leaflets. The hexagonal subcell shown in the XRD WAXS patterns and FT-IR
spectra of d(CH2) and r(CH2) modes lead to a disordered aliphatic conforma-
tion. No specific structure was shown for the olefinic conformation, because
no detectable IR band of g CH was seen (61) and the two carbons adjacent
to the cis-double bonds were equivalent because of the NMR spectra.
2. g form. The long spacing value of 7.05 nm assumes a triple-chain-length
structure, in which the oleoyl and stearoyl leaflets are separated through the
chain sorting during the ag transformation. The stearoyl leaflet assumes a
specific parallel packing, and the hexagonal subcell structure still remains in
the oleoyl leaflets, as verified by FT-IR spectral bands of SOS containing fully
deuterated stearoyl and hydrogenated oleoyl chains (61).
3. b0 form. The long spacing value 7.00 nm determined by the SAXS peak
assumes the triple-chain-length structure, which is constructed by the stearoyl
leaflet with the O? subcell and the oleoyl leaflet with hexagonal subcell, as
shown by FT-IR spectral bands of SOS containing fully deuterated stearoyl
and hydrogenated oleoyl chains (61). The 13C NMR spectra showed clear
differences between the two carbons adjacent to the cis-double bond and the
three glycerol carbons.
4. two b forms. The long spacing values of the triple-chain-length structure were
6.75 nm for b2 and 6.60 nm for b1. The subcell structures of the stearoyl and
oleoyl leaflets are T// in b1 . The subcell structure in b2 was very close to T//
for the two leaflets, but very subtle differences were detectable between the
two b forms.
TABLE 4. Polymorphism of POP.
H Long Chain-Length
Form Tm (?C) (kJ/mol) Spacing (nm) Structure
a 15.2 68.1 46.5 double

g 27.0 92.5 65.4 triple
d 29.2 107.5 62.5 triple
b02 30.3 95.5 42.4 double
b01 33.5 98.3 42.4 double
b2 35.1 124.4 61.0 triple
b1 36.7 130.2 61.0 triple
POP (1,3-palmitoyl-2-oleoyl-sn-glycerol) is a homologous substitution of the

stearoyl moiety in SOS with the palmitoyl moiety. It was anticipated that POP
might show the same polymorphic behavior as SOS. However, a few differences
were observed regarding intermediate forms as explained in the following
(Table 4) (58):
1. Two b0 forms appeared, having a double chain-length structure.

2. During the polymorphic transformation from a to b1 forms, the chain-length
structure changes as double (a) ! triple (g) ! double (two b0 forms) ! tri-
triple (two b forms). The alternative variations of the chain-length structure
among double and triple are only detected in POP.
3. Another intermediate form with triple chain-length structure, d, was
observed. The cause of complexity in the polymorphism of POP is still open
to question.
However, the polymorphism of SOS is common to the other Sat.Unsat.Sat.TAGs,

as evidenced in the polymorphism of SRS(R, ricinoleoyl) and SLS (L, linoleoyl).
Although SOS, SRS, and SLS share the same polmorphic nature illustrated in
Figure 13, remarkable differences are seen in the presence or absence of the stable
forms of b0 or b among the three TAGs. Namely, SRS has no b form and two b0
forms (64), whereas b0 and b forms are absent in SLS (65). Thermal data of the
TABLE 5. Thermal Properties of Polymorphism of SOS, SRS, and SLS.
SOS SRS SLS
a g b0 b2 b1 a g b02 b01 a g
Tm ( C) 23.5 35.4 36.5 41.0 43.0 25.8 40.6 44.3 48.0 20.8 34.5
Hm (kJ/mol) 47.7 98.5 104.8 143.0 151.0 58.1 119.64 171.19 184.76 40.9 137.4
Sm (J/mol/K) 160.8 319.2 338.5 455.2 477.6 194.35 381.32 539.29 575.31 139.2 448.7
a
Tm : temperature of melting; Hm : enthalpy of melting; Sm : entropy of melting.
Figure 14. Wide-angle X-ray diffraction patterns of polymorphic forms of SRS and SLS
(unit, nm).
polymorphic forms of SRS and SLS are shown in Table 5 together with those of
SOS. The a and g forms present in the two TAGs showed the same molecular
structures, as revealed in wide-angle XRD patterns of SRS and SLS shown in
Figure 14. It is postulated that hydrogen bonding in the ricinoleoyl chains is so tight
that the O? subcell is stabilized through the glycerol groups, probably making b0
the most stable in SRS (Figure 15a) (64). The hydrogen bonding in SRS may make
the enthalpy and entropy values for melting of the b0 forms much higher than b0
forms of SOS.
Figure 16 shows SR-XRD patterns of SLS taken during a temperature variation
from 50 C to 10 C, kept at 10 C for about 10 min, and heated rapidly to 50 C (65).
S
S
bs

as

---- Hydrogen bonding
S
S
S
(a) (b)
Figure 15. Schematic model of polymorphic structures of SRS and SLS. (a) A postulated
structure of b0 forms of SRS, in which a ricinoleic acid leaflet is shown. (b) Polymorphic
transformation from a to g in SLS.
The occurrence of a on cooling, transformation from a to g, and melting of g on

heating of SLS were clearly observed. In particular, rapid heating from 10 C to
50 C clearly showed the transformation from a to g at 20 C. No b0 or b forms
were detected during further incubation of SLS after melting of g, or also after
long incubation of g below its melting point. In SLS, the interactions among the
linoleoyl chains at the sn-2 position, each of which has two cis-double bonds,
may stabilize the g form, prohibiting the transformation into more stable forms
of b0 or b. For this reason, the enthalpy and entropy values for melting of the g
form of SLS are much larger than those of SOS and SRS (see Table 5). The trans-
formation from g to b0 or b in SOS is associated with an inclined chain arrangement
with respect to the lamellar interface, which might be prohibited by the linoleoyl
chain-chain interactions in SLS.
4.2.2. 1,3-Diunsaturated-2-Saturated Mixed Acid Triacylglycerols The

polymorphic behavior of symmetric diacid TAGs, 1,3-dioleoyl-2-staearoyl-sn-gly-
cerol (OSO), 2-elaidoyl(OEO), and 2-vaccinoyl (OVO) glycerols was studied by
Kodali et al. (66). On quenching from the melt, OEO and OVO formed a double-
chain- length b0 form, whereas OSO formed the a form. At 7 C, a of OSO
quickly transformed to b0 . Long-time incubation of OVO, OEO, and OSO trans-
formed b0 form into b form of the triple chain-length structure, in which the two
oleoyl chain leaflets are segregated from the vaccinoyl, elaidoyl, and stearoyl chain
leaflet. It can be assumed that the driving force to form the triple chain-length b
Figure 16. Time-resolved SR X-ray scattering patterns of SLS (unit: nm). At left is the
temperature change with time.
form of the three TAGs may be the fact that the saturated or trans-unsaturated acyl
chains at the sn-2 position do not pack with the bent oleoyl chains at the sn-1 and
sn-3 positions in the stable polymorphic forms. This mechanism is essentially the
same as that present in the Sat.-Unsat.-Sat. TAGs.
The saturated-unsaturated mixed-acid TAGs involving trans-unsaturated acids
have recently been examined with and without the effects of surfactant additives
(67, 68). It is notable that b0 is most stable in PEP (1,3-dipalmitoyl-2-elaidoyl-
sn-glycerol). On the other hand, SES (1,3-distearoyl-2-elaidoyl-sn-glycerol) has
the most stable form of b. In contrast to the stabilization of b in ESS and SEE,
the most stable form of EPP and PEE is b0 . The mechanisms for the stabilization
of the b0 in PEP, EPP, and PEE remain unknown. It seems that the methyl end stack-
ing mode may be a key factor, although further clarification is needed.
By now, we have discussed the polymorphism of mono-acid and mixed-acid
TAGs. To summarize, Table 6 shows the number and types of polymorphic forms
of principal TAGs.
TABLE 6. Occurrence of Polymorphic Forms in Representative

Triacylglycerols.
Triacylglycerol Polymorphic Forms
SSS (tristearoyl-glycerol) a, b0 , b
OOO (trioleoyl-glycerol) a, b03 , b02 , b01 , b2 , b1
EEE (trielaidoyl-glycerol) a, b0 , b
PP14 (1,2-dipalmitoyl-3-myristoyl-sn-glycerol) a, b02 , b01
PP10 (1,2-dipalmitoyl-3-decanoyl-sn-glycerol) a, b03 , b02 , b01 , b
CLC (1,3-dicaproyl-2-lauroyl-sn-glycerol) a, b0
SOS (1,3-distearoyl-2-oleoyl-sn-glycerol) a, g, b0 , b2 , b1
POP (1,3-dipalmitoyl-2-oleoyl-sn-glycerol) a, g, d, b02 , b01 , b2 , b1
BOB (1,3-dibehenoyl-2-oleoyl-sn-glycerol) a, g, b0 , b2 , b1
SRS (1,3-distearoyl-2-rycinoleoyl-sn-glycerol) a, g, b02 , b01
SLS (1,3-distearoyl-2-linoleoyl-sn-glycerol) a, g
OSO (1,3-dioleoyl-2-stearoyl-sn-glycerol) a, b0 , b
5. FAT MIXTURES AND POLYMORPHISM
Fats are multicomponent in two ways: (1) a fat phase contains many different types
of TAGs and (2) each TAG molecule involves different types of fatty acid moieties,
namely, mixed-acid TAGs. Therefore, it is important to precisely analyze physical
and chemical properties of the TAGs in multicomponent systems to understand
thermal, structural, and rheological properties of the real food fat systems
(69, 70). Particularly, one may note that the kinetic properties of the molecular
compound-forming mixture phase are closely related to fat blending and interester-
ification in food technology (71, 72) and separation of liquid/solid fractions from
natural oil resources (73, 74). As a first step in the investigation of multicomponent
fat systems, the phase behavior of binary TAG mixture systems has been studied by
many researchers (7584).
The phase behavior of the binary TAG mixtures is classified into three cases:
solid-solution, eutectic, and molecular compound formation, as introduced in Sec-
tion 1. Peculiarities in the mixtures of the TAGs may be explained by the following:
1. The TAGs with similar chemical structures tend to form a solid-solution phase.
2. A eutectic phase is formed between TAGs whose molecular shapes are
largely different.
3. Specific interactions result in the formation of a molecular compound as
reviewed elsewhere (17, 54).
4. In addition, influences of polymorphism make the phase behavior more
complicated.
5.1. Binary Mixtures of Saturated-Acid Triacylglycerols

Rossell (85) suggested that a eutectic phase with a limited region of solid solution
was formed for the stable b form; yet, the solid solution phase was formed in the
FAT MIXTURES AND POLYMORPHISM 101
metastable a and b0 in the mixtures of the saturated-acid TAGs. In the 1990s, this
behavior was precisely analyzed by a time-resolved SR-XRD study for the PPP/
SSS mixtures (78). The 50:50 mixture of PPP/SSS crystallized in a by quenching
the mixed liquid. A single long-spacing peak was the evidence of the solid-solution
of the mixture of a. Upon heating, the a form transformed to b0 and subsequently
to b. The miscible b0 form also appeared on cooling from the liquid phase. The
miscibility was, however, disrupted when the b0 transformed to b upon heating,
as expressed in a splitting of the long spacing pattern.
Very recently, the phase behaviors of the other types of saturated mono-acid
binary mixtures TAGs, LLL (trilauroyl-glycerol)/(trimyristoyl-glycerol), LLL/PPP,
and LLL/SSS, were examined by a SAXS/WAXS simultaneous measurement of
SR-XRD (86).
As an example, the SRXRD patterns of the LLL/PPP 60/40 mixture taken
during cooling and heating processes are shown in Figure 17. During cooling, it
was clearly shown that the b0 form of LLL and the a form of PPP were crystallized.
As for the LLL fraction, direct crystallization of b0LLL with a SAXS peak at 3.2 nm
and WAXS peaks at 0.42 and 0.38 nm occurred without the crystallization of aLLL .
Almost at the same time, aPPP with a SAXS peak at 4.6 nm and a WAXS peak at
Figure 17. Time-resolved synchrotron radiation X-ray diffraction patterns of concentration ratio
LLL/PPP 60/40. At left is the temperature change with time (unit: nm).
PPP LLL
70 Liquid
60
Temperature (C)
PPP + L
50
PPP
40 LLL + PPP
30 LLL + PPP LLL
PPP
20
'LLL + PPP LLL
10
0 20 40 60 80 100
LLL (molar %)
Figure 18. Diagram of the polymorphic occurrence for the LLL/PPP mixtures.
0.42 nm appeared at 35.2 C. Upon heating, b0LLL transformed to bLLL at about

25 C, as identified by the SAXS peak at 3.1 nm and WAXS peaks at 0.46, 0.39,
and 0.38 nm. On the other hand, aPPP transformed to bPPP with a 4.0-nm SAXS
peak at 35.2 C. The intensity of the (001) SAXS peak of bPPP (4.0 nm) started
to increase soon after the melting of bLLL at 46.0 C. This suggests that the presence
of bLLL might hinder the transformation from aPPP to bPPP . The results were
obtained for the LLL concentrations from 50% to 90%. Hence, the immiscible phases
were formed in the LLL/PPP mixture system for the three polymorphic forms.
Figure 18 shows the phase behavior of the LLL/PPP mixtures, which is
subdivided into the three regions:
1. In the LLL concentrations above 90%, the phase behaviors of the LLL/PPP
mixtures were mainly governed by LLL.
2. In the LLL concentrations from 50% to 90%, the b0 -b transformation of the
LLL fraction and the a-b transformation of the PPP fraction occurred
separately. This indicates that phase separation occurred in the three poly-
morphic forms.
3. In the LLL concentrations below 50%, the LLL fraction was dissolved in the
PPP fraction. Hence, the phase behaviors of the LLL/PPP mixtures were
mainly governed by PPP.
Figure 19 shows the phase behavior of the polymorphic occurrence for the LLL/
MMM mixtures obtained from DSC and SR X-ray scattering experiments. This
diagram indicates the following three points:
MMM LLL
70
Liquid
60
Temperature (C)
50 MMM + L
40 LLL + MMM
30 LLL-MMM
20
LLL-MMM
10
0 20 40 60 80 100
LLL (molar %)
Figure 19. Phase behavior of the polymorphic occurrence for the LLL/MMM mixtures.
1. b0 was formed in the mixture system, whereas a transformed directly to b in

LLL and MMM.
2. Miscible solid-solution phases were formed in the metastable a and b0 forms
of the mixtures.
3. A eutectic phase was formed in the most stable b form.
These three results are consistent with the results from the PPP/SSS system.
Consequently, it can be concluded for the mixtures of LLL-MMM, LLL-PPP,
LLL-SSS, MMM-PPP, and PPP-SSS that the TAG binary mixtures are miscible
in metastable polymorphs of a and b0 forms when the difference in the number
of carbon atoms of the fatty acid moieties, n, equals 2, whereas immiscible mix-
tures are found in all polymorphic forms when n is larger than 2. Results obtained
for these mixture systems may indicate a relationship between polymorphism and
phase behavior of the binary mixtures of the saturated-acid TAGs in such a way that
rotational freedom of hydrocarbon chains and entropy of methyl-end stacking are
crucial factors determining the polymorph-dependent phase behavior.
As discussed in Section 1, hexagonal-packed a has the ability of the carbon
atoms to rotate several degrees and form disordered conformations. Hydrocarbon
chains of b0 and b are all ordered conformations except for near methyl-ends, which
have a little rotational freedom. When two types of saturated monoacid TAGs
with different fatty acids are mixed, the molecules are arranged in a double
chain-length structure because of the interactions among glycerol backbones. Thus,
this crystal structure contains the methyl-end stacking gap. For the LLL/MMM
mixtures, a and b0 polymorphs form solid-solution phases. These polymorphs con-
tain disordered methyl-end groups so that they can accommodate a methyl-end
stacking gap. In contrast, b polymorph has all-trans-hydrocarbon chains and these

rigid chains cannot adjust themselves to their circumstance. Therefore, b poly-
morph shows a eutectic phase. As for the LLL/PPP and LLL/SSS mixtures, eutectic
phases occur for all polymorphs. Because of large differences in carbon numbers
for fatty acid chains between LLL and PPP n 4, and between LLL and
SSS n 6, there are very large methyl-end stacking gaps in these crystals.
Therefore, the increased entropy of methyl-end stacking becomes predominant
and phase separation must be favored thermodynamically for all polymorphs.
5.2. Binary Mixtures of Saturated-Unsaturated Mixed-Acid

Triacylglycerols
In the 1960s, Rossell reported (85) that the binary mixtures of saturated monoacid
TAGs and unsaturated monoacid TAGs form an immiscible phases, whereas Moran
has suggested (87) that a molecular compound is formed in some binary mixtures of
saturated-unsaturated mixed-acid TAGs. It was early 1990s when the formation of
molecular compounds was observed in various mixtures of saturated-unsaturated
diacid TAGs. Moreover, a miscible mixture phase was discovered in the mixtures
of SOS-SLS. Recent studies of various types of the TAG binary mixtures clarified
the formation of the molecular compound crystals at the 50/50 concentration ratio:
1,3-distearolyl-2- oleoyl-sn-glycerol/ 1,2-distearoyl-3-oleoyl-rac-glycerol (SOS/
SSO) (77), SOS/1,3- dioleoyl-2-stearoyl-sn- glycerol (SOS/OSO) (79), 1,3-dipalmi-
toyl-2-oleoyl-sn- glycerol/ 1,2-dipalmitoyl- 3-oleoyl-rac-glycerol (POP/PPO) (81),
and POP/1,3-dioleoyl-2- palmitoyl-sn-glycerol (POP/OPO) (82). These properties
are related to molecular-level understandings of the chain-chain interactions occur-
ring in biomembrane lipids containing the saturated-unsaturated mixed acid moi-
eties (16, 23, 8891). Table 7 summarizes the phase behavior of binary mixtures
of saturated-unsaturated diacid and triacid TAGs.
A molecular compound of b form, bC , was formed at the 1:1 concentration ratio
of the binary mixtures of PPO-POP, and SSO-SOS, giving rise to two monotectic
TABLE 7. Phase Behavior of Binary Mixtures of

Saturated-Unsaturated Diacid and Triacid TAGs.
Phase Behavior Mixture Systems
Miscible SSS-SSE
POS-SOS
SOS-SLS
Immiscible
Eutectic POP-PPP
Molecular compound POP-PPO
forming SOS-SSO
POP-OPO
SOS-OSO
Abbreviations: S: stearoyl, P: palmitoyl, O: oleoyl, L: linoleoy,

E: elaidoyl.
TABLE 8. Thermal Properties of Polymorphism of Molecular Compound of PPO-POP

and SSO-SOS, Together with PPO and SSO.
PPO-POP 50-50 PPO
ac b0c bc a b
Tm ( C) 15.5 28.0 31.2 18.5 35.2
Hm (kJ/mol) n.a 90 97 n.a 104
Sm (J/mol/K) n.a 299 319 n.a 337
Ling spacing (nm) 4.6 4.2 4.1 7.8 6.5
Chain length double double double triple triple
SSO-SOS 50-50 SSO

Tm ( C) 27.5 34.0 40.6 131.6 41.4
Hm (kJ/mol) 51.2 73.8 138.2 65.8 101.0
Sm (J/mol/K) 170.3 240.3 440.5 206.1 321.1
Long spacing (nm) 5.4 5.0 4.5 8.50 7.08
Obtain length double double double triple triple
n.a.: not available.
phases of the component TAG and the compound in juxtapositional. Table 8 shows
the physical properties of the polymorphs of PPO/POP compound and SSO/SOS
together with PPO and SSO. The a and b0 forms of PPO and SSO are triple-
chain-length structure, and all the polymorphic forms of POP and SOS except
for a (SOS) or a and b0 forms (POP) are triple-chain-length structure. However,
the three polymorphs of the molecular compounds of PPO-POP and SSO-SOS
are a double chain-length structure. Figure 20 shows the phase diagram of the stable
c
SOS + L SOS + L
40
c + SSO
Temperature (C)
SOS + c
30
c SSO + L
SOS + L c + SSO
20 SOS + c
0 50 100
SSO concentration (%)
Figure 20. Phase behavior in the SOS/SSO mixtures.

Figure 21. Structure models of stable polymorphs of SOS, SSO, SOS/SSO compound crystal.
b form and metastable a forms of the SOS/SSO mixtures examined by DSC and
XRD experiments (92). The most stable forms of SOS and SSO are b1 and b0 ,
respectively. The phase diagrams of a, b0 , and b polymorphs in the binary mixture
of PPO-POP are already explained elsewhere (54, 81). Therefore, the formation of
the molecular compound, which is accompanied with the conversion from triple- to
double-chain-length structure, is a common feature of the binary mixtures of Sat.-
Oleic-Sat. TAGs and Sat.-Sat.-Oleic TAGs. Figure 21 illustrates structure models of
SOS, SSO, and SSO-SOS molecular compounds, in which one leaflet is formed of
palmitoyl or stearoyl chains and the other leaflet contains the mixture of palmitoyl
and oleoyl chains or stearoyl and oleoyl chains.
It should be noted in Figure 20 that bc form is a congruent-type molecular com-
pound, and its melting point is lower than the most stable forms of SOS and SSO.
This raises interesting questions, such as how the molecular compound could be
structurally stabilized, and how its crystallization is kinetically favored. The latter
question is important, because the supercooling value with respect to bc is lowest
compared with the supercooling values of b1 of SOS and b0 of SSO. Nevertheless,
bc is crystallized when the mixture liquid is cooled below its Tm. Moreover, an
in situ XRD study showed that the rate of crystallization of bc is remarkably higher
than b1 of SOS and b0 of SSO. When the molecular compound is formed, steric
hindrance between saturated and oleic acid chains may be caused. Supposing
that the double-chain-length structure is formed in a liquid phase (93, 94), we
assume that formation energy of a crystal nucleus of the double-chain-length
40
: endothermic
liquid : exothermic
Temperature (C)
30

20

0 20 40 60 80 100
SLS concentration (%)
Figure 22. Phase behavior in the SOS/SLS mixtures.
structure may be minimized, compared with those having the triple chain-length
structure, and the nucleation frequency of the double-chain-length molecular com-
pound crystals may become higher. In this regard, a small-angle diffraction pattern
peak at 4.5 nm without the presence of a wide-angle diffraction pattern occurred,
long before the crystallization of bc from the molten sample of SSO/SOS 1/3
when the mixture liquid was cooled to 37 C (92). This indicates the presence of
the smectic liquid crystalline phase, which may be a precursor of the nucleation
of bc .
The binary mixture systems of SOS-OSO (79) and POP-OPO (83) were exam-
ined by DSC, XRD, and FT-IR, giving rise to two monotectic phases of POP (SOS)/
compound and compound/OPO(OSO) in juxtaposition.
A new result was obtained for the mixture of SLS/SOS in which a solid-solution
mixture was observed in the a and g forms in all concentration ranges: the double
chain length in a phase and the triple chain length in g, as shown in Figure 22 (92).
The miscible g form did not transform to the b0 form, when the mixtures were sub-
jected to simple cooling from high-temperature liquid to low-temperature solid
phase. Incubation of g around its Tm also did not result in transformation to the
other polymorph; instead the miscible phase was retained. The a-melt-mediated
transformation, however, caused the disruption of the solid-solution phase, and
immiscible phases of gSLS and b0SOS were formed in concentration ranges of SLS
below 30%. By contrast, in SLS concentration ranges above 30%, the a-melt-
mediated transformation caused the crystallization of only the g form, and b0 or
b of SOS did not appear. The structural model of solid-solution phases of a and
g forms and the eutectic phase of the g of SLS and b0 of SOS are shown in
Figure 23 (92).
Figure 23. Structure models of the polymorphic forms of the SOS/SLS mixtures.
The a form of the SLS/SOS mixture is stacked in the double-chain-length struc-

ture, in which the stearoyl and unsaturated (oleoyl and linoleoyl) chains are packed
in the same leaflets (Figure 23a). Further thermodynamic equilibration induced
the transformation from the double-chain-length a form to the triple-chain-length
g form. In this transformation, chain segregation occurs during the a-g transforma-
tion in the solid state, because the steric hindrance between the stearoyl and lino-
leoyl chains is limited. In the g form, the stearoyl and unsaturated chains are packed
in the different leaflets (Figure 23b). Because of olefinic interactions between
oleoyl and linoleoyl chains, coexistence between oleoyl and linoleoyl chains in
the oleic/linoleic acid leaflet takes place in the g form having the triple-chain-length
structure. This interaction makes the g form the most stable polymorph of the SLS/
SOS mixture, and the transformation form g to b0 or b2 forms does not occur in
the mixture during simple cooling and heating processes. Disappearance of the
miscible g form of the mixture can be achieved by melt-mediated transformation
from the miscible a and g forms to immiscible b0 and b forms in the SOS fraction
(Figure 23c).
6. POLYMORPHISM OF NATURAL FATS
Most natural fats are composed of many different kinds of acylglycerols whose
fatty acid compositions are diverse, and the acyl positions esterified at the glycerol
groups are complicated. This situation makes the polymorphsm of natural fats very
complicated.
Take for example, milkfat that consists of TAGs, diacylglycerols (DAGs), mono-
acylglycerols (MAGs), free fatty acids, phospholipids, sterols, and other polar lipids
(9597). As for TAGs, milkfat is made of about 400 different TAGs containing var-
ious kinds of saturated and unsaturated fatty acids whose carbon numbers range
from 2 to 24. Because of this fact, milkfat has a wide range of melting temperature
from about 30 C to 40 C, and three polymorpic forms of a, b0 , and b reveal com-
plicated chain-length structures and occurrence behavior that are affected by ther-
mal treatment. Another example is cocoa butter (CB) in which stearic, plamitic, and
oleic acids account for about 80 % out of its total fatty acids. This property causes
sharp melting behavior of CB. However, polymorphism of cocoa butter is compli-
cated, and its origins are still unanswered.
POLYMORPHISM OF NATURAL FATS 109
It is difficult to simply define the polymorphism of natural fats composed of mul-

tiple TAGs because of the following two reasons:
1. The polymorphic nature of the multicomponent TAG systems is related

to phase behavior that is affected by molecular interactions among the
component TAGs. The fat crystals in a miscible phase may exhibit simple
polymorphic properties. By contrast, the immiscile eutectic phase may show
complicated polymorphic properties as a superposition of the polymorphic
forms of the component TAGs. Furthermore, if the molecular compound is
formed by specific TAG components, the polymorphic behavior becomes
complicated, as shown for the case of POP-OPO (see Section 5.2). Therefore,
knowing the phase behavior of the principal TAG components is a prerequi-
site for precise understanding of the polymorphism of natural fats.
2. The phase behavior of the mixed TAG system is influenced by polymorphism.
For example, a miscible phase is formed in a and b0 polymorphs, but it
transforms into a eutectic phase in b, as revealed in the SSS-PPP mixture.
Then, the polymorphic occurrence is largely affected by cooling rate and
temperature fluctuation, and it is therefore necessary to observe the poly-
morphic properties of the natural fats by varying the rate of cooling or by
fluctuating the temperature (so-called tempering).
In this section, the polymorphic properties of natural fats are briefly discussed by
highlighting milkfat, cocoa butter, and palm oil fraction based on recent research
into the effects of external factors on the polymorphic crystallization such as shear
stress, ultrasound stimulation, and addition of food emulsifiers.
6.1. Milkfat
Crystallization of milkfat is an important process for fractionation of its contents
and production of butter, whipped cream, and ice cream. As the quality of these
products strongly depends on polymorphism of milkfat, physical chemical proper-
ties of milkfat have been studied by many researchers (98).
As mentioned above, milkfat is characterized as a very complicated mixture of
TAGs, and thereby it is almost impossible to clarify how every TAG component
crystallizes in a cooperative way with the other TAG components. Instead, milkfat
is fractionated in accordance with different melting ranges to obtain three major
fractions: high melting fraction (HMF), medium melting fraction (MMF), and
low melting fraction (LMF). Marangoni and Lencki concluded that HMF
and MMF are fully miscible in the solid state, and mixtures of LMF with HMF
and MMF showed monotectic property with nature of partial solid solution (70).
As for the polymorphism of milkfat, a and b0 forms frequently appear, and b
form appears under special conditions when HMF and milkfat are stored for long
duration (99101). In regard to the effects of thermal treatment and emulsification
on the polymorphic crystallization of milkfat, Lopez et al. recently performed syn-
chrotron radiation X-ray diffraction and DSC studies, using anhydrous milkfat
TABLE 9. Comparison of Thermal and Structural Behaviors of

Anhydrous Milkfat and Cream Observed at a Slow Cooling Rate.
Anhydrous Milkfat Milkfat Globle Cream

0
Polymorphic occurrence b, then b and a a, then a b0
Chain-length structure 4.15 (double): vs* 4.65 (double): m
(unit, nm) 4.83 (double): w 4.0 (double): m
6.22 (triple): s 7.13 (triple):s
3.92 (double): m 6.5 (tripl): w

vs: very storng, s: strong, m: medium, w: weak.
(AMF) as a bulk fat system and milkfat globule of cream as an emulsion system
(102106).
No difference in the polymorphic occurrence was observed between cream and
AMF when the samples were rapidly cooled from 50 C to 8 C: a form first
appeared and b0 form appeared during subsequent heating after the melting of
a form. On the other hand, crystallization by slow cooling (<0.15 C/min) caused
remarkable differences between the emulsion and bulk systems. As summarized in
Table 9, b0 form first crystallized and b0 and a forms coexisted until the end of cool-
ing in the bulk AMF. By contrast, a form first crystallized and b0 form started to
crystallize during further cooling. In the heating process after the crystallization,
a first melted and then b0 form melted in both samples.
The chain-length structures largely differed between the bulk AMF and cream
samples. Table 9 also shows four different crystals formed in the AMF bulk sam-
ples, and four crystals in cream. Quite interestingly, the lamellar spacing values are
all different from each other, and double-chain-length and triple-chain-length struc-
tures are coexisted. The occurrence domains of the four crystals in AMF during
the slow cooling are shown in Figure 24 that shows relative intensity of small-angle
X-ray diffraction peaks of the four crystals observed in the bulk AFM and DSC
thermopeaks taken during the slow cooling. It is clearly shown that a large exother-
mic peak around 22 C, a large exothermic peak around 13 C, and a small peak at
4 C are caused by the crystallization of b0 form, a form, and b0 form, respectively.
It is assumed that the crystallization behavior of milkfat is different between
emulsion and bulk, and the lack of nucleation centers in the emulsion droplets
may delay the nucleation, making less stable a form nucleated in the first. The
occurrence of multiple forms of double-chain-length and triple-chain-length struc-
tures may be caused by segregated crystallization of multicomponent TAGs exhi-
biting complicated mixing behavior, but its details are open to future study.
6.2. Cocoa Butter

Cocoa butter is the most popular fat used for confectionery. CB consists of three
major TAGs, POP, POS, and SOS, and other minor components (107, 108). The
three TAGs determine the polymorphic nature of CB that exhibits six polymorphs,
Form I through Form VI in accordance with the nomenclature given by Wille and
Figure 24. Relative intensity of small-angle X-ray diffraction peaks and DCS thermopeaks of
bulk anhydrous milkfat taken during a slow cooling process.
Lutton (109). This section employs this nomenclature, although the other nomen-
clatures such as b0III and bV, are used in other researchers (110). As Form V func-
tionally works for chocolate, crystallization of CB in Form V, and preservation of
this polymorph during long storage are the prerequisites for quality control of the
end products. For this purpose, a tempering method including cooling from a mol-
ten state, reheating, and recooling has widely been applied (107, 108). The other
technique is to use seed crystals of BOB b2 whose polymorphic structure is
identical to that of Form V of CB and whose melting point is higher than Form
V of CB (107). The BOB b2 seed crystals can be put in molten chocolate during
a simple cooling process without tempering to obtain Form V of CB.
Recently, interesting work has been done to examine the effects of shear stress
(19, 110) and ultrasound irradiation on the polymorphic crysallization of CB
(111).
100
Form V
Intensity (arb. unit)
Form IV
Form III
50
Form III
0
0 30 60
Time (min)
Figure 25. Relative intensity of small-angle X-ray diffraction peaks of cocoa butter without shear
stress (closed) and shear stress (open).
As for the effects of the shear stress, it was shown by a Synchrotron radiation
X-ray diffraction study that transformations from metastable to more stable forms,
especially to Form V, were accelerated by high shear stress (110). Figure 25 shows
the time variation of relative intensities of X-ray diffraction peaks of CB crystals
formed after cooling from 50 C to 18 C at a rate of 3 C/min. In the case of no
shear, Form III appeared at first after the temperature reached at 18 C, and then
Form IV crystallized at the expense of Form III. On the other hand, applying the
shear stress at 1440 s1 caused accelerated transformation from Form III to Form V,
without the occurrence of Form IV. The same result was observed with lower shear
rates (19), and the persistence time of Form III was reduced as the shear rate was
increased. Mazaanti et al also observed that the orientation of CB crystals are
aligned with the shear flow (110). These results indicated that temperature and
shear treatments are the tools for tailoring the desired polymorphic structures
of fats.
It was observed that ultrasound stimulation (ultrasonication) also accelerated the
crystallization of the more stable polymorphs of CB (111). A fundamental study of
the effects of ultrasonication on the polymorphic crystallization of PPP and LLL
showed that several factors, such as pressure effect, shear flow, cavitation, and ther-
mal energy caused by absorption of attenuated ultrasound wave, may play concur-
rent effects of ultrasonication. As a result, there are optimal conditions for
temperature and duration of ultrasonication to increase the rate of crystallization
and the occurrence of the more stable polymorphs (20). This effect was also
observed in CB (111).
Figure 26 shows wide-angle X-ray diffraction profiles of CB with ultrasonication
of three durations and without ultrasonication taken after cooling at 20 C from
15sec
9 sec Form II
3 sec Form V+
Form II
10
0 sec Form V
15
Diff
rac 20
tio na
ngl
e (2 25
: d Form II
eg)
Figure 26. Wide-angle X-ray diffraction patterns of cocoa butter with ultrasonication.
60 C. Ultrasound (200 kHz, 300 W) was stimulated to a 250-mL sample of CB at

32.3 C during cooling before crystallization. Form II occurred without ultrasonica-
tion, whereas Form V was observed when ultrasonication was done for 3 seconds.
Further ultrasonication for 9 seconds formed a mixture of Form II and Form V, and
only Form II was observed by the ultrasonication for 15 seconds. It is assumed that
there are conflicting effects by ultrasonication: promotion of nucleation by pressure
effect and retardation of nucleation by thermal energy caused by absorption of atte-
nuated ultrasound wave. The former effect may prevail at the ultrasonication for
3 seconds. The temperature rise, however, of the sample caused by absorption of atte-
nuated ultrasound wave was 2 C for 9 seconds and 3.9 C for 15 seconds, and the
latter effect may result in the case of cooling from above the melting point of CB.
6.3. Palm Oil

Palm oil is a common fat and oil resource for many industrial uses. For example, of
food applications, it is used as cooking oil, margarine, shortening, and in confec-
tionery products (73, 112). Palm oil has several advantageous properties such as
high productivity and high thermal and oxidative stability and plasticity at room
temperature. However, the crystallization properties of palm oil are disadvanta-
geous because of a low rate of nucleation and crystal growth of granular crystals
(113114). The granular crystals are easily formed during long storage, causing
sandy taste and inhomogeneity of fat crystal morphology of the end products
(8, 115117). Although many TAGs, DAGs, free fatty acids, and so on, are involved
in it, palm oil exhibits two polymorphic forms, a and b0 under normal cooling rate,
and b form also appears at a very slow crystallization rate. The addition of food
emulsifiers into palm oil has been an efficient external factor to modify the poly-
morphism crystallization of palm oil.
Polyglycerol fatty acid esters are biograded surfactants that are used widely in
industries such as food, cosmetics, toiletries, and pharmaceuticals (118). Advanta-
geous properties of the PGFEs are derived from the easy modification of their
hydrophobicity and hydrophilicity by changing the degree of polymerization of
glycerol and esterification with fatty acid moieties and by modifying the chemical
structures of fatty acid moieties.
It was observed that the addition of polyglycerine fatty acid ester to palm oil
affected the polymorphic crystallization and morphological properties of palm oil
(119). In particular, polyglycerol behenic acid ester showed a remarkable effect.
Optical microscopy observation confirmed that palm oil crystals with the addition of
1 wt.% of polyglycerol behenic acid esters were smaller and the number of palm
oil crystals larger than without the additives, as shown in Figure 27. This indicated
that the polyglycerol behenic acid ester promoted nucleation and inhibited crystal
growth of palm oil. X-ray diffraction patterns of palm oil without the additives
revealed that palm oil crystallized in the a form after rapid quenching of melted
palm oil at 10 C. During the heating process from 10 C to 45 C, the a form trans-
formed to the b0 form around 15 C, and the b0 form changed to the b form around
40 C. The X-ray diffraction patterns of palm oil with the addition of polyglycerol
Figure 27. Optical micrographs of isothermal crystallization of palm oil with and without an
additive (1 wt.% of polyglycerine behenic acid ester). (A) 60 min at 20 C without the additive; (B)
60 min at 20 C with the additive; (C) 60 min at 27 C without additive; (D) 60 min at 27 C with the
additive. (This figure is available in full color at http://www.mrw.interscience.wiley.com/biofp.)
SUMMARY 115
behenic acid esters showed that palm oil crystallized in the b0 form at 10 C, and
it did not transform to the b form during the heating process.
It is generally considered that the nucleation rates of a and b0 are largely differ-
ent, with a crystallizing much more rapidly than b0 and b. The size of the crystal is
associated with the crystal form; b form tends to produce granular crystals, and b0 is
recommended for producing small crystals. Figure 27 shows palm oil with polygly-
cerol behenic acid esters added, the b0 form is preferentially crystallized, and the
crystal does not grow to a granular crystal. One reason for this may be that poly-
glycerol behenic acid esters, which promote the nucleation of palm oil, have higher
melting points than that of palm oil; namely, the emulsifier as an additive may
crystallize faster than palm oil, inducing heterogeneous nucleation of palm oil as
template.
7. SUMMARY
This chapter described polymorphic properties of principal TAGs and natural fats
based on recent research work to clarify fundamental aspects of polymorphsim of
fats and oils. The authors hope that the basic understanding of the polymorphism of
the principal TAGs would be useful to elucidate rather complicated polymorphic
properties of natural fats and oils that contain TAGs with very heterogeneous fatty
acid compositions.
It may be worth noting the following subjects, which are nowadays still open to
question and therefore should be worked out in future.
1. Polymorphism of diacid and triacid TAGs

Little knowledge of polymorphism of diacid and triacid TAGs have been
obtained, because these TAGs have high relevance to natural fats and oils. The
main reasons for this must be difficulty in preparing pure materials, which may pro-
vide convincing results. It is highly interesting and important to work on diacid and
triacid TAGs with an emphasis on the effects of chirality on polymorphism that has
rarely been known.
2. Polymorpic structures of saturated-unsaturated mixed-acid TAGs
Although much effort has been done to unveil precise structures of saturated-
unsaturated mixed-acid TAGs, no atomic-level structure analysis has been success-
ful because of difficulty to obtain high-quality single crystals suitable for X-ray
analysis. In particular, SOS and POP are the most representative TAGs that are
major components of cocoa butter and palm oil. It is expected that structures
data of the stable b forms of SOS and POP may give key ideas to resolve fat bloom
phenomena in confectionery fats. Molecular simulations have been done (120,
121), but the experimental results so far obtained are the XRD patterns and FT-
IR spectra using powder and single crystals, which do not provide detailed mole-
cular structures, in particular, about olefinic conformation and interfacial structures
of oleic acid and saturated acid leaflets.
3. Phase behavior of TAG mixtures

As the natural fats and oils are mixtures of different TAGs, their polymorphism
is influenced by the phase behavior of the mixture phases. In order to mimic natural
fats and oils, the mixture phases of diacid TAGs were studied, as reviewed in Sec-
tion 4. However, we more need basic research of the mixtures of TAGs whose
fatty acid compositions are heterogeneous, e.g., the mixtures of between saturated
diacid TAGs of saturated diacid TAGs and unsaturated diacid TAGs. These studies
may be significant in understanding the polymorphism of palm oil and its fractions
(palm stearin, palm olein, etc.), or milkfat and its fractions (high-melting fraction,
etc.).
REFERENCES
1. C. E. Clarkson, and T. Malkin, J. Chem. Soc., 666 (1934).

2. K. Larsson, Acta Chem. Scand., 20, 2255 (1966).
3. H. D. Goff, in F. D. Gunstone, and P. B. Padley, eds., Lipid Technologies and
Applications, Marcel Dekker, New York, 1997, pp. 355368.
4. J. M. de Man, in M. N. Widlak, ed., Physical Properties of Fats, Oils, and Emulsifers,
AOCS Press, Champaign, Illinois, 1999, pp. 7995.
5. R. E. Wainwright, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products 5th ed.,
Vol. 3, Wiley, New York, 1996, pp. 353408.
6. E. S. Lutton, J. Amer. Oil Chem. Soc., 27, 276 (1950).
7. R. E. Timms, Chem. Phys. Lipids, 21, 793, (1978).
8. K. Sato, in P. B. Padley, ed., Advances in Lipid Research, JAI Press, New York, 1996,
pp. 213268.
9. D. Aquilano and G. Spualdino, in N. Garti and K. Sato, eds., Crystallization Processes in
10. K. Sato and T. Kuroda, J. Amer. Oil Chem. Soc., 64, 124 (1987).
11. T. Koyano, I. Hachiya, T. Arishima, K. Sato, and N. Sagi, J. Amer. Oil Chem. Soc., 66,
675 (1989).
12. M. Kellens, W. Meeussen, and C. H. Riekel, Chem. Phys. Lipids, 52, 79 (1990).
13. M. Kellens, W. Meeussen, R. Gehrke, and H. Reynaers, Chem. Phys. Lipids, 58, 131
(1991).
14. M. Kellens, W. Meeussen, A. Hammersley, and H. Reynaers, Chem. Phys. Lipids, 58,
145 (1991).
15. C. W. Hoerr and F. R. Paulicka, J. Amer. Oil Chem. Soc., 45, 793 (1968).
16. D. M. Small, J. Lipid Res., 25, 1490 (1984).
17. D. M. Small, The Physical Chemistry of Lipids, Plenum Press, New York, 1986, pp. 475
522.
18. J. W. Mullin, Crystallization, 4th ed., Butterworth-Heineman, Oxford, 2001, pp. 135
180.
19. S. D. MacMillan, K. J. Roberts, A. Rossi, M. A. Wells, M. C. Polgreen, and I. H. Smith,
Cryst. Growth Design, 2, 221 (2002).
REFERENCES 117
20. S. Ueno, R. Ristic, K. Higaki, and K.Sato, J. Phys. Chem. B, 107, 4927 (2003).
21. D. J. Cebula, D. J. McClements, and M. J. W. Povey, J. Amer. Oil Chem. Soc., 67, 76
(1990).
22. D. J. Cebula, D. J. McClements, M. J. W. Povey, and P. Smith, J. Amer. Oil Chem. Soc.,
69, 130 (1992).
23. F. Kaneko, J. Yano, and K. Sato, Curr. Opin. Struct. Biol., 8, 417 (1998).
24. F. Kaneko, in N. Garti, and K. Sato, eds., Crystallization Processes in Fats and Lipid
25. D. Chapman, J. Chem. Soc., 4489 (1957).
26. J. Yano, F. Kaneko, M. Kobayashi, and K. Sato, J. Phys. Chem. B, 101, 8112 (1997).
27. J. Yano, F. Kaneko, M. Kobayashi, D. R. Kodali, D. M. Small, and K. Sato, J. Phys.
Chem. B, 101, 8120 (1997).
28. J. Yano, K. Sato, F. Kaneko, D. M. Small, and D. R. Kodali, J. Lipid Res., 40, 140 (1999).
29. J. Schaefer, E. O. Stejskal, and R. Buchdahl, Macromolecules, 8, 291 (1975).
30. S. M. Bociek, S. Ablett, and I. T. Norton, J. Amer. Oil Chem. Soc., 62, 1261 (1986).
31. H. Saito, Magn. Reson. Chem., 24, 835 (1986).
32. T. Eads, A. E. Blaurock, R. G. Bryant, D. J. Roy, and W. R. Croasman, J. Amer. Oil
Chem. Soc., 69 1057 (1992).
33. T. Arishima, K. Sugimoto, R. Kiwata, H. Mori, and K. Sato, J. Amer. Oil Chem. Soc., 73,
1231 (1996).
34. L. H. Jensen and A. J. Mabis, Nature, 197, 681 (1963).
35. L. H. Jensen and A. J. Mabis, Acta Crystallogr., 21, 770 (1966).
36. K. Larsson, Arkiv. Kemi, 23, 1 (1965).
37. E. S. Lutton, J. Amer. Oil Chem. Soc., 48, 246 (1971).
38. A. J. van Langevelde, K. van Malssen, F. F. A. Hollander, R. Peschar, and H. Schenk,
Acta Crystallogr., B55, 114 (1999).
39. A. J. van Langevelde, K. van Malssen, E. Sonneveld, R. Peschar, and H. Schenk, J. Amer.
Oil Chem. Soc., 76, 603 (1999).
40. In ref. 17, pp. 368369.
41. A. J. van Langevelde, K. van Malssen, R. Driessen, K. Goubits, F. Hololander,
R. Pescher, P. Zwart, and H. Schenk, Acta Crystallogr. B, 56, 1103 (2000).
42. K. Sato, M. Goto, K. Honda, J. Yano, D. R. Kodali, and D. M. Small, J. Lipid Res., 42,
338 (2001).
43. D. H. Wheeler, R. W. Riemenschceider, and C. E. Sandro, J. Biol. Chem., 132, 687
(1940).
44. R. H. Ferguson and E. S. Lutton, J. Amer. Chem. Soc., 69, 1445 (1947).
45. J. W. Hagemann, W. H. Tallent, and K. E. Kolb, J. Amer. Oil. Chem. Soc., 49, 118 (1972).
46. M. G. R. Carter, and T. Malkin, J. Chem. Soc., 554 (1947).
47. A. Desmedt, C. Culot, C. Deroanne, F. Durant, and V. Gibon, J. Amer. Oil. Chem. Soc.,
67, 653 (1990).
48. D. R. Kodali, D. Atkinson, T. G. Redgrave, and D. M. Small, J. Amer. Oil. Chem. Soc.,
61, 1078 (1984).
49. D. R. Kodali, J. Lipid Res., 25, 1490 (1987).
50. D. R. Kodali, D. Atkinson, and D. M. Small, J. Phys. Chem., 93, 4683 (1989).
51. D. R. Kodali, D. Atkinson, and D. M. Small, J. Dispersion Sci. Technol., 10, 393 (1989).
52. D. R. Kodali, D. Atkinson, and D. M. Small, J. Lipid Res., 31, 1853 (1990).
53. M. Goto, D. R. Kodali, D. M. Small, K. Honda, K. Kozawa, and T. Uchida, Proc. Natl.
Acad. Sci. USA, 89, 8083 (1992).
54. K. Sato and S. Ueno, in N. Garti and K. Sato, eds., Crystallization Process in Fats and
Lipid Systems, Marcel Dekker, New York, 2001, pp. 194209.
55. A. J. van de Streek, P. Verwer, R. de Gelder, and F. Hollander, J. Amer. Oil. Chem. Soc.,
76, 1333 (1999).
56. H. M. Zacharis, D. Mitcham, N. Lovegren, and M, Gray, Adv. X-ray Analysis, 18, 535
(1975).
57. Z. H. Wang, K. Sato, N. Sagi, T. Izumi, and H. Mori, J. Jpn. Oil. Chem. Soc. (Yukagaku),
36, 671 (1987).
58. K. Sato, T. Arishima, Z. H. Wang, K. Ojima, N. Sagi, and H. Mori, J. Amer. Oil Chem.
Soc., 66, 664 (1989).
59. T. Arishima and K. Sato, J. Amer. Oil Chem. Soc., 66, 1614 (1989).
60. T. Arishima, N. Sagi, H. Mori, and K. Sato, J. Amer. Oil Chem. Soc., 68, 710 (1991).
61. J. Yano, S. Ueno, K. Sato, T. Arishima, N. Sagi, F. Kaneko, and M. Kobayashi, J. Phys.
Chem., 97, 12967 (1993).
62. S. Ueno, A. Minato, H. Seto, Y. Amemiya, and K. Sato, J. Phys. Chem. B, 101, 3498
(1997).
63. K. Sato, S. Ueno, and J. Yano, Prog. Lipid Res., 38, 91 (1999).
64. K. Boubekri, J. Yano, S.Ueno, and K. Sato, J. Amer. Oil Chem. Soc., 76, 949 (1999).
65. M. Takeuchi, S. Ueno, J. Yano, E. Floter, and K. Sato, J. Amer. Oil Chem. Soc., 77, 1243
(2000).
66. D. R. Kodali, D. Atkinson, T. G. Redgrave, and D. M. Small, J. Lipid Res., 28, 403
(1987).
67. P. Elisabettini, A. Desmedt, V. Gibon, and F. Durant, Fat Sci. Technol., 97, 65 (1995).
68. P. Elisabettini, G. Lognay, A. Desmedt, C. Culot, N. Istasse, E. Deffence, and F. Durant,
J. Amer. Oil Chem. Soc., 75, 285 (1998).
69. A. G. Marangoni, and R. W. Hartel, Food Technol., 52, 46 (1998).
70. A. G. Marangoni and R. W. Lencki, J. Agric. Food Chem., 46, 3879 (1998).
71. F. B. Padley, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and
72. A. G. Marangoni and D. Rousseau, J. Amer. Oil Chem. Soc., 75, 1265 (1998).
73. R. E. Timms, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and
74. K. Sato, Chem. Eng. Sci., 56, 2255 (2001).
75. R. E. Timms, Prog. Lipid Res., 23, 1 (1984).
76. V. Givon, A. Durant, and C. Deroanne, J. Amer. Oil Chem. Soc., 63, 1047 (1986).
77. L. Engstrom, J. Fat Sci. Technol., 94, 173 (1992).
78. M. Kellens, W. Meeussen, A. Hammersley, and H. Reynaers, Chem. Phys. Lipids, 58,
145 (1991).
REFERENCES 119
79. T. Koyano, I. Hachiya, and K. Sato, J. Phys. Chem., 96, 10514 (1992).
80. A. Minato, S. Ueno, J. Yano, Z. H. Wang, H. Seto, Y. Amemiya, and K. Sato, J. Amer. Oil
Chem. Soc., 73, 1567 (1996).
81. A. Minato, S. Ueno, K. Smith, Y. Amemiya, and K. Sato, K. J. Phys. Chem. B, 101, 3498
(1997).
82. A. Minato, S. Ueno, J. Yano, K. Smith, H. Seto, Y. Amemiya, and K. Sato, J. Amer. Oil
Chem. Soc., 74, 1213 (1997).
83. A. Minato, J. Yano, S. Ueno, K. Smith, and K. Sato, Chem. Phys. Lipids, 88, 63 (1997).
84. M. Takeuchi, S. Ueno, E. Floter, and K. Sato, K. J. Amer. Oil Chem. Soc., 79, 627 (2002).
85. J. B. Rossell, Adv. Lipid Res., 5, 353 (1967).
86. M. Takeuchi, S. Ueno, and K. Sato, Cryst. Growth Design, 3, 369 (2003).
87. D. P. Moran, J. Appl. Chem., 13, 91 (1963).
88. L. Di and D. M. Small, J. Lipid Res., 34, 1611 (1993).
89. L. Di and D. M. Small, Biochemistry, 34, 16672 (1995).
90. M. Goto, K. Honda, L. Di, and D. M. Small, J. Lipid Res., 36, 2185 (1995).
91. J. Yano, and K. Sato, Food Res. Int., 32, 249 (1999).
92. M. Takeuchi, S. Ueno, and K. Sato, Food Res. Int., 35, 919 (2002).
93. K. Larrson, Fette. Seifen. Anstrichm., 74, 136 (1972).
94. K. Larsson, in Lipids-Molecular Organization, Physical Functions and Technical
Applications, The Oily Press, Dundee, U.K., 1994, pp 746.
95. R. G. Jensen, A. M. Ferris, and C. J. Lammi-Keefe, J. Dairy Sci., 74, 3228 (1991).
96. W. W. Christie, in P. F. Fox, ed., Advanced Dairy Chemistry 2: Lipids, 2nd ed.,
Chapmann and Hall, London, 1995, pp. 136.
97. P. Walstra, T. J. geurts, A. Noomen, A. Jellema, and A. J. S. van Boekel, Dairy
Technology, Marcel Dekker, New York, 1999, pp. 3105.
98. R. W. Hartel and K. E. Kaylegian, in N. Garti, and K. Sato, eds., Crystallization
Processes in Fats and Lipid Systems, Marcel Dekker, New York, 2001, pp. 381427.
99. P. Walstra, in J. M. V. Blanshard and P. Lillford, eds., Food Structure and Behaviour,
Academic Press, New York, 1987, pp. 6786.
100. I. L. Woodrow and J. M. de Man, J. Dairy Sci., 51, 996 (1968).
101. J. E. Schaap, H. T. Badings, D. G. Schmidt, and E. Frede, Neth. Milk Dairy J., 29, 242
(1975).
102. C. Lopez, P. Lesieur, G. Keller, and M. Ollivon, J. Colloid Int. Sci., 229, 62 (2000).
103. C. Lopez, F. Lavigne, P. Lesieur, C. Bourgaux, and M. Ollivon, J. Dairy Sci., 84, 756
(2001).
104. C. Lopez, P. Lesieur, C. Bourgaux, G. Keller, and M. Ollivon, J. Colloid Int. Sci., 240,
150 (2001).
105. C. Lopez, F. Lavigne, P. Lesieur, G. Keller, and M. Ollivon, J. Dairy Sci., 84, 2402
(2001).
106. C. Lopez, C. Bourgaux, P. Lesieur, and M. Ollivon, Lait, 82, 317 (2002).
107. K. Sato and T. Koyano, in N. Garti and K. Sato, eds., Crystallization Processes in Fats
and Lipid Systems, Marcel Dekker, New York, 2001, pp. 429456.
108. R. E. Timms, Confectionery Fats Handbook, The Oily Press, Dundee, U.K., 2003.
109. R. L. Wille and E. S. Lutton, J. Amer. Oil Chem. Soc., 27, 491 (1996).
110. G. Mazzanti, S. E. Guthrie, E. B. Sirota, A. G. Marangoni, and S. H. J. Idziak, Crystal
Growth & Design, 3, 721 (2003).
111. K. Higaki, S. Ueno, T. Koyano, and K. Sato, J. Amer. Oil Chem. Soc., 78, 513 (2001).
112. K. Smith, in N. Garti and K. Sato, eds., Crystallization Processes in Fats and Lipid
113. H. Ishikawa, T. Mizuguchi, and S. Kondo, J. Jpn Oil Chem. Soc., 29, 235 (1980).
114. A. Watanabe, I. Tashima, N. Matsuzaki, J. Kurashige, and K. Sato, J. Amer. Oil Chem.
Soc., 69, 1077 (1992).
115. W. L. Ng, J. Amer. Oil Chem. Soc., 67, 879 (1990).
116. K. Kawamura, J. Amer. Oil Chem. Soc., 56, 753 (1979).
117. K. Kawamura, J. Amer. Oil Chem. Soc., 57, 48 (1980).
118. V. K. Babayan, T. G. Kaufman, H. Lehman, and R. J. Tkaczuk, J. Soc. Cos. Chem., 15,
473 (1964).
119. M. Sakamoto, K. Maruo, J. Kuriyama, M. Kouno, and K. Sato, J. Oleo Sci., 52, 639
(2003).
120. S. de Jong, T. C. van Soest, and M. A. van Schaick, J. Amer. Oil Chem. Soc., 68, 371
(1991).
121. J. W. Hagemann and J. A. Rothfus, J. Amer. Oil Chem. Soc., 69, 429 (1992).
4
Fat Crystal Networks
Geoffrey G. Rye, Jerrold W. Litwinenko, and
Alejandro G. Marangoni
University of Guelph
Guelph, Ontario, Canada
1. INTRODUCTION
Fats provide fundamental structural and textural attributes to a wide range of con-
sumer products, including lipstick, chocolate, and everyday products such as butter
and margarine (1, 2). Within these fat-based products, certain textural properties
are required to meet desirable sensory attributes to gain consumer acceptance (3).
This has led to an increase in research efforts on the physical properties of fats,
particularly their rheology.
The goal of this chapter is to investigate the effects of processing conditions
on the physical properties of fats, using anhydrous milkfat (AMF) as an example.
The approach proposed may lead to an increased understanding of product
quality and characteristics, while offering insight into future methods for the
determination and prediction of the rheological and textural attributes of fat-based
products.
121
122 FAT CRYSTAL NETWORKS
2. MECHANICAL PROPERTIES OF MILKFAT
Most mechanical tests developed for fats are empirical in nature and are usually
designed for quality control purposes, and they attempt to simulate consumer sen-
sory perception (3, 4). These large-deformation tests measure hardness-related
parameters, which are then compared with textural attributes evaluated by a sensory
panel (3, 5). These tests include penetrometry using cone, pin, cylinder and several
other geometries (3, 612), compression (13), extrusion (13, 14), spreadability (15,
16), texture profile analysis (2), shear tests (13), and sectility measurements (14).
These methods are usually simple and rapid, and they require relatively inexpensive
equipment (3, 4, 17). The majority of these tests are based on the breakdown of
structure and usually yield single-parameter measurements such as hardness,
yield stress, and spreadability, among others (4, 1720). The relationship
between these mechanical tests and the structure of a fat has, however, not been
established. The ultimate aim of any materials science endeavor is to examine
the relationship between structure and macroscopic properties.
As a starting point, we need a model that can be used to describe the formation
and interaction among structural elements that affect the macroscopic properties of
a fat crystal network. One such model is shown in Figure 1. This model suggests
that lipid composition, directly under the influence of the processing and storage
conditions, will affect the solid fat content, polymorphism, and microstructure of
a fat crystal network. The model also suggests that interactions among these
Macroscopic
Legend Properties
Strong Influence
Potential Influence
Nucleation &
Crystal Growth
Microscopic
Properties
Storage Time
Processing
Storage
Conditions Conditions
Solid Fat
Polymorphism
Content
Molecular/Lipid
Composition
Figure 1. Hierarchical model of the factors affecting the macroscopic properties of a fat crystal
network.
LIPID COMPOSITION 123
factors also play an important role, because they all ultimately affect the structure
of the fat crystal network, and in turn influence the physical properties and
sensory perception of a fat. Each of the factors within the model will be examined
in turn.
3. LIPID COMPOSITION
3.1. Triacylglycerols
Fats are composed primarily of triacylglycerols (TAG). A TAG consists of three
fatty acid residues esterified to a glycerol (three-carbon sugar alcohol) backbone
at specific locations known as sn-1, sn-2, and sn-3. Fatty acid residues exist in a
wide variety of forms, including short and long chain, saturated and unsaturated,
odd or even carbon number, trans- or cis-, linear or branched, as well as any com-
bination thereof (21). TAG variety in a fat system is very large, as there are poten-
tially hundreds of different fatty acid residues and their isomers available for
reaction at any of the three sn-positions.
3.1.1. Anhydrous Milkfat (AMF) Composition and Properties AMF is het-

erogeneous in nature and is the most complex naturally derived fat (2224).
Previous research has revealed greater than 400 different fatty acids in AMF
with carbon chain lengths ranging from 4 to 24 with varying degrees of saturation
and molecular arrangement (24, 25). In bovine AMF, fatty acids are primarily in
the form of TAGs. TAGs make up 9698% of the total fat and are the primary
component in the formation and structure of the fat crystal network (23, 25).
The remaining 24% are minor components, which include monoacylgly-
cerols, diacylglycerols, phospholipids, free fatty acids, cholesterol, and some
protein (26).
Milkfat complexity dramatically increases when one considers the number of
combinations for 400 fatty acid species esterified to glycerol at three different bond-
ing sites4003 (64 million) possible TAG structures are theoretically possible (24,
25). In AMF, however, there are typically only 13 fatty acids present that are in
concentrations greater than 1% (w/w), and therefore, a theoretical maximum of
133 (2197) TAG isomers typically exist (25). However, if locations sn-1 and sn-3
are considered equivalent, the number of possible TAGs is reduced to 455 (25).
For the purpose of investigating the effects of lipid composition on the physical
properties of milkfat, we will use the fatty acid concentrations determined using
gas-liquid chromatography. Table 1 shows the fatty acid compositions of milkfat
and milkfat diluted using canola oil. Dilution with canola oil allows for the con-
trolled variation in solids content. By diluting with canola oil, AMF composition
is altered through the addition of the long-chain fatty acids, primarily 18:1, 18:2,
and 18:3, with only small modifications to the saturated fatty acid concentrations
(16:0, 18:0, and 20:0). This allows for the study of the effects of lipid composition
on physical properties by altering the TAG makeup using a diluent that neither
TABLE 1. Fatty Acid Composition (wt%) of Anhydrous Milkfat and Dilutions of

Anhydrous Milkfat in Canola Oil.
Fatty Acid 100% AMF 90% AMF 80% AMF 70% AMF
4:0 2.76 2.48 2.21 1.93

6:0 2.18 1.96 1.74 1.53
8:0 1.39 1.25 1.11 0.97
10:0 3.16 2.84 3.01 2.63
12:0 3.76 3.38 2.53 2.21
14:0 12.34 11.15 9.96 8.76
14:1 1.92 1.73 1.54 1.34
15:0 1.42 1.28 1.14 0.99
15:1 0.35 0.32 0.28 0.25
16:0 30.37 27.92 25.46 23.01
16:1 2.93 2.69 2.44 2.20
17:0 0.91 0.82 0.73 0.64
17:1 0.48 0.43 0.38 0.34
18:0 9.48 8.77 8.05 7.34
18:1 21.92 25.69 29.46 33.23
18:2 2.79 4.52 6.24 7.97
18:3 1.31 2.11 2.92 3.72
20:0 0.52 0.61 0.71 0.80
22:0 0.00 0.05 0.09 0.14
significantly solubilizes milkfat TAGs nor crystallizes in the temperature range of

interest (usually above 0 C) (27).
AMF complexity gives it unique textural, physical, and sensory properties. The
crystallization and melting properties of AMF are directly related to its complex
composition. The large variety of TAG species that exist in AMF cause it to
have a wide melting range, spanning from 40 C to 40 C. Within this melting
range, at least three distinct fractions have been identified and designated as low-
melting (LMF), medium-melting (MMF), and high-melting (HMF) fractions (28).
These fractions are not well defined; however, the LMF typically contains short-
chain (4:0 to 8:0) and long-chain unsaturated (18:1, 18:2, 18:3, 20:1) fatty acids.
MMF contains medium-chain (10:0-14:0) fatty acids, and HMF contains longer
chain (>16:0) fatty acids. The heterogeneous nature of the material, and its com-
plex phase behavior, causes drastic differences in the solid nature of AMF crystal
networks when storage temperature or crystallization conditions are modified. AMF
lipid composition is the fundamental property affecting all of the structural charac-
teristics of the crystallized bulk fat (Figure 1), because it will influence nucleation,
crystal growth, and the formation of the final crystal network.
There have been various attempts to modify the lipid composition of AMF as a
means of producing butter with improved spreadability. These studies have
included the fractionation and removal of portions of the HMF region to provide
a more spreadable product at refrigeration temperatures (24). Other attempts
PROCESSING CONDITIONS 125
have included the modification of bovine diets to incorporate particular fatty acids
into MF TAGs, and other work has concentrated on the chemical interesterification
of different fatty acids to the TAGs of AMF (9, 2932). These methods typically
affect the flavor of bulk AMF, and labeling the product butter is not permitted
because of the chemical adulterations performed. For this reason, processors are
continually trying to modify crystal structure, and thereby improve textural charac-
teristics solely through the modification of processing conditions.
4. PROCESSING CONDITIONS
The structure of most processed food depends not only on the ingredient formula-
tion, but also on the processing history of the material (33). Processing conditions,
such as crystallization, storage temperature, cooling rate, storage time, shear, and
tempering, affect the crystal structure and rheological properties of fats. In the model
proposed in Figure 1, processing conditions are considered an external factor, and
like lipid composition, affect the underlying physical properties of the fat crystal
network.
Adjusting processing parameters, such as cooling rate and/or crystallization tem-
perature, will cause fats to exhibit differences in physical properties. Altering cool-
ing rate has been shown to induce the formation of different polymorphs in AMF
(34). This research concentrated on determining the initial polymorph present in
AMF upon cooling at various controlled rates. AMF was cooled from 70 C to
65 C at different cooling rates, and polymorphism was monitored using differen-
tial scanning calorimetry (DSC) and powder X-ray diffraction. At temperatures
above 0 C, the a-polymorphic form was predominant at all cooling rates above
1 C/min. At 1 C/min, a mixture of both a- and b0 -polymorphs was present, and
at 0.5 C/min, only the b0 -polymorph was detectable. This indicates that the proces-
sing conditions affect the nucleation and crystal growth of AMF, as indicated by
differences in polymorphism.
Herrera and Hartel (35, 36) demonstrated differences in microstructural size and
structure as a result of cooling rate using milkfat and milkfat blends. Cooling a sam-
ple more rapidly resulted in smaller and more numerous crystallites. Additionally,
they demonstrated differences in SFC and polymorphism. Differences were induced
at the nucleation and crystal growth stages. They further demonstrated the effects of
cooling rate on the rheological characteristics of the system using the compression
storage modulus (E0 ) as an indicator (37). They found that, under shear, the storage
modulus decreased with increases in cooling rate (smaller particle sizes), therefore
making a link among processing conditions, solid fat content, and microstructure
and macroscopic properties (3537).
These various examples depict the effects of processing conditions on the phy-
sical properties of fats and demonstrate that external crystallization conditions, such
as cooling rate and storage temperature, can have dramatic effects on the final mea-
surable properties of a fat.
5. NUCLEATION AND CRYSTAL GROWTH
Crystal shape, size, and density all affect the physical properties of the final solid fat
matrix. Crystal growth, primary nucleation, and secondary nucleation in fat systems
are influenced by many factors, including diffusion, molecular compatibility, TAG
structure, nuclei composition and surface properties, number of nuclei, and proces-
sing conditions (temperature and/or shear) (38, 39). It is during the crystallization
process of fats that the template for the final physical properties of the material is
created.
5.1. Nucleation Mechanisms

Crystallization does not take place until the melt becomes supersaturated or
undercooled (38). Like all substances, fats cannot crystallize until tiny embryo-
nic crystals, known as nuclei, are formed. Supersaturation and consequent homoge-
neous nucleation is unlikely in MF because MF is of a heterogeneous nature with
virtually no pure TAG species in concentrations greater than 2 mol% (39, 40). In the
melt, there are two primary mechanisms by which nucleation takes place. These
nucleation processes, illustrated in Figure 2, are referred to as homogeneous and
heterogeneous, the latter being predominant in AMF crystallization (39, 40). After
nucleation, crystal growth takes place on the surface of existing nuclei (38).
Figure 2. Schematic depicts nucleation and crystal growth processes.

NUCLEATION AND CRYSTAL GROWTH 127
5.1.1. Homogeneous Nucleation Homogeneous nucleation, resulting from

bimolecular reactions between the TAG species, leads to the formation of nuclei
in the absence of foreign particles (39, 40). To achieve homogeneous nucleation,
no solid substrate or contaminant can exist in the fat. TAGs interact only with
one another, usually at higher degrees of supercooling (in most cases are greater
than 30 C below the final melting temperature). AMF typically will not nucleate
in a homogeneous fashion because of the following
1. The presence of catalytic impurities such as water, dirt, monoacylglycerols,

diacylglycerols, phospholipids, and proteins.
2. The presence of a large variety of different and incompatible TAGs in
concentrations of less than 1 mol% of the entire AMF volume (40).
3. The presence of agitation, temperature gradients, inducing inconsistent
nucleation, and diffusion-limited gradients (39).
5.1.2. Heterogeneous Nucleation Heterogeneous nucleation is the most com-

mon process in AMF crystallization (39, 40). Nucleation is induced by the presence
of foreign particles or catalytic impurities, and therefore, it requires a far lower
degree of undercooling (as low as 3 C) than that required for homogeneous nuclea-
tion (39, 40). Therefore, because of the large quantity of structurally incompatible
TAGs in AMF, nucleation on catalytic impurities is entropically favored over homo-
geneous nucleation.
5.1.3. Secondary Nucleation and Crystal Growth Secondary nucleation is

the process whereby nuclei are created as a result of inhomogeneous growth on pri-
mary crystals, or crystal breakage as a result of processing, which yield new inter-
faces for the creation of secondary nuclei. Secondary growth is the continued
solidification of TAGs onto the surface of already existing nuclei, which allows
crystals to form larger structures. During crystal growth, fat crystals take shape
by forming complex spherulitic or needle-like crystal structures (41). In AMF,
spherulitic crystal structures are predominant. Spherulites tend to have a very dense
crystal center with decreasing density as the distance from the crystal center
increases (41). Spherulites can continue to grow and aggregate to form a three-
dimensional network.
5.2. Factors Directly Influenced by Nucleation and Crystal Growth

5.2.1. Solid Fat Content Many methods for measuring SFC have been devel-
oped. These include dilatometry, calorimetry, and pulsed nuclear magnetic reso-
nance (pNMR). Dilatometry and calorimetry use measurements of volume or
heat content ratios between the completely liquid and the completely solid states
(42). Dilatometry and calorimetry methods are time consuming and tend to be
applicable only when the SFC is less than 50% (42). Therefore, pNMR has become
the most commonly used method for SFC determination.
5.2.1.1. Solid Fat Content by pNMR NMR techniques allow for a rapid determi-
nation of SFC with increased accuracy and precision relative to dilatometry and
calorimetry, while offering additional advantages of being noninvasive and rapid
(6 seconds per sample) (43). The NMR method for SFC determination is based
on the principle that nuclear spins align in a magnetic field. In a pNMR measure-
ment, the realignment of the nuclear spins of 1H atoms after a strong electromag-
netic pulse is measured (43). Application of a short 90 radio frequency
electromagnetic pulse provides excitation to the 1H atoms, causing them to leave
their equilibrium, aligned state. The pulse is then withdrawn leading to the relaxa-
tion of the 1H atoms and a return to the aligned, equilibrium state (43). Differences
in the time scale for the relaxation process of the solid and liquid protons are used
to determine the SFC. Protons, in general, have less mobility in the solid phase than
in the liquid phase, therefore leading to large differences in the relaxation times of
the 1H atoms (43).
A typical excitation and relaxation cycle measured using NMR is shown in Fig-
ure 3. The determination of SFC requires measurements at two time periods, one
within the short solid response region, S0 (11 ms), and one within the longer liquid
response region, S00 (70 ms). The signal intensity S0 provides an estimate of the
amount of both solid and liquid protons. Moreover, as the measurement cannot
be taken immediately because of receiver dead time, the true amount of solid
and liquid protons cannot be directly measured. For this reason, an extrapolation
factor, known as the F-factor (F), is used to approximate the maximum response
signal (S) value (43). SFC is calculated using the terms derived from the curve
in Figure 3 and the NMR digital offset (D) using the following equation:
S0 S00 F
SFCDirect : 1
S00 S0 S00 F D
Figure 3. Typical signal decay for a partially crystallized fat, following a 90 r.f. electromagnetic
pulse. Parameters required for measurements of solid fat content (SFC) are shown.
The value calculated is then compared with the signal obtained from either standard
oil or polymer calibration standards to determine the actual SFC value. This method
is very reproducible and provides standard deviations of less than 1%, indicating a
reliable and rapid method for SFC determination in fats (44).
5.2.1.2. Solid Fat Content and the Fat Crystal Network The solids content of a
fat crystal network is of critical importance to the final physical properties of the
system. Generally, an increase in SFC leads to an increase in fat firmness. The SFC
measurement has been widely used as a determinant quantity for the structural
properties of fat systems. Estimations for commercial plastic fats, including butter,
predict firmness increases of 10% for every percent increase in SFC (45). As a
result, models used to describe the rheological properties of fats incorporate refer-
ences to SFC values.
The primary factor affecting the SFC of any fat network is molecular composi-
tion. In general, longer chain fatty acids will have higher melting and crystallization
temperatures than shorter chain fatty acids. Typically, the larger the amount of satu-
rated, long-chain fatty acids, the greater the SFC. Processing conditions, including
cooling rate, agitation, and tempering, have the ability to affect the SFC of fats.
5.2.1.3. Effects of Cooling Rate and Storage Time on SFC in AMF AMF solid
fat content is affected by processing conditions, in particular by cooling rate. The
effects of cooling rate and storage time on SFC at 5 C can be appreciated in Fig-
ure 4. The effects of cooling rate, storage time, and their interaction have been
determined to be statistically significant (p < 0:05). Slowly cooling AMF at a
rate of 0.1 C/min results in a 48% lower SFC than for samples that are cooled
more rapidly (1 C/min and 5 C/min). The SFC of slowly cooled samples also
55.0
52.5
50.0
SFC (%)
47.5
0.1C/min
45.0 1C/min
5C/min
42.5
0 2 4 6 8 10 12 14
Storage Time (days)
Figure 4. Solid fat content of anhydrous milkfat cooled at 0.1C/min, 1C/min, and 5 C/min and
stored for a period of 14 days at 5 C.
increased slightly in time. On the other hand, at the higher cooling rates of 1 C/min
and 5 C/min, SFC does not statistically change in time. This demonstrates that
cooling rate has a significant effect SFC of AMF, and that small changes in SFC
can occur as a result of storage at 5 C for 14 days.
5.2.1.4. Crystallization Kinetics Monitored using SFC by pNMR Solid fat con-
tent measurements can be used to monitor the kinetics of crystallization. SFC
was monitored by pNMR as a function of time while samples were being cooled
in a water bath. These results are presented in a plot of SFC versus time in Figure 5.
The 0.1 C/min crystallization curve indicates that crystal growth begins at
approximately 23 C, or 170 min, and continues over time once 5 C is reached.
Crystal growth continues at a relatively constant rate until 205 min (19.5 C). At
this point, a plateau-like region occurs in the SFC profile, indicating a significant
reduction in crystal growth until a time of 260 min (14 C). After this plateau region,
crystal growth begins to occur at a more rapid and constant rate until the final tem-
perature of 5 C is reached. At this point, the SFC increases very slowly until an
equilibrium SFC is achieved.
The break in the crystallization curve for cooling at 0.1 C/min possibly indicates
that different AMF fractions crystallize at different times, as a result of the slow
cooling rate. This fractionation is common in AMF because of the large variety
of TAGs and fatty acid species that are present. Within the first crystal growth
region, the longer chain saturated HMF triacylglycerols (16:0 to 20:0) crystallize
and contribute to crystal growth and subsequent increase in SFC. Within the pla-
teau-like region, TAGs containing both long- and short-chain fatty acids reach their
60
50
40 reaches 5C
SFC (%)
30
0.1C/min
20 19.5C
1C/min
5C/min
10 14C
17C 23C
0 100 200 300 400 1400 1800

Time (min)
Figure 5. Crystallization and crystal growth of anhydrous milkfat at controlled cooling rates of
0.1C/min, 1C/min, and 5 C/min monitored using SFC by pNMR.
supersaturated state and crystallize. Beyond the plateau region, medium-chain satu-
rated MMF triacylglycerols (10:0 to 14:0) become supersaturated and contribute to
crystal growth. The implications of this crystal growth pattern will become more
apparent in the discussion of the microscopic and macroscopic properties of the
crystallized AMF. Beyond the final temperature of 5 C, any remaining TAGs that
can become supersaturated continue to crystallize until an equilibrium state is
achieved.
The 1 C/min and 5 C/min crystallization profiles indicate that crystal mass
becomes detectable at approximately 17 C in both cases, which occurs after
23 min and 4 min, respectively (Figure 5). Once this temperature is reached, there
is a constant and very rapid rate of crystal growth until the final temperature of 5 C
is reached. At approximately 12 C, there is an observable fluctuation in the SFC
growth curve that is followed by a continued rapid crystal growth until the equili-
brium temperature of 5 C is reached. This minor fluctuation in the SFC growth
curves may indicate a polymorphic transition or more likely the occurrence of frac-
tionation. The rapid and more spontaneous nucleation and crystal growth that
occurs at higher cooling rates leads to less fractionation and the formation of
more numerous mixed crystals containing a larger array of TAGs and a final SFC
that is 48% higher than the samples cooled at 0.1 C/min after 24 hours of storage.
5.3. Polymorphism
Polymorphism is another characteristic of fat networks that affects their rheological
characteristics. Polymorphic forms are crystalline phases with different structural
characteristics, but of identical chemical compositions in their liquid state, when
melted (45, 46). Polymorphic forms are usually categorized according to character-
istic X-ray diffraction patterns, specific volume, and/or melting points (34, 45, 47).
The polymorphic form developed during crystallization of bulk fat can be influ-
enced by several factors, including fat purity, TAG compatibility, temperature,
supercooling, cooling rate, catalytic impurities, solvents, and seed crystals (45).
Phase transitions from one polymorphic form to the next may also occur during
processing and storage depending on many factors (45, 46).
The three common polymorphic forms that exist in fat crystal networks are the
suba, a, b0 , and b modifications. These polymorphic modifications and their char-
acteristic crystallization patterns are shown in Figure 6. The suba and a forms are
metastable (34). Each polymorphic form yields different crystal structures depen-
dent on the magnitude of the crystallization driving force. The polymorphic mod-
ifications also have varying thermodynamic stability, which determines their
lifetime within a crystal matrix, with the tendency toward greater stability in the
order a to b0 to b (24, 34).
5.3.1. Polymorphism and Milkfat High cooling rates (>1 C/min), or high
levels of supercooling (>15 C), lead to the rapid formation of metastable a nuclei
(34). The persistence of these unstable nuclei is dependent on thermal treatments
that occur after crystallization. These nuclei may remain in the a form or convert
Figure 6. Polymorphic forms of fat crystals, including the possible polymorphic transitions,
subcell packing structures, stability characteristic, and triacylglycerol stacking conformations.
to the more stable b0 structure. At lower degrees of supercooling or low cooling

rates (>0 C/min to 1 C/min), b0 crystals predominate with trace amounts of meta-
stable a nuclei detectable (34). Many factors influence the formation of different
polymorphic forms, including catalytic impurities, agitation, viscosity, TAG con-
centration, among others (48).
5.3.1.1. The alpha (a) Polymorph The a polymorph is most readily formed, but
it is very unstable. Under most conditions, it will transform within a short period of
time into the b0 polymorphic form (34, 40). The a crystals are composed of TAGs
whose long, alkane-like fatty acid chains are packed in a loose hexagonal subcell
conformation, as shown in Figure 6 (3, 46). The hexagonal a subcell has a lower
density because of the loose packing of the TAG molecules and is characterized by
a single X-ray short spacing (wide-angle reflection) at 4.15 A (34). This loose pack-
ing allows for a large number of reaction sites, the incorporation of a wide variety
of TAGs into the solid fat matrix, and ample room for molecular rotation and re-
orientation. This allows for transitions from the a to b0 polymorphic form to occur
readily (3, 24). This open lattice allows mixed crystal formation, incorporating a
mixture of all of the supersaturated TAG present in the melt within the crystals
(3). In AMF, the formation of the a form takes place during crystallizations at
high degrees of supercooling or at high cooling rates, and it may persist as the pre-
dominant polymorphic form when the fat is stored at subzero temperatures (34). In
most situations, the a form is the first crystal structure formed during crystallization
because of the loose packing and lower supersaturation requirements. However, a
crystals tend to remain small and eventually transform, through the melt or mole-
cular reorientation in the solid state, to the more stable b0 modification (3).
5.3.1.2. The Beta Prime (b0 ) Polymorph The b0 polymorph is the most common
polymorphic form in AMF. It will form directly from the melt or through transfor-
mation of the a polymorph (Figure 6) (24). The b0 polymorphic form exists in either
a double- or triple-chain length configuration, the latter being more common in
AMF, and orients in a more dense orthorhombic subcell structure (Figure 6), char-
acterized by X-ray short spacings at 3.8 and 4.2 A (45, 46). Processors prefer this
form because it is structurally stable and maintains small- to-moderate crystal sizes
allowing for soft and smooth products (3). The b0 structure does not allow for the
incorporation of a large variety of TAGs within the crystal lattice, compared with
the a structure (3). Lower TAG variety is a direct result of the tighter packing
required to form the more stable structure, which requires similar length TAGs.
During crystallization in the a form, dissimilar TAGs will often cocrystallize tem-
porarily; however, in time, the shorter chain, or more unsaturated, less supersatu-
rated molecules will redissolve. This, in turn, allows the remaining, more similar
longer chain supersaturated TAGs to convert to the more stable b0 form (3, 40).
TAGs in AMF under moderate undercooling conditions will crystallize in both
the a and b0 polymorphic forms.
5.3.1.3. The Beta (b) Polymorph The presence of the b polymorph is uncommon
in AMF, but it has been known to exist in some situations (34). It is the most stable,
highest packing density polymorphic form, and it may exist in either double- or
triple-chain length conformation. Fatty acid chains within TAGs pack in a hexago-
nal arrangement (Figure 6), with a characteristic short spacing at 4.6 A (34, 45).
The b form is uncommon in AMF because of the dense packing arrangement of
similar TAGs. Because of the wide variety of TAG in low individual concentrations
(<1 mol%) in AMF, the conditions required to form b structures are only met when
the fat is crystallized at very slow rates, or at low degrees of supercooling (34, 46).
These crystallization conditions are unusual in commercial practice; thus, the b
polymorph is rare. The formation of b solids is also undesirable in AMF because
it leads to the development of large crystals and sand-like texture (3, 34).
5.3.1.4. Polymorphism in Milk Fat as a Result of Processing Conditions Work

by ten Groetenhuis et al. (34) has shown that AMF typically crystallizes in two frac-
tions. The highest melting group of TAG species represents one fraction, and the
middle and low-molecular-weight TAGs form the second. This group also suggests
that the application of different cooling rates lead to variations in the initial poly-
morphic form during crystallization of MF (34).
Two regions of crystallization can be identified from the two major endotherms
in the melting profiles of samples subjected to different cooling rates, as shown in
Figure 7. Immediate melting profiles showed differences, suggesting that the initial
polymorphic forms differ in AMF crystallized at various rates. Figure 7a depicts the
melting profile of samples cooled at various rates after 10 minutes of storage at 5 C.
Figure 7. Differential scanning calorimetry curves for anhydrous milkfat cooled at rates of 0.1 C/
min, 1 C/min, and 5 C/min to 5 C and stored for time periods (A) 10 minutes, (B) 1 day, (C) 7
days, and (D) 14 days.
The sample cooled at 0.1 C/min was expected to have had sufficient time for rear-
rangement, resulting in more dense packing of the TAGs, and crystallized in the
stable b0 polymorph (34). For this reason, the peak of the endotherm at 17.5 C is
assumed to correspond to the melting temperature of the b0 polymorph.
The samples cooled at 1 C/min and 5 C/min (Figure 7a) both have minor inflec-
tions in the curve at this same temperature (17.5 C), but they also have a major
peak endotherm at around 15 C. This may correspond to an a-polymorphic melt.
It is thus suggested that cooling at 1 C/min results in a mixture of a and b0 poly-
morphs, and cooling at 5 C/min results predominantly in the a polymorph. These
predictions correspond to the findings of previous studies on AMF employing simi-
lar crystallization conditions using DSC and powder X-ray diffraction (34).
The samples cooled at the three cooling rates were also stored at 5 C for 1, 7,
and 14 days, yielding the melting profiles shown in Figures 7b, c, and d, respec-
tively. After storage at 5 C, the melting profiles began to look similar irrespective
of the cooling rate used during crystallization. As it has been shown that AMF tends
to crystallize in the b0 polymorph upon storage (some believe that the a polymorph
survives for only minutes), similar shapes and peak melting points would be
expected (3, 40).
5.4. Microstructure
The textural properties of a fat are influenced by all levels of structure, particularly
microstructure. The microstructure includes the spatial distribution of mass, particle
size, interparticle separation distance, particle shape, and interparticle interaction
forces (4951). Methods that can be used for the characterization of microstructure
in fat systems include, among others, small deformation rheology and polarized
light microscopy, employing a fractal approach (4951).
5.4.1. Fat Crystal Network Theory Fat crystal growth is dictated by external
processing conditions and TAG composition. TAGs nucleate and grow from the
melt into certain polymorphic and polytypic states. These primary crystals then
aggregate into larger polycrystalline particles, also known as microstructural ele-
ments. Aggregation continues, leading to the formation of larger clusters, or micro-
structures, until a space-filling three-dimensional network is formed through
interactions among microstructures. Crystalline mass within microstructures is dis-
tributed in a heterogeneous, disordered fashion, which can be characterized using
fractal scaling principles (4951).
Perfect fractal objects display exactly self-similar at all levels of magnification.
Natural systems do not exhibit exact self-similarity, but statistical self-similarity;
i.e., the microstructure, on average, is similar in appearance, distribution, and struc-
ture within a limited range of magnifications. In particles networks, fractal scaling
is usually encountered within the range of magnifications corresponding to the size
of microstructural elements to the size of microstructures (51). An example of sta-
tistical self-similarity uses the analogy of a tree branch structure. The tree begins as
a trunk, the tree trunk has branches, these branches have branches, and so on. When
the scale of observation is changed, a statistically self-similar pattern is observed.
This theory was first developed for colloidal aggregate networks and was later
adapted to fat crystal networks (5254). In colloidal systems (with a disordered dis-
tribution of mass and statistical self-similar patterns), the mass of a fractal aggre-
gate (or the distribution of mass within a network), M, is related to the size of the
object or region of interest (R) in a power-law fashion:
M RD ; 2
where D is the mass fractal dimension of the object, or the distribution of mass
within a region of the network.
The elasticity of a fat crystal network is dependent on the microstructure of the
fat crystal network, particularly the spatial distribution of mass. The shear modulus
(G) scales with the volume fraction of solids in a power-law fashion (4954):
G m ; 3
where is the volume fraction of solid fat, or SFC/100, and m is dependent on D

(see below).
Figure 8. Idealized fat crystal network under extension. Particles (a) are packed in a fractal
fashion within flocs (x). A force (F) acting on the network causes the links between flocs to yield,
and the original length of the system in the direction of the applied force (L) to increase L.
Thus, the inter-floc separation distance (l), also increases.
The scaling behavior of the elastic modulus was first characterized for colloidal
gels within two specific rheological regimes (55) and was later adapted to fat crystal
networks (5254). These two regimes depend on the strength of the links that exist
between individual clusters, relative to inner cluster strength. These regimes are
referred to as the strong-link regime and weak-link regime (55). Figure 8
illustrates the behavior of a fat crystal network under extension in relation to these
theoretical rheological regimes. The strong-link regime is only applicable at very
low solid fat contents (usually below 10%). In the strong-link regime, crystal clus-
ters grow large, and the links between these flocs are stronger than the flocs them-
selves. The elastic response of the material in this regime is a function of the elastic
response of the flocs, and it is thus dependent on internal aggregate structure.
The contrary is true for the weak-link regime, which is observed at high solid fat
contents, where flocs are small and stronger than the links between them. In this
regime, the elastic response of the material is a function of the elastic response
of the links between flocs, and it is not dependent on the structure within the flocs.
In both regimes, the macroscopic elastic modulus (K) of a system of size L is the
sum of the elastic constants of the flocs (kx ) or the links between flocs (kL) in the
direction of the applied stress (one-dimensional treatment). In the weak-link
regime,

L
K kL x1 ; 4
x
where x is the size of a floc. Floc size (x) scales with the volume fraction of solids
() in a power-law fashion:
1
x aDd ; 5
where d is the Euclidean dimension (usually d 3), and a is the size of a primary
particle. By inserting Equation 5 into Equation 4, we obtain the following relation-
ship:
1
K 3D G: 6
This equation can also be expressed as

1
G l3D ; 7
where l is a constant independent of the volume fraction, but dependent on several

primary particle structural parameters as well as intermolecular forces.
Our group has derived expressions for the Youngs modulus (E) of a fat as it
relates to the structure of the material (5658):
6d 1 A 1
E 3D 3D ; 8
ae 2pae d02
where d is the crystal-melt interfacial tension (about 0.01 J/m2 for TAGs), a is the
primary particle size, is the volume fraction of solids (SFC/100), D is the fractal
dimension, A is Hamackers constant (about 5
1020 J for alkanes), e is the
extensional/compressional strain at the limit of linearity, and do is the equilibrium
inter-microstructural (flocs or clusters) separation distance. Values for the shear
modulus (G) could be obtained from knowledge of the Poisson ratio of the material.
For a material where no volume change takes place when it is stretched or com-
pressed, the Poissons ratio is 0.5 and E 3 G.
Careful scrutiny of Equation 8 would suggest that the stress at the limit of lin-
earity (s ) can be determined from the product of the Youngs modulus (E) and the
strain at the limit of linearity (e ), s E e , yielding the expression:
6d 1
s 3D : 9
a
This expression for the stress at the limit of linearity, proposed in Marangoni and
Rogers (58), provides an approximation to the yield stress (and thus hardness) of a
fat. This model would allow for the prediction of the yield stress of a fat based on
easily determined structural characteristics.
5.4.2. Fractal Dimension Evaluation using Small Deformation Rheology

(Dr) To determine Dr for a particular system by rheological methods, the shear
storage modulus G0 must be measured at different solids volume fractions.
The solids volume fraction, which is equal to the SFC/100, is varied by diluting
the fat with an inert solvent that will not cocrystallize, or otherwise alter the beha-
vior of the system. Canola oil has been used as a suitable AMF diluent for these
purposes. A loglog plot of G0 versus yields a slope m from which Dr can be
TABLE 2. Rheologically Determined Fractal Dimensions Dr and Pre-

Exponential Terms (l) for Anhydrous Milkfat Crystallized at Various
Rates of Cooling and Storage Times at 5 C.
Processing Conditions Microstructural Parameters

Cooling Rate Storage Time
C=min (days) Dr lMPa
1 2.82 1150
0.1 7 2.78 376
14 2.79 400
1 2.67 91
1.0 7 2.59 65
14 2.60 64
1 2.57 65
5.0 7 2.50 52
14 2.47 49
determined. The pre-exponential term l, which is influenced primarily by particle

properties, can be determined from the value of G0 , where 1:0 (SFC 100%).
The effects of cooling rate on Dr are shown in Table 2. In general, an increase in
cooling rate results in a decrease in Dr. Previous work in our laboratory has also
shown that higher fractal dimensions occur in networks that are more ordered;
thus, it can be expected that lower cooling rates would display this trend. The higher
the Dr, the more ordered the crystal packing. For this reason, the value for Dr is
expected to vary in the order of Dr0:1 C=min > Dr1 C=min > Dr5 C=min . When
observed by time-lapsed microscopy, the samples cooled slowly (0.1 C/min) and
yielded less numerous (see Figure 12 below), but more sporadically formed nuclei
relative to the more rapidly cooled samples. This allows for a more ordered crystal
growth. The 1 C/min sample initially demonstrated sporadic nucleation and con-
trolled crystal growth of more numerous and smaller crystals than the slower cooled
sample until reaching 13 C. Beyond 13 C, the AMF began to nucleate sponta-
neously and fill space very rapidly, leading to a more disordered final system. Finally,
samples cooled at 5 C/min demonstrated spontaneous nucleation and rapid crystal
growth at 16 C. This resulted in a more disordered system with a fractal dimension
lower than for slower cooling rates.
The values for l, shown in Table 2, are also affected by cooling rate and storage
time at 5 C. As with the Dr, the values for l decrease with increases in cooling rate.
During storage at 5 C, the l values decrease moderately at each cooling rate and
then appear to equilibrate at 7 days, with only minor differences detectable between
the 7-day and 14-day values. This decrease may be attributable to continued crystal
growth (demonstrated by small increases in SFC) and by the restructuring that
occurs during polymorphic transitions and molecular rearrangement.
5.4.2.1. Fractal Dimension by Microscopy The use of fractal geometry as a

means of characterizing microstructure has been extensive in recent years. Various
applications have included investigating the structure of protein gels (59), depicting
the irregular nature of solid adsorbents such as zeolites (60), and predicting optical
responses of colloid-adsorbate films (61). Recent work by Kim and Berg (62) has
focused on the use of fractal scaling as a link between aggregation kinetics and
structure that result from both diffusion and reaction-limited cluster aggregation
processes of colloidal materials.
To obtain the fractal dimension of a network of particles, acquiring images of the
microstructure is necessary. Many forms of microscopy can be used, including
brightfield microscopy, confocal laser scanning microscopy, scanning electron
microscopy, and in the case of fat crystal networks, polarized light microscopy.
Grayscale images of the network (usually consisting of pixel intensities ranging
from pure white to pure black) must be converted to a binary format prior to image
analysis. The discriminatory conversion of grayscale images to binary, requires a
process termed thresholding. Images of fat crystal networks are typically
obtained as 8-bit (256 color) images and consist of a histogram of pixel values
that span the range of 0 (pure white) and 255 (pure black). Thresholding involves
the selection of a particular pixel value within the range of 0 and 255 that will result
in a binary image that best isolates the features of the network. This process is
subject to much controversy because the selection of the cutoff pixel value is sub-
jective. To minimize subjectivity, an autothresholding algorithm was used that pro-
vides accurate and reproducible results.1 This algorithm works by scanning the
histogram to find an intensity value where the average moments of the histogram
counts about an intensity value are balanced. This means that a threshold value is
chosen where the average pixel intensities are equal above and below the threshold
(63). The general process of autothresholding can be seen in Figure 9. Once the
image is converted to binary and the features isolated, it can be subjected to image
analysis. Features are measured, and result in data, which can then be interpreted
and related to structure. A flow chart outlining the general process from start to fin-
ish is depicted in Figure 10 (64). Emphasis is put on acquiring initial images that
are of high quality (high distribution of pixel values, no saturation of white features,
and evenly distributed lighting) to avoid unnecessary image processing to enhance
features, thereby altering pixel values.
Assuming a statistically constant microstructural element (or particle) size, the
relationship between radius and mass (Equation 2) can be used to determine the
fractality of crystal networks from two-dimensional PLM images (54).
In this case, the scaling relationship between the number of discrete particles (N)
and the length of the region of interest containing the particles (L) can be expressed
as:
N cLD : 10
1
Performed on a Macintosh computer using the public domain NIH Image program, developed at the U.S.
National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/
Figure 9. The conversion of 256-color grayscale micrograph to a binary image suitable for
image analysis. A threshold value is chosen to isolate the featues that will be measured
(AdobePhotoshop1 6.0).
Figure 10. Flowchart depicts the steps involved in obtaining and evaluating the microstructure
of materials from digitally acquired images.
The first method used was a particle counting algorithm performed using NIH-
Image2 as follows: c is a constant, and the number of distinct particles (N) are first
counted within the entire image of known length (L). At 5% increments, the image
is cropped, thereby making a new, smaller length L, and the number of particles in
each new image are counted. This process is repeated until the length of the image
is 35% of the original size. A diagram depicting the overlaying of the boxes of
decreasing size can be appreciated in Figure 11. The number of particles counted
within each box size of various lengths (L) is plotted on a loglog scale, and the
slope of the line is determined by linear regression. This slope corresponds to Df.
This method for calculating Df was modified from previous procedures in our
laboratory to eliminate artifacts and improve accuracy. This was accomplished
by performing these iterative counts twice on each imageonce including all par-
ticles touching the edge of each region of interest, and once excluding those that
touch the edges. By taking the average of these two counts, a Df that best represents
the spatial distribution of mass is obtained. This improved method is equivalent to
well-established methods that involve counting features that touch two sides of the
region of interest (64).
The fractal dimension arrived at by this method reveals information on the
degree of order in the packing of the microstructural elements within the micro-
structures. Systems that display a high degree of order have characteristically lower
Figure 11. Schematic diagram shows the incremental decreases in box size used for the
particle counting method for the determination of the microscopic fractal dimension D f .
2
Analysis performed on a Macintosh computer using the public domain NIH Image program, developed at
the U.S. National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/
fractal dimensions than those of a more disordered network (54). Previous simula-
tions and studies employing mass-radius techniques for the determination of D have
found that fast, diffusion-limited cluster aggregation (DLCA) typically results in D
values of 1.751.8, whereas slow, reaction-limited cluster aggregation (RLCA) pro-
cesses result in D values of 2.02.1 (60).
Another method used to examine microstructure is the box counting or grid
dimension analysis. This dimension is referred to as a grid dimension because
for mathematical convenience, the boxes are usually part of a grid that is laid
over the image (65). The box dimension is defined as the exponent D in the relation-
ship:
1
NL ; 11
LD
where N(L) is the number of boxes of linear size L necessary to cover a data set
of points distributed in a two-dimensional plane. If the network is indeed fractal,
plotting the logarithm of N(L) versus the logarithm of L results in a linear plot
with a negative slope equal to D (65). This analysis appears to be sensitive to
the degree of fill of the solids in a crystal network. It can be expected that a
network that is more empty (many large void spaces) will result in a lower propor-
tion of full boxes counted, and thus, it will result in a lower fractal dimension and
vice versa.
It is important to note that in some cases, the fractal dimensions estimated by
different methods are not the same. It is therefore necessary that one be aware of
the particular method used and its interpretation; this is mandatory for comparisons
of dimensions of different data sets (65). One of the future goals within our labora-
tory is to continue investigating and quantifying the link between the fractal dimen-
sions determined by rheology and microscopy.
5.4.3. Polarized Light Microscopy Microscopy allows for the visual observa-
tion of crystallization, and crystal growth of fat networks in realtime and subse-
quent image analysis can be used to quantify particle size, degree of order, and
space-filling mass. Polarized light microscopy was used to examine crystal network
properties of AMF during nucleation and crystal growth, and over time as a result
of the processing conditions and storage time.
Cooling at 0.1 C/min, 1 C/min, and 5 C/min resulted in detectable birefringent
crystal mass at onset temperatures of 26.8 C, 20 C, and 16 C, respectively. The
onset temperatures of the 0.1 C/min and 1 C/min samples do not correspond
exactly to those determined by DSC or NMR. This is because of the small quantity
of matter crystallizing at the onset of nucleation, which although visible by PLM,
does not release enough heat to be resolved by the DSC or enough solid mass to be
detected using pNMR. The sample cooled at 5 C/min, on the other hand, shows
visual signs of crystallization at 16 C, which corresponds closely to the values
obtained in the DSC and pNMR experiments. This similarity in measurement
and resolution by each of the methods is attributable to the large change in state that
occurs at higher cooling rates.
Microscopy also allows for the observation of dynamic changes that occur dur-
ing nucleation and crystal growth. Figure 12 depicts still images of crystallizations
at cooling rates of 0.1 C/min, 1 C/min, and 5 C/min. The images shown are at 5 C
intervals in the range of 30 C to 5 C and portray dramatic differences in kinetics,
and crystal properties, as a result of altering the cooling rates.
The crystallization profile for AMF cooled at 0.1 C/min shown in Figure 12
spans a total time period of 250 minutes. At 26.8 C, sporadic nucleation begins,
and as time progresses, these early nuclei intermittently grow outward in a radial
fashion as the temperature decreases. Until late in the crystallization period (at
5 C), there is no visible evidence of secondary nucleation. Between the tempera-
tures of 26.8 C and 23 C, there is a discrete and continuous growth region where
the microstructures increase dramatically in size as a function of time. This is fol-
lowed by a period of very little growth (little evidence of changing microstructure
sizes) until 14.5 C, when significant growth occurs once again up until a tempera-
ture of 12 C is reached. From approximately 8 C until 5 C, minor space-filling
crystal growth occurs. These intermittent, temperature-dependentrelated growth
regions are likely attributable to the crystallization of the AMF fractions of HMF
(26.8C24 C), followed by MMF (14.5C12 C), and then the space-filling crystal-
lization of some of the LMF. This is made possible by the very slow change in
supercooling that occurs over time when cooling AMF at 0.1 C/min, which permits
adequate time for diffusion-related interaction to occur, leading to the fractionation-
mediated growth of the network. The resulting fat crystal network is made up of a
small number of very large crystal structures and relatively large regions of void
space as seen in Figure 12.
The crystallization profile for AMF cooled at 1 C/min shown in Figure 12 spans
a real-time period of 25 minutes. The first sign of crystal structure occurs at
20 C and is followed by a large amount of spontaneous nucleation and radial crys-
tal growth until a temperature of 16 C is reached. After the rapid growth period,
there is minimal visible growth until approximately 12 C is reached and a very
rapid, spontaneous, space-filling secondary nucleation and growth occurs and con-
tinues until 5 C is attained. This rapid period of space filling crystallization leads to
the masking of the larger microstructures formed during the earlier stages of crys-
tallization. Like the system cooled at 0.1 C/min, AMF cooled at 1 C/min demon-
strates fractionation-mediated growth where most of the HMF and some MMF
likely crystallizes during the early stages, followed by crystallization of the
MMF and LMF during the later stages. The resulting microstructure consists of a
large number of small crystal structures; however, previous micrographs collected
at the higher temperatures indicate the presence of somewhat larger structures
(Figure 12).
The crystallization profile of AMF cooled at 5 C/min spans a real-time period of
just 5 minutes (Figure 12). The first visible sign of a crystal structure occurs at a
temperature of 16 C, after which very spontaneous, and rapidly space-filling,
nucleation and crystal growth occurs. The progression of growth until 5 C shows
Figure 12. Polarized light micrographs of anhydrous milkfat cooled at 0.1C/min, 1C/min, and
5 C/min. Images were acquired during crystallization in the range of 30 C to 5 C at intervals of
5 C.
minimal differences over time with the exception of the occurrence of significantly
increasing birefringence with decreases in temperature. This is an indicator of the
densification and continued crystal growth of the microstructural elements within
the crystal network. The resulting network is a densely packed, space-filling
matrix of very small crystal structures. In general, as cooling rate increases, the
number of crystal aggregates also increases, and the size of the constituent particles
decreases.
The effects of storage time on microstructure at 5 C are represented in Figure 13.
For each cooling rate, the network structure and particle size established
during crystallization remain relatively unchanged during storage. Minor changes
in the appearance of the 0.1 C/min sample occur as a result of the filling of
void space, during storage time, by small crystal structures similar to those present
in the originally formed crystal matrix. Rapidly cooled samples demonstrate no
significant change in network or particle appearance as a function of storage
time at 5 C.
The effects of cooling rate and storage time on mean particle size, as determined
by image analysis, are shown in Table 3. Cooling at 0.1 C/min resulted in average
particle sizes that were approximately two times larger than those resulting from
cooling at 1 C/min, and 5 C/min. Also, the minor changes in the appearance of
samples cooled at 0.1 C/min can be further characterized by a slight decrease
Figure 13. Polarized light micrographs of anhydrous milkfat cooled at 0.1C/min, 1C/min, and
5 C/min followed by storage for 1 day, 7 days, and 14 days at 5 C.
TABLE 3. Particle-Counting Fractal Dimension Df , Box-Counting Fractal Dimension

Db , and Mean Microstructural Element Area MEA for Anhydrous Milkfat Crystallized
at Various Rates of Cooling and Storage Times at 5 C.
Processing Conditions Microstructural Parameters (n 1012)

Cooling Rate Storage Time
C=min (days) Df Db MEAmm2
1 1.98 0.11 1.72 0.01 24.76 7.83

0.1 7 1.98 0.09 1.73 0.01 15.71 1.43
14 1.98 0.08 1.73 0.01 15.72 1.42
1 1.92 0.08 1.85 0.01 4.12 0.47
1.0 7 1.88 0.04 1.85 0.01 3.91 0.22
14 1.86 0.07 1.85 0.01 3.44 0.17
1 1.91 0.03 1.87 0.01 3.88 0.40
5.0 7 1.91 0.04 1.88 0.00 3.50 0.19
14 1.89 0.04 1.88 0.00 3.26 0.07
in mean particle area (MEA total crystal area/no. of particles) as a function of

time. Samples cooled at higher rates of 1 C/min and 5 C/min remain relatively
constant in time.
Fractal dimensions of micrographs obtained at different cooling rates, deter-
mined using the particle counting method, reveal information about the degree of
order in the resulting networks (Table 3). The Df values decreased in the order of
Df0:1 C=min > Df5 C=min Df1 C=min , with no significant changes detectable over
time.
Fractal dimensions determined via the box-counting or grid dimension method
are also shown in Table 3. Db values decrease in the order of Db1 C=min >
Db5 C=min > Db0:1 C=min , and they do not significantly change in time. Db is
sensitive to the degree of fill within the network; therefore, higher values indicate
an increase in space-filling mass (i.e., the smaller, more-numerous particles of the
1 C/min sample fill more space relative to the larger, less-numerous 0.1 C/min
particles).
5.4.4. Importance of Nucleation and Crystal Growth Kinetics on the Final

Crystal Properties and Network As stated in the hierarchical model (Figure 1),
the processing and storage conditions affect the physical properties, including the
SFC, polymorphism, and microstructure during nucleation and crystal growth.
Therefore, processing conditions exert their most dramatic effect during crystalliza-
tion, during which the final crystal network properties are developed. Scrutiny of
the micrographs in Figure 12 demonstrates that the crystal structures formed during
the early stages of nucleation and crystal growth dictate the final fat crystal network
structure. The samples cooled at 0.1 C/min show fewer, more sporadic nuclei,
which appear at higher temperatures. These nuclei then grow slowly to form
large crystal structures that continue to grow until the final temperature of 5 C is
reached. Similar trends are seen in the faster cooled samples, where spontaneous
nucleation occurs, leading to a large number of small crystals. This distribution
remains at the final temperature of 5 C/min. Analysis of these micrographs using
the fractal dimensions (Df and Db) also indicates that the crystal network structure
is determined before the final temperature is reached. These results can be seen in
Figure 14. These results demonstrate that at 10 C15 C prior to the final
temperature of 5 C, the final equilibrium value (or close to the final value) for
2.5
2.0
Fractal Dimension (Df)
1.5
1.0
0.1C/min
0.5
1C/min
5C/min
0.0
35 25 15 5
Temperature (C)
(a)
2.0
Fractal Dimension (Db)
1.5
1.0
0.1C/min
0.5
1C/min
5C/min
0.0
35 25 15 5
Temperature (C)
(b)
Figure 14. Cooling profile analysis of polarized micrographs of 100% AMF collected during the
static crystallization process at the cooling rates of 0.1C/min, 1C/min, and 5 C/min depecting
the microscopic properties (A) D f and (B) D b monitored against changing temperature.
microscopically determined fractal dimensions are reached. This indicates

that the final structure is predetermined far before the final equilibrium state is
reached.
6. MECHANICAL PROPERTIES
6.1. Small Deformation

Small deformation rheometry refers to testing procedures that do not cause struc-
tural damage to the sample. Constant stress rheometers, such as dynamic mechan-
ical analyzers or oscillatory constant stress rheometers, are often used.
There have been a number of studies that demonstrate that crystallized AMF and
butter exhibit linear (ideal) viscoelastic behavior at low levels of stress or strain (4),
where the strain is directly proportional to the applied stress. For most materials,
this region occurs when the critical strain (strain where structure breaks down) is
less than 1.0%, but for fat networks, the strains typically exceed 0.1% (4, 66).
Ideally, within the LVR, milkfat crystal networks will behave like a Hookean solid
where the stress is directly proportional to the strain (i.e., s / g), as shown in
Figure 15 (66, 68). Within the elastic region, stress will increase linearly with strain
up to a critical strain. Beyond that critical strain (strain at the limit of linearity),
deformation of the network will occur at a point known as the yield point. The elas-
tic limit quickly follows, beyond which permanent deformation and sample fracture
occurs. Beyond these points, the structural integrity of the network is compromised
and the sample breaks down.
From small-deformation oscillatory methods, several useful parameters can be
obtained to describe the mechanical properties of a material. These measurements
Figure 15. Stress-strain behavior of a typical elastic system, including (A) yield point, (B) elastic
limit, (C) irreversible deformation, and (D) fracture.
MECHANICAL PROPERTIES 149
Figure 16. Time profile of an applied sinusoidal stress wave and the corresponding resulting
sinusoidal strain wave as they apply to small deformation rheological testing.
include the complex modulus (G*), the shear storage modulus G0 , the shear loss
modulus G00 , and the tangent of the phase shift or phase angle tan d.
These rheological parameters can be determined using controlled stress rhe-
ometers using dynamic oscillatory testing within the LVR region (68). The oscilla-
tory method collects strain information by applying a controlled stress via the
application of a sinusoidal stress wave. The rheometer measures the variation in
strain as a function of the applied stress, in terms of the magnitude of the strain
and the phase angle (d) between the applied stress wave and the resulting strain
wave. A typical stress-strain sinusoidal relationship is shown in Figure 16. The vec-
torial resolution, shown in Figure 17, of the stress-strain ratio is used to calculate
the complex modulus G*, which is derived from the following equation:
p
jG j G02 G002 : 12
The strain response can be broken down into its elemental components of stress,
which are in phase or out of phase, to derive the values for G0 and G00 . The storage
modulus G0 is the ratio of the applied stress that is in phase with the strain (d 0 ).
This means that G0 is an expression of the magnitude of the energy stored in the
material, recoverable per deformation cycle (68). The loss modulus G00 is the ratio
of the applied stress that is out of phase with the strain (d 90 ), meaning that it is a
measurement of the energy lost as viscous dissipation per deformation cycle (66
68). These two moduli are dependent on the phase angle of the system and are
Figure 17. Vectorial resolution of the complex modulus components.
derived from the vectorial components of G* and are calculated using the following
relationships:

0 s0
G cos d; 13
g
0
s0
G00 sin d: 14
g0
The tangent of the phase angle tan d can be expressed as the ratio of the loss to the
storage moduli, and it represents the relative balance of elastic to viscous compo-
nents in a material:
G00
tan d : 15
G0
These rheological parameters have been successfully correlated to textural attri-

butes of hardness and spreadability and provide information pertaining to the fat
crystal network (69). The value of G0 is useful in assessing the solid-like structure
of the fat crystal network. Increases in the value of G0 typically correspond to a
stronger network and a harder fat (66). Alternatively, G00 represents the fluid-like
behavior of the fat system. This value can be related to the spreadability of a fat
system, because increases in G00 indicate more fluid-like behavior under an applied
shear stress. The tan d is the ratio of these two values. As the value of d approaches
0 (stress wave in phase with stress wave), the G00 value approaches zero, and there-
fore, the sample behaves like an ideal solid and is referred to as perfectly elastic
(68). As d approaches 90 (stress is completely out of phase relative to the strain),
the G0 value approaches zero; all of the energy will be dissipated as heat, and the
sample will behave predominantly as a fluid (68). At intermittent values, the sam-
ples are considered to be viscoelastic in nature. Therefore, tan d is an excellent indi-
cator of the structural integrity of the sample, indicating the proportion of the
material structure attributable to the crystal network and to the liquid phase.
6.1.1. Effects of Processing ConditionsSmall Deformation Rheology The

effects of cooling rate and time on small deformation rheological measurements of
G0 , G00 , and tand are shown in Figures 18, 19, and 20, respectively.
Figure 18 illustrates the effects of cooling rate and storage time on the storage
modulus G0 measured using small deformation rheology at 5 C. There are no sig-
nificant effects of cooling rate or storage time on the G0 of AMF samples, indicating
the absence of differences in solid-like properties and potentially hardness.
Figure 19 illustrates the effects of cooling rate and storage time on the loss mod-
ulus G00 , measured using small deformation rheology at 5 C. Results show lower
G00 for samples cooled at 5 C/min and 1 C/min than at 0.1 C/min. Additionally, the
G00 values for the 5 C/min and 0.1 C/min cooling rates do not change significantly
with time, and the values for the 1 C/min samples decrease slightly with time,
which may be indicative of time-dependent hardening.
The tangent of the phase angle tan d offers a better indicator of structural integ-
rity than either the G0 or G00 measurements do individually. The phase angle (d) or
its tangent tan d are an indicator of the systems structural behavior or integrity
because it indicates whether the system behaves predominantly as a solid, liquid, or
viscoelastic material. The greater the tan d, the more liquid-like the samples will
behave, and conversely, low values indicate more solid-like properties. A tan d
value of 1 is indicative of a viscoelastic material.
20
15
G' (MPa)
10
0.1C/min
5
1C/min
5C/min
0
0 2 4 6 8 10 12 14
Storage Time (days)
Figure 18. Storage moduli G 0 of anhydrous milkfat cooled at rates of 0.1C/min, 1C/min, and
5 C/min followed by storage for 14 days at 5 C.
2.0
1.5
G(MPa)
1.0
0.1C/min
0.5
1C/min
5C/min
0.0
0 2 4 6 8 10 12 14
Storage Time (days)
Figure 19. Loss moduli G 00 of anhydrous milkfat cooled at rates of 0.1C/min, 1C/min, and
The graphs in Figure 20 demonstrate the effects of cooling rate and storage time
on tan d at 5 C. Values of tan d increase in the order of tan d5 C/min < tan d1 C/min
tan d0.1 C/min. The tan d value for AMF cooled at 0.1 C/min is significantly high-
er than the values at 1 C/min and 5 C/min. The values do not show large increases
with time; therefore, the changes during storage at 5 C are minimal.
The data from small deformation rheology indicate that at higher cooling rates,
the fat samples are more solid-like in nature. Conversely, the low cooling rate of
0.1 C/min results in a softer, less elastic system.
0.15
0.10
tan()
0.05
0.1C/min
1C/min
5C/min
0.00
0 2 4 6 8 10 12 14
Storage Time (days)
Figure 20. Values of tan d of anhydrous milkfat cooled at rates at 0.1C/min, 1C/min, and 5 C/min
followed by storage for 14 days at 5 C.
6.2. Large Deformation

Large deformation methods usually are based on the determination of the amount of
applied force required to induce a change in a sample. Parameters determined
include hardness, spreadability, cutting force, or yield force (4). One such method
involves the compression of a sample between two parallel plates to determine rela-
tive hardness values via the measurement of yield force.
The compression of uniform samples to the point where the force exceeds the
structural capacity causes it to permanently deform and essentially break (4). A
typical load-deformation curve can be used to derive values for yield stress, yield
strain, and compressive yield work, and depending on the linearity of the onset of
compression, a compressive modulus may be obtained (4). These measurements
can be used to provide an index of hardness for fats, which have been successfully
correlated to the textural attributes of hardness and spreadability obtained through
sensory evaluation (4). Unfortunately, these tests are destructive in nature and yield
minimal information about the native microstructure of the system.
6.2.1. Effect of Processing ConditionsLarge Deformation Rheology The

effects of cooling rate and time on the yield force of AMF and its dilutions are
shown in Figure 21. Yield force measurements (Fy) demonstrate that cooling rate
and, to a lesser extent, storage time affect the hardness of AMF. In general, the Fy
increases with increasing cooling rate; however, the difference between 1 C/min
and 5 C/min is relatively small compared with the difference observed between
0.1 C/min and 1 C/min. AMF samples cooled at 0.1 C/min have yield force values
(3035 N) that are approximately half that of the higher cooling rates of 1 C/min
(5863 N) and 5 C/min (6472 N). Over time, the yield force increases slightly, by
75
50
Yield Force (N)
25
0.1C/min
1C/min
5C/min
0
0 2 4 6 8 10 12 14
Storage Time (days)
Figure 21. Yield force F y of anhydrous milkfat cooled at rates of 0.1C/min, 1C/min, and
a few Newtons, at each of the cooling rates. This indicates that the samples may be
hardening slightly; however, the yield force does not significantly increase
p > 0:05 during the 14-day storage period.
7. ASSESSING THE VALIDITY OF THE MODEL: CORRELATING

EXPERIMENTALLY DETERMINED PARAMETERS
In order to further demonstrate the existence of inter-relationships between para-

meters that influence macroscopic properties, linear correlations were performed
on data obtained from the study of cooling rate and storage time (Table 4). As
the physical properties of a fat system are dictated primarily by lipid composition,
it is interesting to observe the effects of composition of other measured parameters.
For example, a significant correlation exists between the saturated fatty acid content
(SFAC) and the final SFC attained (Figure 22a). This agrees with expectations,
because the saturated FAs generally have higher melting points, and thus, they
are likely to crystallize and contribute to the measured SFC at any given tempera-
ture. Significant positive correlations also exist for relationships between SFAC and
55 20
A B
50
15
G' (MPa)
SFC (%)
45
10
40
5
35
2
r =0.905 r2=0.785
30 0
50 55 60 65 70 50 55 60 65 70
Saturated Fatty Acid Concentration Saturated Fatty Acid Concentration
(%) (%)
75
C
Yield Force (N)
50
25
r2=0.392
0
55 60 65 70
Saturated Fatty Acid Concentration
(%)
Figure 22. Linear correlations between saturated fatty acid content (%) and (A) solid fat content,
(B) storage modulus, and (C) yield force. Data shown represent all points collected at all cooling
rates and storage times.
TABLE 4. First Order Linear Regression Coefficients (2) for Various Physical Properties of Anhydrous Milkfat. Correlations were Performed
After Pooling all Results from Experiments Involving Various Cooling Rates and Storage Times at 5 C.
SFAC SFC Df Db Dr MEA l G0 G00 tan (d) Yield Force
SFC 0.91 ()
Df 0.37 () 0.80 ()
Db 0.62 () 0.89 () 0.78 ()
Dr na 0.78 () 0.70 () 0.90 ()
MEA 0.31 () 0.98 () 0.77 () 0.90 () 0.77 ()
l na 0.86 () 0.54 () 0.60 () 0.54 () 0.90 ()
G0 0.78 () 0.87 () 0.19 () 0.11 () 0.04 () 0.15 () 0.11 ()
G00 0.71 () 0.57 () 0.80 () 0.92 () 0.87 () 0.76 () 0.50 () 0.90 ()
tan (d) 0.08 () 0.23 () 0.81 () 0.95 () 0.87 () 0.79 () 0.52 () 0.34 () 0.10 ()
Yield Force 0.39 () 0.77 () 0.87 () 0.97 () 0.90 () 0.86 () 0.61 () 0.61 () 0.02 () 0.63 ()
Work 0.58 () 0.81 () 0.66 () 0.48 () 0.28 () 0.53 () 0.36 () 0.63 () 0.14 () 0.31 () 0.76 ()
Strain 0.04 () 0.02 () 0.58 () 0.72 () 0.73 () 0.61 () 0.44 () 0.01 () 0.29 () 0.52 () 0.23 ()
Legend
na comparisons unachievable due to method of calculation
bold p<0.05
SFAC saturated fatty acid concentration (%)
() or () indicates positive or negative correlation
30 20
Element Area (m2)
A r2=0.983 B
25
Microstructural
15
G' (MPa)
20
15 10
10
5
5
r2=0.874
0 0
45.0 47.5 50.0 52.5 55.0 25 30 35 40 45 50 55
SFC (%) SFC (%)
75
C
Yield Force (N)
50
25
r2=0.773
0
30 35 40 45 50 55
SFC (%)
Figure 23. Linear correlations between solid fat content and (A) microstructural element size,
(B) storage modulus, and (C) yield force. Data shown represent all points collected at all cooling
rates and storage times.
measures of apparent hardness. SFAC is linearly correlated with both the storage
modulus G0 and yield force (Fy) (Figure 22b,c). Higher saturate levels, their sub-
sequent crystallization, and contribution to the solids content of the network trans-
late into increases in the materials hardness.
SFC, which is commonly used as a primary indicator of hardness, shows a high
positive correlation to both the storage modulus G0 and yield force (Fy) in milkfat
(Figure 23a,b). This relationship forms the basis of methods used to determine the
fractal dimension of fat crystal networks. Evidence of a relationship between SFC
and microstructural element size was also established (Figure 23c). Smaller crystal
aggregates are a result of higher cooling rates, which also dictate the final SFC. This
relationship between factors further exemplifies the interdependence of processing
conditions, microstructure, and SFC.
Correlations shown in Figure 24 highlight the relationships between various
microstructural parameters and yield force. As stated previously, microstructure
strongly influences macroscopic hardness. Networks consisting of smaller crystal
aggregates are generally harder than those made up of larger microstructures (Fig-
ure 24a). Correspondingly, networks consisting of particle distributions that are
more disorderly (lower Df and Dr) and more space filling (higher Db) are also harder,
as indicated by higher yield force values (Figure 24.b, c, and d).
ASSESSING THE VALIDITY OF THE MODEL: CORRELATING 157
80 80
A r2=0.865 B r2=0.865
Yield Force (N)
Yield Force (N)

60 60
40 40
20 20
0 0
0 5 10 15 20 25 30 1.85 1.90 1.95 2.00
Microstructural Element Area Fractal Dimension (Df)
(m2)
80 80
C D
Yield Force (N)
Yield Force (N)
60 60
40 40
20 20
2
r =0.969 r2=0.897
0 0
1.70 1.75 1.80 1.85 1.90 2.4 2.5 2.6 2.7 2.8 2.9
Fractal Dimension (Db) Fractal Dimension (Dr)
Figure 24. Linear correlations between yield force and (A) microstructural element area, (B)
fractal dimension by particle-counting. (C) fractal dimension by box-counting, and (D) fractal
dimension by rheology. Data shown represent all points collected at all cooling rates and storage
times.
7.1. Summary of the Effect of Processing Conditions on the Physical

Properties of AMF
The schematic illustration of the combined effects of the majority of the physical
properties examined can be appreciated in Figure 25.
Figure 25a illustrates the effect of rapid cooling (5 C/min) on the physical prop-
erties of AMF. The diagram depicts a large number of small, randomly packed crys-
tal structures. These structures are highly space filling and therefore have a larger
number of particleparticle interactions. The result is a crystal network with an
even higher SFC (5153%), high Db (highly space filling), and low Dr and Df
values, resulting in an even harder fat crystal network.
Figure 25b illustrates the effects of the midrange cooling rate (1 C/min) on the
physical properties of AMF. This diagram shows a combination of both small- and
medium-sized crystal structures that are more randomly distributed in space. The
increased number of smaller crystal structures lead to increases in the number of
particleparticle interactions and a decrease in void space. This resulted in a system
with a higher SFC (5153%), higher Db (more space filling), and decreased Dr and
Df (less packing order), which all translated into a harder fat.
Figure 25.
Finally, Figure 25c depicts an illustrative model for the effects of the slow
(0.1 C/min) cooling rate on the physical properties of AMF. As shown, the crystal
network is composed of a small number of large highly ordered crystal structures
that are homogeneously distributed in space. The decrease in crystal structure num-
bers and increased size lends results in fewer particleparticle interactions and a
larger amount of void space. The final result was a lower SFC (46 48%), lower
values for Db (less space filling crystal mass), and increases in Dr and Df (increased
structural packing order), which all resulted in a softer fat (50% lower yield force
than samples cooled at 1 C/min and 5 C/min).
Therefore, through experimentation using AMF as a model fat, it has been
shown that processing conditions affect the underlying physical properties of a
fat, and it is these physical properties and their interrelationships that ultimately
effect the final macroscopic properties of a fat crystal network.
ACKNOWLEDGMENTS
The authors acknowledge the financial support of the Natural Sciences and Engi-
neering Research Council of Canada, the Ontario Ministry of Agriculture and Food,
the Canadian Foundation for Innovation, and the Ontrario Innovation Trust.
REFERENCES
1. M. Kalab, J. Dairy Sci., 68, 32343248.

2. A. L. Halmos, Food Australia, 49(4), 169173 (1997).
3. W. D. de Bruikne and A. Bot, in A. J. Rosenthal, ed., Food Texture Measurement and
Perception, Aspen, Gaithersburg, Maryland, 1999 pp. 185227.
4. A. J. Wright, M. G. Scanlon, R. W. Hartel, and A. G. Marangoni, J. Food Sci., 66(8), 1056
1071 (2001).
5. D. Rousseau and A. G. Marangoni, Food Res. Int., 31, 381388 (1998).
6. J. H. Prentice, J. Texture Studies, 3, 415458 (1972).
REFERENCES 159
7. L. de Man, J. M. deMan, and B. Blackman, J. Amer. Oil Chem. Soc., 66, 128132 (1987).
8. J. M. deMan and A. M. Beers, J. Texture Studies, 18, 303318 (1987).
9. D. Rousseau, A. R. Hill, and A. G. Marangoni, J. Amer. Oil Chem. Soc., 73, 983989 (1996).
10. A. M. Fearon and D. E. Johnston, J. Food Quality, 12, 2338 (1989).
11. A. J. Haighton, J. Amer. Oil Chem. Soc., 36, 345348 (1959).
12. K. R. Davey and P. N. Jones, J. Texture Studies, 16, 7584 (1985).
13. M. Kawanari, D. D. Hamann, K. R. Swartzel, and A. P. Hansen, J. Texture Studies, 12,
483505 (1981).
14. H. Rohm and F. Ulberth, J. Texture Studies, 20, 409418.
15. V. R. Huebner and L. C. Thomsen, Amer. Dairy Sci., 40, 834838 (1957).
16. C. Pompei, E. Casiraghi, M. Lucisano, and B. Zanoni, J. Food Sci., 53, 597602 (1988).
17. H. T. Lawless and H. Heymann, in Sensory Evaluation of Food, Aspen, Gaithersburg,
Maryland, 1999, pp. 379405.
18. B. K. Mortensen and H. Danmark, Milchwissenschaft 36, 393395 (1982).
19. H. Corey, CRC Crit. Rev. Food Technol., May, 161198 (1970).
20. A. S. Szczesniak, Food Technol., October, 5258 (1966).
21. D. M. Small, in D. J. Hanahan, ed., The Physical Chemistry of Lipids, from Alkanes to
Phospholipids, Vol. 4, Plenum Press, New York, 1966, pp. 475522.
22. A. Kuksis, L. Marai, and J. J. Myher, J. Amer. Oil Chem. Soc., 50, 193201 (1973).
23. J. Gresti, M Bugaut, G. Maniongui, and J. Bezard, J. Dairy Sci., 76, 18501869 (1993).
24. B. Breitschuh and E. J. Windhab, J. Amer. Oil Chem. Soc., 75, 897904 (1998).
25. H. E. Swaisgood, in O. R. Fennema, Food Chemistry, 3rd ed., 1996, pp. 841878.
26. W. W. Christie, in P. F. Fox, ed., Advanced Dairy Chemistry, Vol. 2, 2nd ed., Chapman &
Hall, London, 1995, pp. 136.
27. A. J. Wright, S. E. McGauley, S. S. Narine, R. W. Lencki, and A. G. Marangoni, J. Agric.
Food Chem., 48, 10331040 (2000).
28. A. G. Marangoni and R. W. Lencki, J. Agric. Food Chem., 46, 38793884 (1998).
29. D. Rousseau and A. G. Marangoni, J. Agric. Food Chem., 46, 23682374 (1998).
30. D. Rousseau and A. G. Marangoni, J. Agric. Food Chem., 46, 23752381 (1998).
31. A. G. Marangoni and D. Rousseau, J. Amer. Oil Chem. Soc. 75, 16331636 (1998).
32. A. G. Marangoni and D. Rousseau, Food Res. Int., 31, 595599 (1998).
33. A. G. F. Stapley, H. Tewkesbury, and P. J. Fryer, J. Amer. Oil Chem. Soc., 6, 677685 (1999).
34. E. ten Groetenhuis, G. A. van Aken, K. F. van Malssen, and H. Schenk, J. Amer. Oil Chem.
Soc., 76, 10311039 (1999).
35. M. L. Herrera and R. W. Hartel, J. Amer. Oil Chem. Soc., 77, 11771187 (2000).
36. M. L. Herrera and R. W. Hartel, J. Amer. Oil Chem. Soc. 77, 11971204 (2000).
37. M. L. Herrera and R. W. Hartel, J. Amer. Oil Chem. Soc. 77, 11891195 (2000).
38. J. Garside, in J. M. V. Blanshard and P. Lillford, eds., Food Structure and Behaviour,
Academic Press, London, 1987, pp. 3549.
39. D. S. Grall and R. W. Hartel, J. Amer. Oil Chem. Soc., 69, 741747 (1992).
40. P. Walstra, T. van Vliet, and W. Kloek, P. F. Fox, eds., Advanced Dairy Chemistry, Vol. 2,
2nd ed., Chapman & Hall, London, 1995, pp. 179211.
41. A. G. Marangoni and D. Rousseau, N. Widlak, ed., Physical Properties of Fats, Oils and
Emulsifiers, AOCS Press, Champaign, Illinois, 1999, pp. 96111.
42. B. L. Madison and R. C. Hill, J. Amer. Oil Chem. Soc., 55, 328331 (1978).
43. J. van Duynhoven, G. J. Goudappel, M. C. M Gribnau, and V. K. S. Shukla, Inform, 105,
479484 (1999).
44. L. van Putte, I. Bruker minispec documentation.
45. W. W. Nawar, in O. R. Fennema, ed., Food Chemistry, 3rd ed., 1996, pp. 226319.
46. J. M. deMan and L. deMan, in N. Widlak, R. Hartel, S. Narine, eds., Crystallization and
Solidification Properties of Lipids, AOCS Press, Champaign, Illinois, 2001, 225235.
47. C. W. Hoerr, J. Amer. Oil Chem. Soc., 37, 539546 (1960).
48. P. Walstra, T. J. Geurts, A. Noomen, A. Jellema, and M. A. J. A. van Boekel, in P. Walstra,
ed., Dairy Technology: Principles of Milk Properties and Processes, Marcel Dekker,
New York, 1999, pp. 5070.
49. S. S. Narine and A. G. Marangoni, Food Res. Int., 32, 227248 (1999).
50. S. S. Narine and A. G. Marangoni, Adv. Food Nutr. Res., 44, 33145 (2002).
51. A. G. Marangoni, Trends Food Sci. Technol., 13, 3747 (2002).
52. R. Vreeker, L. L. Hoekstra, D. C. den Boer, and W. G. M. Agterof, Food Hydrocolloids, 6,
423435 (1992).
53. A. G. Marangoni and D. Rousseau, J. Amer. Oil Chem. Soc., 73, 991994 (1996).
54. S. S. Narine and A. G. Marangoni, Phys. Rev. E, 59, 19081920 (1999).
55. W. H. Shih, W. Y. Shih, S. I. Kim, J. Liu, and I. A. Aksay, Phys. Rev. A, 42, 47724779
(1990).
56. S. S. Narine and A. G. Marangoni, Phys. Rev. E, 60, 69917000 (1999).
57. A. G. Marangoni, Phys. Rev. B, 62, 1395113955 (2000).
58. A. G. Marangoni and M. A. Rogers, Appl. Phys. Lett., 82, 13 (2003).
59. T. Hagiwara, H. Kumagi, and K. Nakamura, Food Hydrocolloids, 12, 2936 (1998).
60. M. Tather and A. Erdem-Senatalar, J. Phys. Chem. B, 103, 43604365.
61. K. Solecka-Cermakova, B. Vlckova, and F. Lednicky, J. Phys. Chem., 100, 49544960
(1996).
62. A. Y. Kim and J. C. Berg, Langmuir, 16, 21012104 (2000).
63. G. Joss. 2002. Available: http://list.nih.gov/cgi-bin/wa?A2ind0101&Limagej&D
0&P5222.
64. J. C. Russ, in D. B. Williams, A. R. Pelton, and R. Gronsky, eds., Images of Materials,
Oxford University Press, New York, 1991, 338373.
65. Trusoft International Incorporated. (2000). Available: http://www.trusoft.com.
66. H. Rohm and K. H. Weidinger, J. Texture Studies, 24, 157172 (1993).
67. J. F. Steffe, in Rheological Behaviour of Butter at Small Deformations, 2nd ed., Freeman
Press, East Lansing, Michigan, 1996, pp. 294349.
68. M. A. Rao, in M. A. Rao, ed., Rheology of Fluid and Semisolid Foods: Principles and
Applications, Chapman & Hall, Gaithersburg, Maryland, 1999, pp. 59152.
69. S. S. Narine and A. G. Marangoni, Lebensmittel Wissenschaft und Technologie, 34, 3340
(2001).
5
Animal Fats
Michael J. Haas
Eastern Regional Research Center, Agricultural Research Service
Wyndmoor, Pennsylvania
1. INTRODUCTION
The use of animal fats by humans may well predate civilization. As the depot fats of
animals are readily noticed during the butchering of a slaughtered animal, are easily
harvested, and are available in the absence of plant domestication and the adoption
of established agriculture, it is probable that animal fats were the first lipids
employed as industrial and as distinct edible lipids by humans. This is evidenced
by the fact that the paints used in prehistoric cave paintings were animal fat-based,
as were the fuels in the lamps that illuminated the cave artists at their work. Despite
a tremendous diversification to include other lipid types over the intervening cen-
turies, animal fats still play a prominent role in our diets, industry, and commerce.
Lipids are biological materials that are insoluble in water but soluble in
nonpolar organic solvents. Here the term will be used interchangeably with
acylglycerol, the primary component of animal lipids. These are classified as
either fats or oils. The former are predominantly solid at room temperature
(24 C), and the latter are liquid. The depot lipids of animals are generally fats.
The major animal fats (also termed meat fats) of contemporary commerce are
produced from pigs (Sus scrofa), in which case they are termed lard and rendered
pork fat, from the fat of cattle (Bos taurus) or sheep (Ovis aries) and termed tallow,
or from poultry (primarily chickens, Gallus gallus) and termed poultry fat. Tallow
161
162 ANIMAL FATS
from domestic cattle is known as beef tallow, whereas that from sheep is termed
mutton tallow.
Animal depot lipids are used in edible applications, sometimes being consumed
directly but more often in such applications as baking, cooking, and deep fat frying.
They are also used in industrial applications, primarily in soap production, as an
energy and nutrient source in animal feeds, in lubricants, and as a source of indus-
trial fatty acids. These applications are discussed in detail elsewhere in this series.
This chapter will consider contemporary aspects of the classification, composi-
tion, properties, production, processing, and consumption of the depot lipids of land
animals. For a more detailed discussion, the reader is referred to the recent compre-
hensive animal fats text (1). Additional discussions of the material covered here,
and related topics, are found in excellent contemporary texts (24). Some consid-
eration of greases will also be included here, because these are largely handled by
and traded within the animal fats sector of the lipid industry. Discussions of milk-
fats and fish oils can be found elsewhere in this series.
In addition to the citations accompanying this chapter, the reader will find
numerous valuable websites on the Internet. These can be located by the use of
search engines such as www.google.com and www.dogpile.com.
2. SOURCES, FATTY ACID CONTENT,

AND ACYLGLYCEROL STRUCTURE
2.1. Acylglycerol Components

The major fat depots of animals include subcutaneous fat (located under the skin
and overlying superficial muscles) and intermuscular fat (located between muscles).
Appreciable amounts of fat also are deposited in the abdominal cavity and other
internal sites. The distribution of fat between different sites varies somewhat
with animal species, breed, and degree of finish.
On the larger species, some of the fat deposits, for example, those located around
the kidneys, heart, intestines, and greater omentum, are stripped from the animal
when it is slaughtered. These are called the killing-floor or killing fats. Additional
fat separated from the carcass when it is cut apart to give wholesale or retail cuts, or
for processed meat products, is called cutting-floor or cutting fat.
White, yellow, and brown greases are also defined and traded in the industry (5).
White grease is derived primarily from the rendering of pork offal. Yellow grease is
composed largely of spent deep fat fryer lipids. Brown grease can be any lipid that
does not meet the specifications for yellow grease. All are specified to contain at
least 90% lipid. Poultry fat is composed of 100% poultry offal from slaughter
operations. Formal subtypes or classifications are not identified. Individual purcha-
sers may make specifications as to its quality or content.
Animal and plant depot lipids consist primarily of triacylglycerols: triesters of
glycerol and three fatty acids. Lipids differ from species to species with regard to
the types and amounts of fatty acids they contain. Table 1 presents the fatty acid
TABLE 1. Typical Fatty Acid Compositions: Subcutaneous Adipose Tissue of Selected Animals, Greases, and Selected Vegetable Oils.
Trivial Beef Mutton Chicken Choice Yellow

Fatty Acid Name Tallowa Tallowa Larda Fata White Greaseb Greaseb Soyc Rapec
10:0 capric 00.1 0.1.2 0.1 0.2 4

12:0 lauric 0.1 0.10.5 0.1 0.2
14:0 myristic 2.74.8 2.84.9 1.41.7 1.3 1.9 1.9 0.1 0.1
14:1 myristoleic 0.82.5 0.70.8 00.1 0.2 0 0
16:0 palmitic 20.928.9 19.521.3 23.128.3 23.2 21.5 16.2 11 2.8
16:1 palmitoleic 2.39.1 1.42.3 1.83.3 6.5 5.7 2.5 0.1 0.2
17:0 margaric 1 1 0.5 0.3 0 0
18:0 stearic 7.026.5 17.628.9 11.724 6.4 14.9 10.5 4 1.3
(19)d (14)d
18:1 oleic 30.448.0 33.240.4 29.745.3 41.6 41.1 47.5 23.4 23.8
(43)d (44)d
18:2 linoleic 0.61.8 1.23.4 8.112.6 18.9 11.6 17.5 53.2 14.6
18:3 linolenic 0.30.7 1.41.9 0.71.2 1.3 0.4 1.9 7.8 7.3
20:0 arachidic tr.0.9 tr.0.3 0.20.3 1.8 for all > 1.0 for all > 0.3 0.7
C20 C20
20:1 gadoleic 0.31.7 0.20.3 0.81.3 12.1
20:2 eicosadienoic 00.1 0 0.30.5 0.6
20:4 arachidonic
22:0 behenic 00.1 0 00.4 0.1 0.4
22:1 erucic 0tr. 0 tr.0.1 0 34.8
24:0 lignoceric 0tr. 0 00.5 1
24:1 nervonic 0 0 00.6
Trans-fatty
acids 1.36.6 11.014.6 1.11.4
a
Data from (1)
b
See (6).
c
See (7).
d
Typical value for U.S. product: From (8).
164 ANIMAL FATS
compositions of the major industrially important animal lipids, with data for some
plant lipids for comparison. Data are also included for white and yellow greases.
In countries such as the United States where animal fats are no longer used
in substantial quantities for deep fat frying, yellow grease consists largely of
hydrogenated vegetable oil, whereas elsewhere it may contain a substantial portion
of beef tallow. Thus, the fatty acid content and physical properties of yellow grease
may depend on country or region of origin and can change over time as food indus-
try practices change. Note that there is a wide range of values stated for some
entries in Table 1, such as a stearic acid (18 carbons, no double bonds, i.e., 18:0)
content between 7% and 27% for beef tallow. This reflects the fact that a number of
factors impact the composition of an animal fat, and thus any general statement of
composition must have a broad range.
Examination of Table 1 shows that the prevalent fatty acids in animal depot
lipids are either 16 or 18 carbons in length and are either fully saturated or contain
one or two double bonds. Animal lipids generally contain a higher proportion of
saturated fatty acids than do the lipids of temperate zone plants.
The physiological role of lipids is to serve as a carbon and energy reserve. To be
biochemically accessible for these purposes, it is best that a lipid be liquid, or at
least semiliquid. Thus, at body temperatures, depot fats are semisolids. They soli-
dify when cooled to room temperature. The largest factor contributing to the freez-
ing or melting temperature of an acylglycerol is its fatty acid composition,
specifically with regard to the chain lengths and number of double bonds. A double
bond in a fatty acid has the effect of introducing a kink into the linear run of carbons
(Figure 1). This kinking interferes with the tendency of the acyl chains of the fatty
acids in the lipid to align and pack with one another. As a consequence, more
energy must be removed from the system to induce crystallization. This is mani-
fested as a reduced freezing or melting point relative to saturated fatty acids, which
can adopt straighter chain configurations, pack more readily, and crystallize (freeze)
at higher temperatures. Differences in the content of unsaturated fatty acids play the
largest role in determining melting point differences between various animal and
temperate zone vegetable lipids. Thus, it is the relatively low content of unsaturated
fatty acids that renders fats solid at room temperature, whereas the liquidity of oils
is due to their higher content of unsaturates. Note also (Figure 1) that the extent of
kinking introduced into a fatty acyl chain by a double bond depends on the config-
uration of that bond, with cis-bonds introducing a considerable kink and trans-
bonds barely perturbing the linear run of the carbon chain. This leads to substan-
tially different physical and biochemical properties for the two types of unsaturated
fatty acids. The great preponderance of naturally synthesized unsaturated fatty
acids contain cis-double bonds. The industrial hydrogenation of oils to produce
more fully saturated lipids can result in the production of a substantial proportion
of trans-double bonds. Differences in the physiologic effects of cis-vs. trans-fatty
acids in dietary lipids have led to concerns regarding consumption of the latter, as
discussed below (see Recent Developments).
It is energetically more costly to synthesize an unsaturated fatty acid than a satu-
rated one of the same chain length. The body fats of warm-blooded animals need to
SOURCES, FATTY ACID CONTENT, AND ACYLGLYCEROL STRUCTURE 165
H H
H H
C C
H H
C C
H H
C C
H H
C C
H H
H H
cis
H H H H
H H H
C C C C
C C C C
H H H
H H H H
trans
Figure 1. The configurations of cis- and trans-carboncarbon double bonds. (Used with the kind
permission of the Institute of Shortening and Edible Oils, Inc. Washington, D.C.)
contain only a sufficient proportion of unsaturated fatty acids to render them semi-
solid at body temperature, which is approximately constant and generally higher
than ambient temperature. The resulting relatively high proportion of saturated fatty
acids causes these lipids to be fats: to have melting points above room temperature.
As plants have no notable mechanism of temperature control, their lipids must stay
fluid over the typical range of temperatures encountered in the field. Plants thus
synthesize and incorporate a higher proportion of unsaturated fatty acids, typically
oleic (18:1), linoleic (18:2), and linolenic (18:3) acids into their lipids, allowing
them to remain fluid at relatively low ambient temperatures. These tendencies
are so general as to allow the term fat alone to be used in some contexts to refer
to animal lipids, whereas oil is used to refer to plant lipids. Tropical plants
experience higher ambient temperatures than those of temperate climates, and
thus their lipids generally contain a higher proportion of saturated fatty acids.
Among animals, another feature that bears on the fatty acid content of the depot
lipids is whether an animal bears a rumen. The rumen harbors a dense, diverse, and
metabolically very active microbial community. By conducting the biochemical
hydrogenation of unsaturated fatty acids, it is this community that modifies the
unsaturate-rich lipid diet of a grazing animal to produce the saturated fatty acids
that are incorporated into the body fat. The biohydrogenation pathway has several
intermediates, and these can be substrates for other reactions within the rumen or
the whole animal. As a result, the depot fats of ruminants have a very diverse fatty
acid composition. Tallow contains hundreds of different fatty acid structures, most
present only in small or trace amounts. The saturated fatty acids of ruminant acyl-
glycerols contain normal and methyl-branched components; the latter may be of
the iso (terminally branched) or anteiso (subterminally branched) form. Odd-and
even-numbered carbon chain lengths are found, and both geometrical and positional
isomers of the unsaturated fatty acids usually are present. This presence of
166 ANIMAL FATS
trans-and positional unsaturated fatty acid isomers is characteristic of ruminants

and animals with ruminant-like digestive systems.
The degree of rumen-mediated fatty acid modification varies from species to
species. For example, the biohydrogenation of dietary unsaturates is greater in
sheep than in cattle, and thus mutton tallow contains 5% to 10% more stearic
acid, and a correspondingly lower amount of oleic acid, than beef tallow. Table 1
illustrates this trend, although it is somewhat obscured by the necessarily wide
ranges of values reported.
In general, the concentration of saturated fatty acids, especially stearic, is higher
in ruminant than in nonruminant fat. Additionally, the polyunsaturates, especially
linoleic acid, the most prevalent natural polyunsaturate, are lower in ruminant fat
than in that from monogastric (i.e., nonruminant) species. Thus, the fat of cattle and
sheep is, in general, firmer than that from pigs.
The fatty acid structural diversity seen in the fats of rumen-containing animals is
not found in plant lipids. Extensive microbial modification of dietary fatty acids is
also absent in monogastric species (e.g., pig, poultry). Consequently, the fatty acid
content of their lipids is not only simpler than that of ruminants, but also more clo-
sely mirrors the dietary fatty acid intake, displays a greater sensitivity to alterations
in dietary fatty acid content, and can be intentionally manipulated to a greater
degree by adjustments to the diet. Enser (9) has summarized work regarding the
relationship between dietary and depot lipid fatty acid contents, among other things,
pointing out a linear correlation between dietary and depot linoleic acid levels in
the pig. Table 2 illustrates the ability of dietary unsaturated oils to increase the
degree of unsaturation of pork fat. The fatty acid content of the pork fat reflected
that of the dietary lipids. In contrast, the fatty acid composition of beef fat varied
TABLE 2. Influence of Diet on the Fatty Acid Compositions (% of total) of Pork Fat
and Beef Adipose Tissue.
Pork Fata Pork Fatb Beef Fatb

Corn & Rapeseed &
Corn & Soy Cottonseed Cottonseed
Fatty acid Tallow Dietc Soy Oilc Meal Canola Oil Meal Oil
14:0 1 0.8 1.3 0.5 3.9 3.6

16:0 26.6 22.1 25.6 10.8 26.7 24.3
16:1 4.1 2.5 0.9 0.4 2 2
18:0 12.1 11.3 12.9 4.3 20.2 20.5
18:1 40.5 33.2 46.9 56.1 41.2 43
18:2 11.2 24.4 11.5 21.5 4.8 5.5
18:3 4 4.9 0.9 6.5
20:4 0.3 0.5
22:6 0.2 0.2
Total
saturates 39.6 34.2 39.8 15.6 50.8 48.4
a
See (10).
b
See (11).
c
Diets contained 3% tallow or soy oil.
SOURCES, FATTY ACID CONTENT, AND ACYLGLYCEROL STRUCTURE 167
little between animals on a corn-cottonseed diet and those on a diet also containing
rapeseed oil, which is enriched in erucic acid (22:1) (Table 2). It has been reported
that the back fat of pigs on a diet containing soy oil had an elevated content of lino-
leic acid (the most prevalent fatty acid in soy oil), although the altered fat did not
pose any problems with regard to carcass appearance or softness during processing
(12). In other studies, pigs were fed a cornsoybean meal diet with added tallow,
safflower oil, or a combination of tallow and safflower oil (13). The increased levels
of dietary safflower oil resulted in a decrease in the contents of stearic and oleic
acids and an increase in linoleic acid (the predominant acid in safflower oil),
20:2, and 20:3 in the subcutaneous fat.
For ruminant fat to become directly responsive to dietary unsaturated fats, it is
necessary to protect the lipids against saturation by rumen microorganisms. The
alteration of the lipid content of mutton by the feeding of such protected oil supple-
ments has been described (14). Also, it has been shown that a diet of extruded
soybeans increased the linoleic acid and linolenic acid contents of steer adipose
tissue (15).
A number of factors other than diet can influence the fatty acid composition of
the depot lipids of a given species. These include genetic and sex effects (16, 17).
Physical environment also plays a role. For example, in sheep, colder temperatures
result in softer body fats with lower melting points and higher iodine numbers (18).
In general, diet has a more marked effect on fat quality than do breed or sex,
especially in nonruminants, which are susceptible to alteration of tissue fatty acids
by dietary modification.
Location within the body also influences the degree of unsaturation in a fat. The
temperature of a warm-blooded animal is not constant throughout the body. Fats
located near the skin experience colder temperatures than those within a carcass.
To remain semifluid and thus metabolically accessible in the face of this thermal
challenge, fats near the surface tend to have lower melting points and be softer,
traits resulting from an elevated content of unsaturated fatty acids relative to fats
from the interior of the carcass. For example, the saturated fatty acid content of
beef tallow located just below the hide has been reported to be 48.7% of the total
fatty acid content, whereas that from the kidney region, deep within the carcass, has
a saturated fatty acid content of 57.9% (19). In general, the fatty deposits increase
in hardness from surface subcutaneous locations through the inter- and intramuscu-
lar fat to deep abdominal and kidney fats in cattle, sheep, and pigs. Thus, the inter-
nal fats from these species, especially those surrounding the kidney, tend to be
harder than those from fats near the surface of the carcass. Products made from
kidney fat will be firmer. They will also have better flavor stability, because rancid
flavors are the result of oxidation of the double bonds of unsaturated fatty acids.
Although relevant in some food applications, this feature is of restricted overall
importance because it is rare for animal fat to be segregated on the basis of anato-
mical origin.
Whether for food or nonedible industrial applications, the choice of raw material
lipid is dependent on a match between the physical properties of the lipid and the
desired performance properties. Physical and performance properties are largely a
168 ANIMAL FATS
function of fatty acid composition. Thus, for example, the higher saturated fatty
acid contents of animal fats, by rendering them solid at room temperature, makes
them desired feedstocks for the production of bar soaps. It also affects their perfor-
mance in baking, and for decades made them the optimal choice for such applica-
tions. Also, because their lower content of polyunsaturated fatty acids confers
greater oxidative stability, animal fats are suggested for oxidation-prone applica-
tions such as deep fat frying and as lubricants in high-speed metal working.
2.2. Nonacylglycerol Components

Animal fats consist mainly of triacylglycerols, containing only minor amounts
(<0.05%) of other compounds such as phospholipids, tocopherols, and carotenoids
(9). However, one minor component of considerable note is cholesterol. Typical
cholesterol levels for animal fats of industrial importance are on the order of
8501100 mg/kg, with poultry fats at the lower end of this spectrum. These levels
are at least an order of magnitude greater than those for vegetable oils. Cholesterol
is a necessary component of the human body. It is both synthesized de novo and
obtained in the diet. In most individuals, the consumption of moderate amounts
of cholesterol poses no health risk, as it is balanced by a reduction in endogenous
synthesis. However, concern over the role of cholesterol as a dominant factor in the
genesis of atheroschlerosis (20), a major cause of mortality in developed countries,
has led to government advisories to restrict the intake of cholesterol (and of satu-
rated fatty acids, which exacerbate its atheroschlerotic effects) (21, 22). As a result,
since the 1980s, there has been a substantial replacement of animal fats by vegeta-
ble oils in many food applications, greatly reducing the utilization of animal fats in
edible applications. To perform properly, the vegetable oils are often hydrogenated
to convert their unsaturated fatty acids to saturates, thereby increasing solid fat con-
tent, conferring desirable performance properties, and increasing oxidative stability.
Methods are available to remove cholesterol from animal fats (23), and a line of
low cholesterol edible lipids, termed Appetize, was marketed in the United States
in the 1990s (24). The product consisted of 7090% tallow whose cholesterol had
been reduced by steam distillation to only 8 mg/100 g. Despite the fine rationale
underlying its production, however, high production costs resulted in the withdra-
wal of the Appetize line from the market. Supercritical fluid extraction can also be
used to reduce the cholesterol content of animal fats (25), although to this authors
knowledge, this approach has not been commercially implemented.
3. ACYLGLYCEROL STRUCTURE AND ITS RELATIONSHIP

TO FUNCTIONALITY AND USE
For nonedible uses, fatty acid content, availability, and price play the largest roles in
determining which lipid is employed in a particular application. In cases where both
animal fats and vegetable oils are able to perform a desired function, animal fats
can be attractive raw materials because their bulk prices have historically been
ACYLGLYCEROL STRUCTURE AND ITS RELATIONSHIP TO FUNCTIONALITY 169
substantially below those of the least expensive vegetable oils. For example, in the
United States, edible tallow prices are typically 40% to 60% of that of soybean oil.
For edible uses, however, aspects of lipid structure other than simply fatty acid
content come into consideration. Chief among these are the pattern of distribution
of fatty acids on the glycerol chain, because the melting profile, plastic range, and
other properties of a fat are affected by this feature, known as acylglycerol struc-
ture. These properties affect performance parameters such as mouthfeel and melt-
ing range. The carbons of the glycerol molecule are not chemically equivalent.
They are identified, based on position and stereochemical numbering conventions,
as sn-1, -2, and -3, with the -2 position being the secondary hydroxyl in the middle
position of the molecule. The distribution of fatty acids among these three carbons
in natural fats (and oils as well) is not random, and varies with species and fatty
acid. Thus, a lipid will contain an array of acylglycerol structures including those
consisting entirely of saturated fatty acids (SSS); those containing saturated, mono-
unsaturated (M), and polyunsaturated (U) fatty acids (SSM, SSU, UM S, etc.) and
all-unsaturated species (MMM, UMU) (Table 3). As expected, a lipid with a higher
content of unsaturated fatty acids will have a higher amount of multiply unsaturated
(polyunsaturated) triacylglycerols (Table 3). Within these groups is further species-
dependent diversity, e.g., content of SSU vs. SUS (Table 3). Recent advances in
mass spectral methods have greatly facilitated this speciation of natural lipids
(2831). Differences in fatty acid content and distribution cause differences in
the physical properties of fats, and thus they can determine the uses of a fat. In
the first instance, fully saturated acylglycerols have high melting points, whereas
those of partially or fully unsaturated ones are much lower. (The difference in melt-
ing points between tristearin and trilinolenin is nearly 100 C.) The melting profile,
TABLE 3. Positional Distribution of Fatty Acids in Triacylglycerols of Animal Depot Fats.a
Fatty Acid, mol%

Fat Position 14:0 16:0 16:1 18:0 18:1 18:2 18:3
Pig 1 1 10 2 30 51 6
(outer back fat) 2 4 72 5 2 13 3
3 tr. 2 7 73 18
Cattle 1 4 41 6 17 20 4 1
(subcutaneous fat) 2 9 17 6 9 41 5 1
3 1 22 6 24 37 5 1
Sheep 1 1 35 2 47 4b
(perineal fat) 2 4 14 2 15 52b 5
3 3 16 1 42 26b 2
Chickenc 1 2 25 12 6 33 14 2
2 1 15 7 4 43 23 3
3 1 24 12 6 35 14 3
a
See (26).
b
Results for 18:1 cis isomers only. 18:1 trans was present in positions sn-1, sn-2 and sn-3 as 5, 2 and 6%
respectively.
c
See (27).
170 ANIMAL FATS
plastic range, and other properties of a fat are affected by the range of its acylgly-
cerol structures. For example, lard is plastic over only a narrow temperature range,
has a grainy texture, and does not cream well. This is in large part a result of its
fatty acid content and pattern of acylglycerol structures, especially to a high content
of acylglycerols with palmitic acid (16:0) in the sn-2 position (Table 3). In lard,
72% of the residues at this site are palmitic acid, whereas this fatty acid constitutes
only 10% of the residues at the sn-1 position and is found in only traces in the sn-3
position. In beef tallow (and sheep and chicken fat as well), palmitic acid is located
at all three positions of the triacylglycerol molecule: In the sn-2 position, 17 mol%
is palmitic; at the sn-1 position, palmitic acid makes up 41%; and in the sn-3 posi-
tion, 22% (Table 3). The greater uniformity of lard triacylglycerol structures leads
to a shorter plastic range, a sharper melting point, and larger crystals in the solid
phase. The latter are largely responsible for the poor creaming abilities and grainy
mouthfeel of lard. Although their fatty acid contents are not widely different, lard
and beef tallow do exhibit some substantial differences in triacylglycerol structure.
Thus, lard contains about 7 mol% trisaturates and 32 mol% disaturates, whereas
beef tallow contains 15 mol% trisaturated triacylglycerols and nearly 40 mol% di-
saturates. Lard has a greater proportion of triacylglycerols with two double bonds
than does beef tallow and about half as many mol% of triacylglycerols with three
double bonds. Differences such as these confer different physical properties on fats.
In contrast to the high proportion of saturated fatty acids in the sn-2 position of
lard, the tallows contain predominantly an unsaturated fatty acid (about 60% rela-
tive abundance), generally oleic (18:1), in this position. This has been exploited to
develop a method to detect the adulteration of beef fat with lard (32). This involves
first the isolation of the acylglycerols containing one saturated and two unsaturated
fatty acids, followed by determination of the fatty acid population at the sn-2 posi-
tion of this fraction. The sensitivity is reported to be sufficient to detect the presence
of 1% lard in the tallow.
Lard is unique among the common fats and oils in having a preponderance of
palmitate at the sn-2 position. This has led to a unique use of this lipid: Human
milkfat also contains primarily palmitate at the sn-2 position of its acylglycerols,
and lard has been employed in infant formulas in attempts to produce a material
more closely resembling human milk. Approaches taken in this work have included
the direct addition of lard (33, 34) as well as the enzymatic restructuring of its acyl-
glycerols. The latter approach employs lipases that specifically interesterify (see
below) the sn-1 and sn-3 positions of acylglycerols, thus retaining the palmitate
at the sn-2 position, to introduce fatty acids from soybean oil into the terminal
positions of the fat molecules, creating structures closely mimicking those in
human milkfat (35).
Texture is one property affected by the content and arrangement of fatty acids in
the acylglycerols of a lipid. When a melted lipid cools and crystalizes to form a
solid, its acylglycerols will generally adopt one of three predominant crystal lattice
forms, a, b, or b0, depending on acylglycerol content and the kinetics of cooling
(36, 37). This ability to adopt more than one crystal form is termed polymorphism.
Crystals of the b type are large and course and confer an undesirable grainy
QUALITY INDICATORS FOR EDIBLE FATS 171
mouthfeel. Crystals of the b0 form are smaller and result in a smoother, more desir-
able, mouthfeel. Tallow tends to adopt the b0 configuration. Lard has a relatively
large proportion (27%) of disaturated acylglycerols, mostly oleoyllpalmitoylstearin,
which has a propensity to crystalize in the b form. This gives lard a grainy texture
and poor creaming ability. Through interesterification (see Section 7.5) a more ran-
dom distribution of fatty acids can be achieved, yielding a product that crystalizes
in the b0 form, which is more desirable for such edible applications as margarines.
The physical structure of the b0 form is less stable than the b form, and can con-
vert to the latter over time. Margarines typically contain mixtures of vegetable oil
and solid fat of b0 form. The small crystal structure of the latter aids in keeping the
mixture plastic and uniform. The spontaneous conversion of the b0 - to a b-
dominated hard fat population over time results in replacement of the fine crystals
by larger coarse one. These are less effective at carrying the oil phase, and the result
can be separation, yielding an undesirable two phase liquid/solid product. Appro-
priate choices of feedstocks, processing conditions, and crystal stabilizers can
reduce the rate of product deterioration due to changes in crystal structure.
4. QUALITY INDICATORS FOR EDIBLE FATS
Raw fat is susceptible to defects caused by (1) oxidation of the double bonds of
fatty acids, which creates degradation products that confer undesirable odors and
flavors; (2) ester bond hydrolysis (lipolysis) by contaminating microbes, which
releases free fatty acids; and (3) the formation of off colors. Best quality is obtained
if contaminants such as blood and manure are kept out of the raw material, and if it
is then kept cold, processed within a few days of slaughter with adequate attention
to process controls, and handled and stored properly. If these conditions are not
achieved, deterioration can occur. Waste greases are especially subject to such
degradation, because they are typically stored for long periods of time without
refrigeration. Because the resulting chemical changes can impact performance
and acceptability, various analytical parameters and specifications have been devel-
oped to identify lipid quality. Some of these also provide basic information on the
properties of a lipid, and they can be used to assess its suitability for a given appli-
cation. Among the properties, analyses, and terms by which lipid samples are
characterized are as follows:
Color: Excessive color in a lipid sample can prevent its use in some
applications or necessitate the application of color reduction technologies.
Several methods for expressing the color of a fat or oil exist, but all rely on the
direct comparison of the color of a sample to that of a series of color
standards. For raw samples, especially of tallow and grease, which can be
deeply colored, this value is often measured by comparing a sample of filtered
liquid fat with a set of 26 color standards designated by the Fat Analysis
Committee (FAC) of the American Oil Chemists Society (AOCS), and
assigning a number from 1 (lightest) to 45 (darkest), sometimes referred to
172 ANIMAL FATS
as simply FAC. Conduct of this measurement is described in AOCS Official

Method Cc13a-43 (38). Color reduction, when necessary, is generally
achieved through the use of bleaching clays. In the United States, after
bleaching, color is typically measured and expressed in terms of the
Wesson color method, described in AOCS Official Method Cc13b-45 (39).
In most other countries, the color after bleaching is determined by means
of the Lovibond Method, AOCS Official Method Cc13e-92, which is the
accepted international standard for the measurement of color in animal
and vegetable fats and oils (40). Particularly for use in bar formulation
soaps for hand and body washing, white color is desired in the feedstock
lipids.
Free fatty acid content: A measure of the amount of acylglycerol hydrolysis.
Free fatty acids can reduce the palatability, acceptability, and performance of
a lipid and are therefore generally considered a negative trait. As with all
parameters described here, acceptable levels vary depending on application.
Iodine number or value (IV): A measure of the degree of unsaturation of a
lipid, as measured by its iodine absorption, a trait proportional to double bond
content. Expressed as the number of grams of iodine absorbed by 100 g of fat.
As fatty acid oxidation occurs at double bonds, a high IV can indicate a fat
sample that will have marginal oxidative stability.
Moisture, impurities, and unsaponifiables (MIU): A summation of the non-
acylglycerol materials in the product. Moisture is undesirable because it will
support microbial growth and facilitate lipid hydrolysis. Unsaponifiables are
any materials that will not saponify (form soap) when incubated with sodium
hydroxide. These include sterols, pigments, and hydrocarbons. These are
natural components of both animal and plant lipids, but excessive amounts can
indicate a sample that will not perform comparably with a sample richer in
acylglycerols, and they can indicate adulteration or contamination by petro-
leum products.
Peroxide value (PV, also referred to as initial peroxide): The presence of
fatty acid hydroperoxides, formed by the oxidative degradation of fatty acids,
is a measure of oxidative abuse and degradation of the lipid. Products
generated by hydroperoxide degradation will confer a rancid note in edible
applications.
Refined and bleached (R&B) color: A measure of the amount of red color in
the rendered fat is an indicator of the quality of both the starting material and
the rendering techniques. This value is measured using the Lovibond
5.25-inch scale according to procedures described in AOCS Official Method
Cc 13b-43 (39). The result is referred to as the AOCS Wesson color, the
AOCS Lovibond color, or simply Lovibond color. The lower the value, the
less colored the sample.
Saponification value (SV): Defined as the number of milligrams of potassium
hydroxide required to hydrolyze (saponify) 1 g of fat. The higher the SV, the
lower the mean chain length of the component fatty acids of an acylglycerol.
QUALITY INDICATORS FOR EDIBLE FATS 173
Titer (titre): The solidification temperature of the free fatty acids derived from
a lipid. The higher the value, the greater the unsaturated fatty acid content. An
especially important characteristic in fats used to produce bar soap or fatty
acids, where degree of hardness is important. The melting point of an intact
fat is not a good indicator of its firmness, because this value depends on the
crystal form adopted by the fat. The titer value does not suffer from this
defect, and thus, it is a much more reliable estimate. Trade practice in the U.S.
rendering industry is to designate animal fats with titers of 40 C and up as
tallow, and those below 40 C as grease. Tallow titers can be as high as
5961 C, although 4050 C is more common. Lard exhibits a slightly higher
value, because of its greater content of high-melting stearic acid.
Formal classification standards for fats and oils have been defined by many
bodies and organizations. Those established by the Code of Federal Regulations
(United States) and the Codex Alimentarius are widely employed to guide com-
merce. The U.S. Code of Federal Regulations (41) specifies that lard should have
the following quality characteristics: free of foreign odors and flavors; maximum
free fatty acid value of 0.5% (as oleic acid equivalents) or an acid value of
1.0 mg of KOH consumed per gram of sample; maximum peroxide value (as
milliequivalents of peroxide per kilogram of fat) of 5.0; moisture and volatile
matter at a maximum of 0.2%; insoluble impurities no greater than 0.05%; and
white in color, with a maximum reading of 3.0 red units in a 5 14-inch cell on the
Lovibond scale.
In the Codex Alimentarius (42), maximum free fatty acid levels are specified as
0.65% for lard, 1.00% for premier jus, and 1.25% for rendered pork fat and edible
tallow. For all these, a peroxide maximum of 10-milliequivalents active oxygen per
kilogram fat is specified. The Codex standards also specify levels for antioxidants
and antioxidant synergists and maximum allowed amounts of impurities, soaps, and
certain metals.
Adulteration is another quality rating factor of commercial lipids. Methods for
the detection of tallow adulteration with lard were discussed above. Tests have also
been developed to detect the presence of beef fat in lard. The best known of these
tests is the Bomer test, which is based on the difference between the melting points
of acylglycerols and the fatty acids they contain (43). This difference is large for
unhydrogenated pork fat and small for tallow. The test is invalidated by the pre-
sence of hydrogenated fat in the lard. As another means of detecting adulteration,
it has been suggested that more than 0.01%, 0.05%, and 0.05% of branched chain
14, 15, and 16 carbon fatty acids, respectively, in lard indicates the presence of tal-
low (44). However, when pigs are fed tallow, they incorporate some of the branched
fatty acids into their depot fat (45).
Triacylglycerol profiles, determined by high-performance liquid chromatogra-
phy (HPLC), may also be a tool for the detection of the adulteration of pork by
beef fat (44). The presence of 5% or more of pork fat in beef or mutton tallow
can be detected and quantified by HPLC analysis of fatty acids in the sn-2 position
of the triacylglycerols, because the ratio of 16:0/18:1o9 at this position is about 5.0
174 ANIMAL FATS
for lard, whereas for edible tallow, it is about 0.4. G as liquid chromatography can
also be employed to make such a determination of the purity of a lipid.
5. REGULATORY AND COMMERCIAL CLASSIFICATIONS

OF ANIMAL FATS
Although Table 1 lists the fatty acid compositions of various lipids, this is not the
only or the final arbiter of their classification. As opposed to vegetable fats and oils
(other than olive oil), where only one oil is generally identified as originating from
an oilseed (e.g., corn oil), a diversity of definitions and specifications is used in the
identification of and trade in animal fat products. These often include statements of
the allowed limits of any number of quality parameters.
In the case of tallow, two broad categories are defined: edible and inedible.
Edible tallow originates from cattle or sheep that are judged by a competent regu-
latory authority to be healthy, sound, and fit for consumption at the time of slaugh-
ter. Tallow obtained from the inedible offal resulting from slaughter, from animals
unfit for consumption, or from outdated meats returned from commercial outlets is
classified as inedible.
The Code of Federal Regulations (41) of the United States concerns itself with
only one pure animal fat, e.g., lard, which is defined as the fat rendered from clean
and sound edible swine tissues. Tissues to be used for lard are to be reasonably free
from blood and shall not include stomachs, livers, spleens, kidneys, brains, or set-
tlings and skimmings. Leaf Lard is prepared from fresh leaf (abdominal) fat.
Lard (when properly labeled) may be hardened by the use of lard stearin (a lard
fraction rich in acylglycerols containing saturated fatty acids) or hydrogenated
lard or both and may contain refined lard and deodorized lard, if so labeled. A
detailed compilation of the killing and cutting fats to be used in producing lard
and rendered pork fat has been provided (46).
The Codex Alimentarius (42) contains international standards for four main pro-
ducts from animal sources: lard, rendered pork fat, premier jus, and tallow. Lard
is defined as the fat rendered from fresh, clean, sound edible-grade fatty tissues
from swine. These tissues must lack bones, detached skin, head skin, ears, tails,
organs, windpipes, large blood vessels, scrap fat, skimmings, settlings, pressings,
and be reasonably free of muscle tissues and blood. Rendered pork fat is defined
similarly to lard, with the exception that the tissues forbidden in lard production are
allowed. Premier jus (or oleo stock) is the product obtained by low-temperature
rendering of the fresh fat of heart, kidney, greater omentum, and mesentery of
bovines, collected at slaughter, as well as cutting fats. This fat has a creamy
white-to-light yellow color, a characteristic mild flavor, and a very low free fatty
acid content (1.0% maximum). Edible tallow (dripping) is the product obtained
by rendering the clean, sound fatty tissues, including trimming and cutting fats, but
also the attendant muscles and bones of bovine animals and/or sheep. It is distin-
guished from premier jus by the allowance of sheep tissues and of a greater diver-
sity of materials from which it can be obtained, by the specification of a higher
REGULATORY AND COMMERCIAL CLASSIFICATIONS OF ANIMAL FATS 175
TABLE 4. Codex Alimentarius Standards for Lard, Rendered Pork Fat, Premier Jus,
and Edible Tallow.a
Rendered Pork
Characteristic Lard Fat Premier Jus Edible Tallow
Relative density 0.8960.904 0.8940.906 0.8930.904 0.8930.904

(40 C/water at 20 C)
Refractive index n 40D 1.4481.460 1.4481.461 1.4481.460 1.4481.460
Titre ( C) 3245 3245 42.547 4049

Saponification value 192203 192203 190200 190202
(mgKOH/g fat)
Iodine values (Wijs) 5556 6072 3647 4053
Unsaponfiable matter 10 12 10 12
(maximum, g/kg)
a
See (42).
maximum free fatty acid content (1% to 1.25%), and by a higher allowed peroxide
level. Lard, rendered pork fat, and edible tallow may contain certain further pro-
cessed forms of the rendered fat, such as refined or hydrogenated product, or stear-
ines, as long as labeling regulations are followed.
The Codex descriptions specify that all edible animal fats must come from ani-
mals determined to be in good health at the time of slaughter and fit human con-
sumption as judged by a competent authority recognized in national legislation. The
main Codex analytical identity standards for lard, rendered pork fat, edible tallow,
and oleo stock are given in Table 4. The ranges for fatty acid composition specified
in the Codex standards for lard, rendered pork fat, edible tallow, and oleo stock are
given in Table 5. Note that these values are not necessarily constant over time,
having undergone revision since the previous edition of this chapter.
In terms of trade in animal fats, another important classification system is the
Specifications for Tallow and Greases established by the American Fats and Oils
Association. These specifications (Table 6) establish 13 categories of lipids, and
guide U.S. industry and commerce. One of the specified categories is for edible
lard, 11 are for various grades of tallow, and the remaining 2 provide specifications
for white and yellow grease. The categories are identified in terms of both species
of origin and the composition-related parameters minimum titer, maximum free
fatty acid content, maximum color, maximum refined and bleached color, and max-
imum moisture, insolubles, and unsaponifiables. The numerous designations of tal-
low are necessary because in various locations not only cattle but also swine and/or
poultry products are rendered together, giving products of varying content depend-
ing on the species mix. Choice white grease is a pork product, consisting of pork
lipids other than leaf lard. Accordingly, its fatty acid composition is very similar to
that of lard (Table 1). Yellow grease is a term given to used fat from deep fryers.
It is defined in terms of its free fatty acid content, with no specification made as to
biological origin. The displacement of tallow by hydrogenated vegetable oils in
deep fat frying in recent years has led to yellow grease that is presently in many
176 ANIMAL FATS
TABLE 5. Ranges of Fatty Acid Composition (%) Specified

by the Codex Alimentarius Standards for Lard, Rendered Pork
Fat, Premier Jus, and Edible Tallow.a
Lard, Premier Jus,

Fatty acid Rendered Pork Fat Edible Tallow
<14 <0.5 <0.5

14:0 1.02.5 26
14:ISO <0.1 <0.3
14:1 <0.2 0.51.5
15:0 <0.2 0.21.0
15:ISO <0.1
15:ANTI ISO <0.1 }<1.5
16:0 2030 2030
16:1 2.04.0 15
16:2 <0.1 <1.0
16:ISO <0.1 <0.5
17:0 <1 0.52.0
17:1 <1 <1.0
17:ISO <0.1
17:ANTI ISO <0.1 }<1.5
18:0 822 1530
18:1 3555 3045
18:2 412 16
18:3 <1.5 <1.5
20:0 <1.0 <0.5
20:1 <1.5 <0.5
20:2 <1.0 <0.1
20:4 <1.0 <0.5
22:0 <0.1 <0.1
22:1 <0.5 Not detected
a
See (42).
places largely a vegetable oil product. Accompanying this change have been
increases in the degrees of unsaturation and of trans-fatty acids in yellow grease.
Brown grease can be anything that does not meet the minimum specifications for
yellow grease, irrespective of origin (Table 6). Only one grade of poultry fat is
included in Table 6. Other poultry fat lipid products are sold, but they are not
detailed here because the standards for poultry fat are generally set by the indivi-
dual customer. Potential users should be acquainted with these specifications when
contemplating performance or purchase needs. Commercial renderers and brokers
of fats are familiar with these terms and specifications and the trading rules asso-
ciated with the purchase of these products.
In the United Kingdom, trade is also governed by a multiplicity of definitions for
tallow. Premier jus is the term applied to the highest, edible grade of tallow, and its
specifications conform to that of the Codex Alimentarius (42). For inedible tallows,
British Standard 3919 (49) specifies six grades and grease (Table 7). Again, these
grades of inedible tallow are defined on a compositional basis, including a specified
REGULATORY AND COMMERCIAL CLASSIFICATIONS OF ANIMAL FATS 177
TABLE 6. Specifications for Some Commercial Grades of Tallows, Animal Fats,

and Greases.a
Specificationsb
Titer FFA FAC R&B MIU
Grade ( C, min.) (%, max.) (max.) (max.) (%, max.)
c d
Lard (edible) 38 0.5 None
d
Edible tallow 41 0.75 3 None
Top white tallow 41 2 5 0.5 1
All Beef Packer 42 2 None 0.5 1
Tallow
Extra fancy tallow 41 3 5 None 1
Fancy tallow 40.5 4 7 None 1
Bleachable fancy 40.5 4 None 1.5 1
tallow
Prime tallow 40.5 6 1311B None 1
Special tallow 40 10 21 None 1
No. 2 tallow 40 35 None None 2
A Tallow 39 15 39 None 2
Choice White 36 4 1311B None 1
Grease
e e
Yellow Grease 39 None 2
Brown Greasef n.s. g
>15 n.s. n.s. n.s.
Poultry Fath 2835 15 19 n.1.i 2
a
As issued by (47).
FFA: free fatty acids, FAC: color as per Fat Analysis Committee, R&B: refined and bleached color, MIU:
moisture, impurites, and unsaponifiables.
c
Lovibond color for 5.25-inch cell: maximum 1.5 red. Lard peroxide value: 4.0 M E/K maximum.
d
Moisture maximum 0.20%. Insoluble impurities maximum 0.05%.
e
When required, to be negotiated between buyer and seller on a contract-by-contract basis.
f
See (5).
g
n.s.: not specified.
h
Provided by (48).
i
n.l.: not listed.
TABLE 7. Trading Grades for Technical Tallows and Animal Greases According
to British Standard 3919.a
FFA Bleached Moisture and Unsaponi- Titre Iodine Plastics

(max.) Colour, red Dirt (basis) fiable (max.) ( C) Value (max.)
Grade (% m/m) (max.) (% m/m) (% m/m) (min.) (max.) (mg/kg)
Tallow 1 3 0.5, 5 1/4 in. 0.5 0.5 40 55 200

cell
Tallow 2 5 1.0, 5 1/4 in. 1 1 40 55 200
cell
Tallow 3 8 3.0, 5 1/4 in. 1 1 40 55 200
cell
Tallow 4 12 4.0, 1 in. cell 1 1.5 40 58 200
Tallow 5 15 12, 1 in. cell 1 1.5 40 58 200
Tallow 6 20 No limit 1 2 40 58 200
Grease 20 No limit 2 2 36.040.0 61 200
a
See (49).
178 ANIMAL FATS
TABLE 8. Naming Conventions for Inedible Tallow and Grease Products.a
Rendered Rendered
Inedible Beef Tallow Tallow Recovered
Region Tallow (Standard Grade) (Lower Grade) Inedible Grease Cooking Oil
U.S. Bleachable Bleachable Prime, Special Choice White Yellow Grease

Fancy Tallow Fancy Tallow Tallow Grease
Packer Grade Rendered Grade
E.U. Technical Bonefat, 4% Tierkupper ROb, RVOb,
RTOb
U.K. UK 2 UK 6 RO, RVO, RTO
Eire Irish 2 Tallow 3/4 RO, RVO, RTO
Irish 6
(High FFA)
Italy Tallow S Tallow A RO, RVO, RTO
a
Provided by D. Dempsey, Unichema, Chicago, IL.
b
RO: recovered oil, may be either vegetable or animal; RVO: recovered vegetable oil; RTO: recovered
tallow oil.
maximum for plastic content. This material makes its way into inedible fats via the
inclusion of outdated commercial meats and fats, still in their wrapping containers,
into the rendering process. These standards are also often applied to edible tallow.
Table 8 provides some insight into the naming conventions of several countries
for fat products. Poultry fat is used virtually exclusively as an animal feed. For feed
grade fats, the accepted U.S. industry definitions are those established by the Asso-
ciation of American Feed Control Officials (Oxford, Indiana, www.aafco.org), an
organization primarily concerned with issues related to animal feeds. It defines
fat product, feed grade as any fat product which does not meet the definitions
for animal fat, vegetable fat, or oil, hydrolyzed fat or fat ester. The Association
defines three categories of material:
1. Animal fat, and within this category, poultry fat, which is fat obtained from
poultry tissues via commercial rendering or extracting. It consists primarily of
acylglycerols and contains no additions of free fatty acids or other materials.
The total fatty acid content exceeds 90%, with unsaponfiables and insoluble
impurities making up no more than 2.5% and 1%, respectively. The presence
of any added antioxidants must be stated.
2. Hydrolyzed animal fat is animal fat obtained via the procedures commonly
used in edible fat processing. Its free fatty acid content is not less than 85%,
with not more than 6% unsaponifiables and not more than 1% insoluble
impurities. A maximum moisture level must be guaranteed, and the presence
of any added antioxidants must be stated. Its source must be stated, e.g.,
hydrolyzed poultry fat.
3. Fat Product, Feed Grade is any fat product that does not meet the
definitions for fat and hydrolyzed fat. It is sold on the basis of its individual
PATTERNS AND TRENDS IN THE PRODUCTION AND USE OF ANIMAL FATS 179
specifications regarding total fatty acid content, unsaponfiables, insoluble

impurities, free fatty acids, and moisture. Again, added antioxidants must be
declared. Around these definitions, individual purchasers of poultry fat
typically set their own standards with regard to fat quality, as measured by
such parameters as free fatty acid content. Thus, for example, pet food
manufacturers will generally specify a higher quality fat than for other animal
feeds. Presently, most poultry fat is consumed as poultry feed, although there
is a trend toward increased use in other animal feeding applications as well.
6. PATTERNS AND TRENDS IN THE PRODUCTION AND USE

OF ANIMAL FATS
Domesticated animals are grown primarily for their meat. Carcass fat is a minor
coproduct, contributing less than 10% of the total market value of an animal. There-
fore, fat production does not drive producer decisions regarding the number of ani-
mals to raise. In this aspect, animal fats are a different type of commodity than most
oilseeds, for which the oil value constitutes a sizeable portion of the value of the
crop, causing producer decisions to be influenced by oil demand and price. Lipids
are desired dietary components, contributing energy and essential dietary nutrients.
Both a continuing increase in world population and increases in the standard of liv-
ing have led to increased lipid consumption in many regions of the world. However,
the increased demand for fats and oils has largely been met by an increase in vege-
table oil production.
Representative data for worldwide annual production of the major fats and oils at
selected intervals from 1968 to 2001, with a projection to 20082012, are presented
in Table 9. It can be seen that in 1968, global vegetable oil production was roughly
23 million metric tons and grease and animal fat production was approximately
9 million metric tons, for a total global lipid production of roughly 32 million
metric tons. Animal fats and grease constituted about 28% of global lipid produc-
tion. In the subsequent 32 years to the turn of the millennium, global production of
the major vegetable oils rose at an annual rate of about 13%, nearly four fold over-
all, to 90 million metric tons. In contrast, the sum production of tallow, grease, and
lard increased at only about 1.5% annually to 15 million metric tons. At the turn of
the century, then, greases and animal fats had fallen to constitute about 14% of the
global fats and oils production of approximately 105 million metric tons. This
percentage drop was not a result of decreased animal fat production but of tremen-
dous increases in the production of vegetable oils, especially palm-based lipids,
which increased about 14-fold, and soybean oil, which registered a roughly
5-fold increase. Projections for the near future (Table 9) suggest continued slow
growth and approximately constant market share for animal fats and oils.
In addition to large increases in vegetable oil production in recent decades, a
trend toward lower carcass fat contents at slaughter has also held down the rate
of growth of animal fat production. This trend has been a result of two factors: con-
sumer preferences for lean meat and economic pressure to produce animals more
180 ANIMAL FATS
TABLE 9. Worldwide Production of Fats and Oils (1000 metric tons).a
Commodity 1968 1978 19881989 19971998 20002001 20082012b
Edible vegetable oils

Cottonseed 2415 3195 3628 3701 3510 5900
Olive 1479 1582 1502 2526 2558 2100
Peanut 3505 3136 3729 4180 4296 5700
Rapeseed 1880 2693 7599 11425 13174 15600
Soybean 5540 11283 14574 23665 27029 25100
Sunflower 3975 4717 7263 8289 8333 12000
Corn 265 436 n.r.c 1,680b n.r 2749
Total 19059 27042 38,295 55466 58,900 69149
Tropical Oils
Coconut 2260 3148 2580 3285 3417 3300
Palm 1480 3578 9467 16973 23676
Palm kernel 395 569 1238 2202 2906 29800
Total 4135 7295 13285 22460 29999 33100
Animal Fats
Tallow and grease 4655 5866 6603 8342 8312 8100
Lard 4440 3663 n.r. 5,800d n.r. 7700
Total 9095 9529 13342 15800
a
See (50).
b
Predicted annual average production for the stated time period (51).
c
n.r.: not reported
d
See (4).
efficiently. As a result, for example, lard production per 100 pounds of pig live
weight was 13.9 pounds in 1959, 10.8 pounds in 1965, and 4.6 pounds in 1983
(46). Such reductions have been achieved by changes in animal breeding and nutri-
tion, and by a movement to younger ages at slaughter, which results in a carcass
with less depot fat. In the more economically developed countries, there has also
been a trend to reduce lipid consumption in an effort to reduce obesity. Further-
more, an increasing awareness over the past decade of the correlation between
the dietary consumption of saturated fats and cholesterol and the incidence of cor-
onary heart disease (5254) has led to the replacement of animal fats by vegetable
oils in many edible applications. Thus, in the United States, for example, animal
fats comprised 2.1% of the margarine and 21.2% of the shortening produced in
1984, whereas by the year 2000, animal fat usage in these items had fallen to
0.7% and 7.9%, respectively (50, 55).
Surprisingly, despite increased public awareness of, and stated dedication to, the
value of low fat diets, annual per capita edible lipid consumption in the United
States has grown in recent years. Between 1991 and 2000, for example, this value
rose from 65.5 to 74.6 pounds per person annually. In addition, the sum of lard and
tallow consumption over this period rose from 3.2 to 5.9 pounds per person. These
trends are attributable to an increase in the consumption in the home of commer-
cially prepared foods, which have a higher fat content and animal fat component,
and to an increase in dining in restaurants. The greater use of animal fats in these
PATTERNS AND TRENDS IN THE PRODUCTION AND USE OF ANIMAL FATS 181
TABLE 10. Prices (cents U.S., per pound) of Selected Commodity Lipids at Randomly
Chosen Times in the Period 20002003.
Commodity March 27, 2000a October 15, 2001b June 23, 2003c
Soy oil, refined 18.9 18.6 28.3

Palm, refined 18.0 14.8 32.3
Tallow, edible 12.5 13.0 19.0
Tallow, inedible 10.5 12.0 18.5
Lard 12.3 13.3 19.0
Grease, white 10.0 11.0 18.0
Grease, yellow 7.5 9.3 14.5
a
See (56).
b
See (57).
c
See (58).
sectors is the result of their historically lower price compared with even the least
expensive refined edible vegetable oils, and the belief that in some applications,
such as deep fat frying, animal fats impart superior flavor to foods. Thus, it may
be that societal factors accompanying affluence can result in unexpected increases
in the consumption of animal fats.
Table 10 presents representative recent U.S. price data for lipids of interest here.
U.S. values are presented because in many cases they constitute the only, or the
most complete, data sets for areas of interest here, and because the United States
is among the top producers and consumers of animal fats. Table 10 shows that in the
United States, edible tallow and lard are typically priced at 6070% the price of
palm and soy oils. Table 10 also shows that in recent years, inedible tallow has
offered a price discount of up to 15% relative to edible tallow over this time period,
although the price gap has nearly disappeared recently. Inedible tallow is used in
industrial applications. As the price difference between it and edible tallow nar-
rows, it is common to see inedible tallow displaced by the edible material, which
often is of higher quality and thus requires less cleanup of starting material or pro-
duct. When the U.S. fast food industry switched from tallow to vegetable oils for
deep fat frying in the mid-1980s, substantial amounts of edible grade tallow flowed
into uses that had previously consumed inedible material. Table 10 also illustrates
the significantly lower costs of yellow and white grease compared with refined
vegetable oils, the latter being nearly twice the price of the greases.
For tallow, although industry identifies and trades in many grades of raw material,
cumulative production and use data are listed in terms of only edible vs. nonedible
tallow, at best. The United States accounts for approximately half of the annual
world production of tallow, with the balance coming primarily from Australia,
Canada, New Zealand, Argentina, and Brazil. In the United States, the production
of inedible tallow typically is slightly more than double that of edible tallow. Thus,
in the year 2000, the U.S. inedible tallow production was 1.7 million metric tons,
and edible tallow production was estimated at 0.76 million metric tons (59). In the
year 2000, estimated lard production, for use mostly in cooking, was estimated at
182 ANIMAL FATS
only 0.24 million metric tons (59). Grease production rivals that of inedible tallow,
and in that year was estimated at 1.5 million metric tons (59).
Edible tallow is used primarily in shortening (i.e., baking and frying fats) and
margarine, with additional uses in the chemical, soaps, and personal care products
areas. About 30% of U.S. inedible tallow is exported. The remainder is used in ani-
mal feeds (48% of U.S. usage in 2002 and rising), for the production of bar soap for
washing (6% ), as a source of industrial fatty acids (32%), and for other uses such as
textile sizing, leather processing, metalworking, lubrication, and paint production.
Substantial growth is occurring in the use of tallow in animal feeds.
In 1999, the top five tallow exporting countries were the United States, with just
over 50% of the 2.3-million-ton export market, Australia, Canada, New Zealand,
and Germany (60). The top five lard exporting countries were the United States,
Germany, Hong Kong, Argentina, and Hungary, with the United States and
Germany accounting for nearly 80% of lard exports (60). Although the United
States exported 80,000 to 130,000 metric tons of tallow and lard to the EU annually
near the turn of the century, none of this was destined for the European food market.
Hormones are used in American meat production, and products so produced are
banned from edible applications in Europe. Cultural and religious practices and
regulations can influence the use of tallow and other animal fats. This may be a
result of the existence of prescribed methods of slaughter, as for those following
the Islamic and Jewish faiths, or to other dictates. Jewish law forbids the use of
pork products, for example, and in Islamic countries, the use of fatty acids from
animal sources is forbidden in toiletries (e.g., toothpaste).
Poultry fat production rates are difficult to determine, because there are no
formal governmental talleys for this commodity. U.S. industry sources estimate cur-
rent annual production of rendered poultry fat in that country to be 900,000 metric
tons, or about one-tenth of global tallow production (61). This probably represents a
sizeable proportion of global poultry fat production. Presently most of this goes into
animal feeds. Historically this was the poultry feed market, although recently there
has been an increase in use in other animal feeds as well.
Grease production data are often lumped with that for inedible tallow. Greases
constitute about 45% of this total and are used exclusively in nonedible applica-
tions. In the United States, the use of yellow grease in animal feeds is increasing
yearly and in 2000 accounted for 36% of total feed fat usage (62).
7. PROCESSING OF ANIMAL FATS
7.1. Rendering
The fatty tissues separated from meat animals at slaughter and during cutting
consist of fat deposited in a connective tissue matrix containing protein and water.
To separate the fat from other components, a technology known as rendering (63)
has been developed that is based primarily on the melting of the fat and its
removal from the nonfat matrix surrounding it. The aim of rendering is to obtain
PROCESSING OF ANIMAL FATS 183
as complete a separation of the fatty tissue components as is possible. Most render-

ing systems rely on heat to release fat from the cells of the fatty tissues, in either
the absence (dry rendering) or presence (wet rendering) of added water/steam. In
other methods, the heat is kept low and the fat is released by mechanical rupture of
the cells.
In dry rendering, the chopped fatty tissues are heated to about 121 C, usually in
a horizontal steam-jacketed vessel, to disintegrate the fat cells, release the melted
fat, and drive off moisture. Most batch cookers for dry rendering are equipped with
rotating agitators that may be steam-heated. Agitation aids in heat transfer. After 4
to 5 hours, the fat is released and sufficient moisture is removed. The material is
pressed or expelled to recover the lipid, which is strained or filtered to remove
cooked proteinaceous residue (the cracklings). Dry rendering can be carried out
at atmospheric pressure, under vacuum, or at elevated pressure. The atmospheric
pressure method is the most common, with operating temperatures of 115 C to
140 C. It is the method of choice for inedible animal lipids. It offers the advantages,
compared with wet rendering, of reductions in water requirements and air and water
pollution, and a reduction in odors. In the United States, it is no longer approved for
use by the U.S. Department of Agriculture for the production of edible grade fats.
Rendering for the production of inedible lipids can use raw materials such as dis-
eased or condemned animals and fats picked up from butcher shops as trimmings or
scraps or outdated meat products. The latter can arrive in the polyethylene contain-
ers in which they were offered for edible sale. These are often thrown directly into
the rendering vessel. If not removed downstream, this polyethylene can later pose a
problem to the operation of continuous fat splitters. For this reason, some fat spe-
cification sheets list a maximum polyethylene content.
In the low-temperature continuous (dry rendering) process, finely ground fatty
material is heated to approximately 38 C and fed to a centrifuge that separates
a fat sludge stream from the remaining material, which is largely undenatured
protein. The fat stream is heated to approximately 99 C and centrifuged again to
recover the majority of the fat. Wet rendering is then used to recover residual fat
remaining in the solids stream. The final lipid product of the continuous low-tem-
perature process is very light in color and low in free fatty acids.
In wet rendering processes, the fatty tissues are heated in the presence of water,
usually at lower temperatures (82 C to 96 C ) than in most dry rendering. Fats ren-
dered at lower temperatures typically have less color and a milder flavor than those
that are dry-rendered. Wet rendering is the method of choice for edible fat render-
ing, often using a steam rendering process. In this method, live steam is injected
into a closed vessel containing the fatty tissues, and the rendering takes place under
pressure to shorten the cooking time. Lard produced by this process is called
prime steam lard.
Continuous processes have been developed for both wet and dry rendering. They
offer the advantages of higher throughput rates and lower operating costs, although
batch systems offer better control of product quality. Nonetheless, high-quality
edible products can be produced by continuous low-temperature rendering systems.
In these systems, the fat is separated from the protein fraction at low temperatures
184 ANIMAL FATS
in a relatively short time. The resulting products are light in color, mild in flavor,
and low in free fatty acids. The features and operation of a low-temperature system
used for edible fat processing have been described in detail (64).
If edible-grade fatty tissues are handled and rendered properly, the resulting pro-
duct is suitable for use as a food fat without further treatment. For the better grades
of tallow and greases, air is excluded from the melting operation in order to main-
tain the color and reduce oxidation of the unsaturated fatty acids. Alakali refining
can be employed if the free fatty acid content is greater than about 0.3% or if col-
lagen or proteinaceous material are present (65). Edible animal fats also may be
subjected to bleaching, hydrogenation, deodorization, interesterification, or frac-
tional crystallization to improve their characteristics or produce fats for specialized
use. For the finer grade products, such as bleachable tallows, care is taken to use
high-quality raw material from the packing plant operation. Materials such as floor
sweepings, catch basin contents, and the carcasses of diseased and condemned ani-
mals go to lower grade tallows.
7.2. Bleaching
Bleaching is conducted to remove components that give fats an undesirable color,
and to decompose peroxides. Most lards do not require bleaching, whereas tallows
may be bleached to remove colored materials. Bleaching is usually accomplished
by adding natural or acid-activated clays (montmorillonite) and, to a lesser extent,
activated carbons. These adsorb the color bodies and certain degradation products
in the fats. The clays have a high density of negative charges and act via cation
exchange interactions to remove colored compounds, which are generally polar
or charged. They also remove metal ion contaminants such as iron and copper,
which can serve as pro-oxidants. Specific recommendations for bleaching lard
and tallow have been reported (66). For lard, a contact time of 15 minutes at 95
100 C is suggested, with a maximum of 0.5% mildly activated clay or 0.25% mod-
erately activated earth. For top grades of tallow, up to 1% mildly activated clay or
about 0.3% well-activated clay is suggested, and a contact time of 20 minutes at
95100 C is normal. For animal fats, activated carbon tends to be less frequently
employed because of its higher cost, although it is often efficacious as an admixture
with clay in the treatment of particularly dark samples or to reduce bleaching earth
consumption.
Color and peroxide values (PV) have been compared for tallows bleached at
atmospheric pressure and under vacuum at several temperatures (67). At 90 C,
atmospheric bleaching gave better color than bleaching under vacuum, but peroxide
values were lower for the vacuum bleached product (PVs of 2 versus 16, respec-
tively). At higher temperatures, color was better and peroxide values were lower.
A bleaching temperature of 90 C to 110 C was recommended for beef tallow.
Treatment of lard with bleaching earth decomposes the peroxides and increases
the content of conjugated trienes, which absorb at 268 nm. This characteristic has
been the basis of a quality control procedure for determining whether lard has been
bleached (66).
7.3. Hydrogenation (67, 68)

In some applications, a relatively hard fat is desired. In the process known as hydro-
genation, a metal catalyst is employed to add hydrogen to fatty acid double bonds,
reducing them to single bonds. As the melting point and solid fat content of lipids
increase with the degree of saturation of their fatty acids, this increases firmness. In
addition, the elimination of oxidation-prone double bonds increases oxidative sta-
bility. The hydrogenation process can be adjusted and controlled, and it is then
termed partial or selective, so that various degrees of hydrogenation can
be attained. Hydrogenation is also employed in the fatty acid industry to reduce
fatty acid nitriles to amines and to produce fatty alcohols from fatty acids, esters,
or acylglycerols. These technologies will not be considered here.
As an intermediate product, hydrogenation isomerizes cis-double bonds to the
more thermodynamically stable trans-configuration. It also catalyzes double-bond
migration along the fatty acid chain. As trans-double bonds confer higher melting
points than do their cis-counterparts, this isomerization again contributes to
increased firmness. Trans-unsaturated acylglycerols are desirable in the context
of margarine, confectionary, and similar applications because, although solid at
room temperature, they have lower melting points than acylglycerols composed
of saturated fatty acids. Thus, they can give a more desirable spreading and melting
performance than lipids hardened by the presence of saturated acylglycerols.
Process technology optimized for the production of trans-fatty acids while reducing
the yield of fully saturated fatty acids is therefore often employed. However, sub-
stantial concern regarding the medical implications of trans-fatty acids in the diet
has led to questions about the role and future of hydrogenation in edible lipid tech-
nology (see discussion in Recent Developments).
Because of their higher contents of unsaturated fatty acids relative to animal fats,
vegetable oils are more commonly subjected to hydrogenation technology. This
particularly improves their performance as margarines and in baking and frying
applications. Being naturally firm, tallows are seldom hydrogenated for edible
uses as intact fats, and then only very slightly. Mild hydrogenation of edible tallows
delays or prevents the development of undesirable flavors thought to be caused by
the peroxidation of trienoic and tetraenoic fatty acids (46). However, for use in the
production of mono- and di-acylglycerols, which are widely used as food emulsi-
fiers, animal fats are generally fully hydrogenated, achieving final iodine values less
than 1.
Unhydrogenated lard can lack desired melting point or plastic range properties,
especially because the melting range characteristics of lard from different animals
or animals fed different diets can vary greatly. A consistent product can be obtained
by blending lards from different sources or by blending them with lards that have
been hydrogenated to almost complete saturation (IV < 5)(46).
For edible applications, lipids are often used intact. For such uses, the whole fats
and oils are subjected to hydrogenation. For industrial uses, lipids are the feed-
stocks for free fatty acid production. When firmer fractions are needed for industrial
applications, the free fatty acids obtained by lipid hydrolysis are subjected to
186 ANIMAL FATS
hydrogenation. Inedible tallows have traditionally been used as an industrial source

of saturated fatty acids, and for this application, their fatty acids are extensively
hydrogenated. When the economics are favorable, usage can shift to edible tallows
because their greater purity makes for easier processing.
Industrial hydrogenation may be conducted in batch or continuous modes. In
either case, the engineering challenge is to achieve efficient contact among liquid
lipid, solid metal catalyst, and gaseous hydrogen. Batch mode is by far the most
common approach in industrial edible oil hydrogenation, being used in more
than 90% of applications, although in the future a shift to continuous processes
may be seen. Batch reactors can have capacities up to 35 metric tons. To increase
turbulence and mixing, they contain rotating central shafts bearing agitators, and
they may also have baffles built into the reactor walls. Agitation also aids in tem-
perature control, and it keeps the catalyst suspended. Continuous reactors, with
typical outputs of 50 200 tons per day, may move a slurry of oil and catalyst along
a pipe, with vigorous introduction of hydrogen at several sites to provide turbulent
mixing. Alternatively, hydrogen and oil may be mixed prior to addition of catalyst
and then passed through the reactor vessel. Fixed bed systems also exist, in which
the catalyst is packed in a column through which the reactants flow.
Hydrogenation is conducted under pressure, typically 300 400 psi (20 27
atmospheres [atm]), and at temperatures of 120 200 C. Nickel is most commonly
employed as catalyst, and it is produced by the decomposition of nickel salts on
either silica or alumina carriers. Platinum is used less extensively and is more
costly. These metal catalysts are sensitive to poisoning by such agents as peroxides,
water, polyethylene, phosphorus, sulfur, nitrogen, halides, and free fatty acids and
their soaps. Inactivation losses increase the cost of hydrogenation. In some cases,
impurities in the starting material can be successfully and economically removed by
screening or filtration (polyethylene) or treatment with bleaching earth. In other
instances, it is expedient to employ a higher quality, cleaner, starting material in
order to extend catalyst lifetime.
The reduction of double bonds is an exothermic reaction. Heat control is impor-
tant during hydrogenation to avoid damage to the product and the equipment. Also,
the capture of heat and its reuse to bring subsequent hydrogenation runs up to reac-
tion temperatures is a significant aspect in overall process economics.
7.4. Deodorization
Animal fats are subjected to deodorization when a very bland or essentially flavor-
less fat is desired, such as in margarines or cooking fats. The fats are heated at
200 C to 260 C in the absence of air (to prevent oxidation) and treated with dry
steam under a vacuum of 510 milliatmospheres. Off-flavor compounds are volatile
under these conditions and are captured and removed in the steam stream. In addi-
tion to flavor components, free fatty acids, which can also contribute undesirable
flavors, and other minor constituents such as peroxides, sterols, sterol esters, toco-
pherols, and other natural antioxidants are partially or completely removed from the
fat by this treatment.
a. Alcoholysis
H2COCOR1 H2COH
HCOCOR2 + R4OH HCOCOR2 + R4OCOR1 & Etc.
H2 COCOR3 H2COCOR3
b. Acidolysis
H2COCOR1 H2COCOR4
HCOCOR2 + R4COOH HCOCOR2 + R1COOH & Etc.
H2 COCOR3 H2COCOR3
c. Transesterification (ester interchange, ester rearrangement, interesterification)
H2COCOR1 H2COCOR4 H2COCOR5

HCOCOR2 + HCOCOR5 HCOCOR2 & Etc.
H2COCOR3 H2COCOR6 H2COCOR3
Figure 2. Types of interesterification reactions of acylglycerols.
7.5. Interesterification
The ester bonds of fats and oils are not immutable. Under appropriate conditions,
they can be broken and reformed. This allows the replacement of the acid and alco-
hol components in an ester bond, a class of reactions generally termed interester-
ification (69 71). Three general types of interesterification reactions can be
identified (Figure 2):
1. Alcoholysis, as in lysis by alcohol, in which a new alcohol species (R-OH)

is exchanged into the ester bonds of the lipid. This converts an acylglycerol
into simpler fatty acid esters, as in the reaction of a lipid with methanol to
form fatty acid methyl esters. A subset of this reaction type is glycerolysis,
where the incoming alcohol is glycerol and the products are predominantly
mono- and di-acylglycerols, which see extensive use as emulsifiers. In a sense
the lysis of ester bonds by water, hydrolysis, is a form of alcoholysis in which
the role of the generic alcohol, R-OH, is played by H-O-H.
2. Acidolysis, lysis by acid, in which a new acid group (R-COOH ) is introduced
into the ester bonds of the lipid. This results in replacement of the fatty acids
within a lipid by other fatty acids. An example would be the acidolysis of
tripalmitin with oleic acid to generate a product containing monooleoyldi-
palmitin, dioleoylmonopalmitin, and triolein.
3. Transesterification, in which fatty acid exchange occurs between esters. For
example, the transesterification of tripalmitin (PPP) and triolein (OOO)
would produce all positional isomers of POO and PPO. The reaction can
be conducted using a single intact fat as the reactant, in which case shuffling
188 ANIMAL FATS
of fatty acids within the component acylglycerols will occur. As the simplest
example, if a homogenous sample of POP was subjected to transesterification,
the product would contain a mixture of PPO, OPP, and POP as well as PPP,
OOO, OOP, POO, and OPO. This reaction can also be conducted with
mixtures of natural lipids, as in the interesterification of tallow and sunflower
oil. This type of reaction is sometimes referred to as ester interchange, ester
rearrangement, or confusingly, interesterification. Of the types of interester-
ification reactions, transesterification is the one most commonly conducted on
edible fats and oils.
Although interesterification will occur in the absence of added catalysts at suffi-

ciently high temperatures, catalysts are employed by industry to speed this reaction,
reducing reaction time and the sample degradation that occur at elevated tempera-
tures. The most commonly used inorganic catalysts are alkaline ones such as
sodium methoxide, sodium ethoxide, sodium or potassium metal, and alloys of
sodium and potassium. Catalyst concentrations of 0.05% to 0.1% are employed.
As the catalysts will react with water, free fatty acids, and oxidized compounds,
it is important to use clean, dry feedstocks. Reaction temperatures are generally
kept below 100 C. The reactions can be run in batch or continuous formats. In
batch mode, the reaction times are typically less than an hour.
In the simplest interesterification reaction, termed nondirected, fatty acids are
rearranged among the acylglycerol species to achieve a fully random, statistically
determined population distribution. This can be useful in modifying the perfor-
mance properties of a natural fat or oil that contains a preponderance of one acyl-
glycerol species, and for this reason has been useful in improving the utility of lard
in baking applications. Shortenings for cakes and related products must cream read-
ily to facilitate incorporation of air into the batter. Unmodified lard has a large
proportion (64%) of acylglycerols with palmitic acid at the sn-2 position, and a
relatively high proportion (approximately 27%) of disaturated acylglycerols, mostly
oleoylpalmitoylstearin (OPS). These features cause the lard to adopt a
b configuration upon crystallization. The resulting large crystals confer a grainy
texture and give poor creaming performance. The uniform acylglycerol population
also results in an undesirable narrow plastic range. Randomization of lard, in the
absence of any additional fats or oils, is the most common application of interester-
ification technology to animal fats. It decreases the proportion of palmitic acid in
the sn-2 position of the lard triacylglycerols to 24% (72, 73) and produces a mixture
of disaturated acyglycerols with a substantially lower melting point than the origi-
nal OPS-rich material. Randomized lard is an improved baking fat. It crystallizes in
the b0 form, which is characteristic of hydrogenated vegetable shortenings, has
smaller crystal structures, lacks graininess, and exhibits improved performance as
a creaming agent.
In directed interesterification, the course of the reaction is shifted away from
production of a population of acylglycerols with a fully random fatty acid distribu-
tion. This is achieved by a modification of reaction conditions to selectively remove
from reaction some of the produced acylglycerol species as they are formed. Most
typically this involves conducting the reaction at a sufficiently low temperature that
formed acylglycerols with melting points above that temperature solidify as they
are formed and precipitate from the reaction. This approach is used to remove
the high-melting trisaturated species from the reaction liquid. It is not that these
species are themselves actually isolated and removed as they form. Rather, their
precipitation simply prevents the further participation of their fatty acids in the
interesterification event. The remaining ester pool is enriched in unsaturated fatty
acids. As additional saturated acylglycerols are formed by continued interesterifica-
tion, they too will be removed from reaction. Thus, as the interesterification-
precipitation reactions continue, the acylglycerol population in the fluid phase of
the reaction mixture becomes increasingly unsaturated. The final acylglycerol dis-
tribution will depend on such variables as the temperature and duration of reaction.
The reaction is not run to such an extent as to drive all saturated fatty acids into
triunsaturated acylglycerols, as this is both undesirable and time consuming. Rather
it is conducted until a sufficient distribution of acylglycerol species is achieved to
confer desired performance properties.
Directed interesterification can be employed to produce lard with an increased
solids content at high temperatures, because of production of a fully saturated acyl-
glycerol population. Such a product would be plastic over a greater range of tem-
perature. Lard produced by nondirected interesterification requires the addition of
stearin for high-temperature stability.
Interesterification reactions can also be conducted between two different natural
lipids, generating a product with an acylglycerol content representing the statistical
random population predicted by the content of the starting materials. As it provides
a means of introducing saturated fatty acids into the acylglycerols of liquid vege-
table oils, interesterification of oils with hard fats represents an alternative to partial
hydrogenation for the production of plastic fats for margarine. With increasing con-
cerns regarding the negative health implications of the consumption of the trans-
unsaturated fatty acids generated by conventional hydrogenation, it is possible
that this approach will be implemented at the industrial scale for margarine produc-
tion. The interesterification of sunflower oil with lard and tallow has been described
(74). Also, soybean oil has been interesterified with beef tallow to produce a plastic
fat suitable for use in making tub-type margarine (75). The interesterified blend of
60% soybean oil and 40% tallow contained 33.4% trans-fatty acids from the tal-
low, substantially less than in commercial margarines produced by hydrogenation
hardening of vegetable oils.
Biological catalysts can also be used to conduct interesterification reactions.
Lipases are enzymes produced by nearly all living organisms to catalyze hydrolysis
of the ester bonds of fats and oils, the first step in their metabolism. Lipases can be
used as applied catalysts for lipid hydrolysis. However, in low water (microaqu-
eous) systems, with water contents of a few percent or less, they will also catalyze
the various interesterification reactions. Lipase catalysis offers several advantages
over nonenzymatic catalysts, among them the fact that because they are active at
ambient temperature and pressure, lipases reduce energy needs, minimize degrada-
tion of the feedstock, and allow reactions on labile polyunsaturated lipids. Various
190 ANIMAL FATS
known lipases exhibit substrate specificity, including specificity regarding the

length or degree of unsaturation of the fatty acid they will accept as substrate,
or the position on the glycerol molecule at which they will act. These can be
exploited to perform directed interesterifications involving positional or fatty acid
selectivity.
Considerable recent research has defined conditions for successful use of lipases
and other enzymes in numerous lipid modification reactions, including a variety of
types of interesterifications (69, 71, 76). For edible applications to date, they have
been employed at industrial scales for the production of (1) cocoa butter substitutes,
for which disaturated, monounsaturated acylglycerols with the unsaturated fatty
acid in the sn-2 position are desired (77); (2) to produce human milkfat analogues,
where 2-palmitoyl acylglycerols are desired (77); (3) in the synthesis of 1,3- di-
acylglycerols (78); and in the production of diacylglycerols for edible applications.
These reactions employ vegetable oils as feedstocks.
The use of enzymatic catalysis for the interesterification of animal fats has also
been reported. For example, blends of tallow and rapeseed oil (79) were interester-
ified using a commercial immobilized 1,3-specific Rhizomucor miehei lipase as a
catalyst. The altered composition of triacylglycerols in the interesterified product
was reflected in significant changes in solid fat content in the temperature range
of 0 45 C. The degree of melting point reduction achieved depended on the
mass fraction of the substrates: the lower the mass fraction of tallow, the larger
the decrease. Similarly, the solid fat content of tallow was altered by interesterifica-
tion with high-oleic sunflower oil or soybean oil using lipases with either positional
or fatty acid selectivity (80, 81).
Combinations of physical fractionation and enzyme-catalyzed interesterification
have been employed to modify lipids. With regard to animal fats, Bhattacharyya
et al. (82) employed low (12 C and 15 C) and intermediate (25 C) temperatures
in acetone to prepare a high-saturates hard stearin and a softer olein fraction
from tallow. Chemical or enzymatic interesterification of these fractions was then
conducted to produce various samples with melting points and solid fat con-
tents similar to shortening, margarine, and vanaspati (a plastic fat, usually prepared
by hydrogenation, popular in India, Pakistan, Bangladesh, and some Eastern
countries). Similarly, tallow was subjected to directed lipase-catalyzed self-
interesterification, in either a batch or continuous mode, to reduce its content of
saturated fatty acids (83). Trisaturated acylglycerols synthesized by enzyme action
were removed from the mixture by a low-temperature step, resulting in the produc-
tion of an unsaturate-enriched tallow derivative. However, the process succeeded in
elevating the content of unsaturates from 45% to only 57% and required a 14-day
incubation for this purpose.
Human milk lipids are unique in that they contain predominantly palmitate at the
sn-2 position. To produce an infant formula whose lipids resemble those of mothers
milk, Yang et al. (84) employed Rhizomucor miehei lipase to catalyze the acidolysis
of lard, which is rich in 2-palmitoyl acylglycerols, with soybean fatty acids. As the
enzyme had tight specificity for reaction at the sn-1 and sn-3 positions of the
acylglycerols, the 2-palmitoyl structure was preserved, creating a product whose

fatty acid composition and acylglycerol structure resembled that of human milk.
A process similar to this, but using palm stearin as the source of 2-palmitoyl
residues, has been commercialized.
The term structured lipid (SL), very broadly defined, refers to acylglycerols
whose fatty acid composition or distribution has been altered by enzymatic or none-
nzymatic catalysis, or any of a number of biological or physical methods (85). The
products of acidolysis and transesterification are thus structured lipids. More speci-
fically, this term is often applied to acylglycerols in which some of the long-chain
fatty acids at the sn-1 and -3 positions have been replaced by medium-chain length
ones, generally caprylic acid (8:0) or capric acid (10:0). In these positions, medium-
chain fatty acids are readily released from fats and oils in the gut by the action of
(1,3-positionally specific) pancreatic lipase and absorbed. Medium-chain fatty acids
are directly metabolized for energy rather than deposited as depot fat. Thus, struc-
tured lipids containing them provide readily available energy and have a reduced
tendency to foster obesity. They are desirable dietary components for those requir-
ing high-density energy sources, such as athletes and individuals recuperating from
burns. The presence of some degree of long-chain fatty acid content in SLs is desir-
able to provide a source of essential fatty acids, especially linoleic acid (18:2). Sub-
stantial work has been conduced to investigate and optimize SL production from
vegetable oils. Animal fats have also been shown to be substrates for SL produc-
tion. Thus, the introduction of caprylic acid into the sn-1 and sn-3 positions of an
unsaturate-rich fraction of tallow (86) or chicken fat (87) and of unfractionated
chicken fat (88) have been described.
Cocoa butter is the premier confectionary fat. Its desirable property of sharp
melting near the temperature of the tongue, which imparts a cooling sensation, is
largely a result of its high content of 1,3-disaturated, 2-monounsaturated triacyl-
glycerols. In an attempt to find economical replacements for cocoa butter, substi-
tutes have been produced from several types of less-expensive lipids by several
investigators, some at the commercial scale. Osborn and Akoh (89) enzymatically
randomized the fatty acid distribution of tallow and conducted the lipase-catalyzed
acidolysis of tallow with a twofold molar excess of stearic acid. The physical prop-
erties of the resulting products suggested their potential usefulness as cocoa butter
replacements in chocolate type coatings.
Other uses of lipases as catalysts for animal fat modification are considered in
the discussion of biodiesel. Despite this variety of research-scale investigations,
enzymatic catalysis has not yet been implemented for animal fat modification at
industrial scales. This may be a result of the recent general decline in the food
use of animal fats or a result of the disadvantages of lipases as catalysts. Among
these are the facts that they are relatively expensive compared with conventional
catalysts, can tend to be unstable, and require relatively expensive process technol-
ogies. Although enzymes can be appropriate catalysts for the production of high-
value food, nutraceutical and pharmaceutical products, these factors have to date
largely prevented their use for the production of bulk food lipids.
192 ANIMAL FATS
7.6. Fractionation
Natural fats are heterogenous in composition, containing acylglycerols with differ-
ent fatty acid compositions. Each acylglycerol exhibits unique chemical properties,
among them melting temperature and solubility in organic solvents, that depend on
the size and degree of unsaturation of these fatty acids and their position on the
glycerol backbone. Physical fractionation relies on these differences in chemical
behavior to isolate specialty subfractions with desirable compositions and perfor-
mance properties.
The most frequent use of fractionation (9092) is to separate a natural fat into
two general categories: (1) fractions whose acylglycerols are enriched in saturated
fatty acids, are firm or solid at room temperature, and are referred to as stearin;
and (2) fractions whose acylglycerols are relatively rich in unsaturated fatty acids,
are liquid at room temperature, and are referred to as olein.
The two predominant types of fractionation are termed dry and wet. Dry
fractionation is the oldest, simplest, and most widely practiced approach. In very
general terms, it involves melting the lipid, cooling it to some desired temperature
below the melting point of the more saturated acylglycerols, and collecting the
crystals of this stearin fraction when they form. Contemporary industrial scale crys-
tallizers have capacities between 5 and 50 tons. To operate effectively, substantial
attention is required to such details as the triacylglycerol composition and purity of
the materials (the latter influences crystal formation), the nature of any pretreatment
of the fat, temperature differential between cooling surfaces, and the melt, cooling
rate, and degree of agitation employed during cooling and holding. The goal is to
obtain large crystal sizes, as these are most readily removed in downstream recov-
ery operations and will carry over the least amount of entrained olein. This method
is also termed fractionation from the melt, simple fractionation, or natural fractio-
nation. Recovery of the solid fraction can be by centrifugation or by vacuum or
membrane filtration.
The relationship among composition, melting point, titer, and solid fat index of
beef tallow and its liquid and solid fractions obtained by dry fractional crystalliza-
tion has been described (93). This study was conducted with Urguayan tallow,
which has been reported to have a higher titer (43.2 47.8 C) and melting point
(45.0 48.8 C) than is average for beef tallow (94).
Alternatively, in a commercial process termed Lipofrac, an aqueous solution
containing a surfactant and an electrolyte are added such that the solid fat crystals
partition into the aqueous phase. This is isolated and heated to melt this stearin frac-
tion, allowing its recovery by centrifugation (95, 96). The Lipofrac method results
in higher stearin yields than obtained by dry fractionation using vacuum filtration
for product recovery. However, the introduction of more efficient means of recover-
ing the stearin, especially the use of membrane filter presses introduced in the
1980s, to dry fractionation technology has increased the yield of stearin to such
a degree that the use of Lipofrac technology has declined.
In wet fractionation, the fat is dissolved in organic solvents, most generally
hexane or acetone, and the solution is brought to a temperature suitable to allow
crystallization of a higher melting acylglycerol fraction of desired composition. The

crystals are separated by decanting, filtering, or centrifuging. The solvent acts to
dilute and reduce the viscosity of the lipid, allowing more effective removal of
the included liquid lipid from the solid fat fraction. Thus, this method gives a clea-
ner solid fat fraction. Crystal nucleation and growth are faster, heat transfer is
easier, and the recovered stearin fraction can be washed with fresh solvent to reduce
the amount of entrained liquid fraction. However, wet fractionation is more costly
than dry fractionation, and it is not presently being implemented at the industrial
scale for the fractionation of animal fats.
Multiple fractionation steps at various temperatures can be conducted. Thus, an
olein fractionation can be chilled to further isolate a solid component enriched in
the least saturated of the unsaturated lipid components of a lipid. Tallow can be
fractionated to yield several fractions with different melting ranges. Tirtiaux (97)
has described a process for fractionating beef tallow to give products ranging
from a very hard stearin (melting point [m.p.] 56 C) to an oil (m.p. less than
20 C). The wet fractionation of beef tallow using acetone has also been described,
producing five fractions using a multistep crystallization process (98). Two of the
five fractions were crystalline, one was a plastic solid, and two were liquid. The
properties of the plastic solid were similar to those of cocoa butter, and this fraction
was reported to have excellent compatibility with cocoa butter. However, the pro-
cess was never commercialized, probably because of complexity, a decline in ani-
mal fat food use, and competition from superior fractions produced from palm oil.
Fractionation technologies are much more frequently employed in the proces-
sing of vegetable oils, especially palm oil. An estimate of the magnitude of the dif-
ference in scopes of the applications can be obtained by examination of the reported
new fractionation capacities installed worldwide in the period 19911993: >8000
tons per day for palm oil vs. 320 tons per day for all other oils and fats combined
(92). Nonetheless, the use of tallow fractions as cost-effective ingredients to substi-
tute for, or extend, cocoa butter has been investigated (99), production of a pourable
tallow oil shortening for deep-fat frying has been reported (100), and tallow frac-
tions have been reported to perform well in producing french-fried potatoes (101).
7.7. Lipid Hydrolysis (Fat Splitting)

Fats and oils are the feedstocks for the industrial production of fatty acids (102). In
addition to a major use in the production of bar soaps, these are also employed as
starting materials for the production of fatty amines, amides, alcohols, polyoxyethy-
lene esters, and other derivatives widely used in the nonfood industrial sector.
Water alone will hydrolyze fatty acid esters, and the greater its concentration in
the lipid phase, the greater its effectiveness in this regard. At temperatures below
boiling, the lipid solubility is too low to achieve the concentrations necessary to
promote hydrolysis. Under elevated pressures and with the addition of heat, condi-
tions are achieved where the lipid solubility of water is sufficient to achieve accep-
table rates of hydrolysis. Accordingly, high-temperature, high-pressure hydrolysis
reactions are the prevalent method of industrial fat splitting. In the United States,
194 ANIMAL FATS
the predominant technology, termed the ColgateEmory process, involves the con-
tinuous countercurrent flow of water and fat or oil at pressures of 4851 atm, and
approximately 260 C. Heated liquid lipid is introduced at the bottom of a vertical
cylindrical reactor. Heated water enters at the top. As the lipid charge rises through
the falling water charge under pressure, a continuous zone of high water solubility
in oil forms, below the bulk lipid layer and above the bulk aqueous layer, wherein
hydrolysis occurs. Effluent from the column is recovered, free fatty acids from one
outlet and an aqueous glycerol stream from the other. Although its high tempera-
tures can degrade polyunsaturated fatty acids, the ColgateEmery process is useful
for the hydrolysis of lipids such as most animal fats that have iodine values less
than 120.
Lipases, developed by nature for the hydrolysis of fatty acyl ester bonds, have
also been explored for the hydrolysis of fats and oils. However, because of issues of
cost, stability, and productivity, they are not presently employed in industrial lipid
hydrolysis.
8. ANTIOXIDANTS IN ANIMAL FATS
The oxidation of fatty acid double bonds is responsible for the generation of off
flavors and performance defects in lipids. Oxidation can be enzymatically catalyzed
by, for example, lipoxygenases produced by microbial contaminants. However,
nonenzymatic oxidation initiated by nonlipid contaminants is a more frequent
danger, as it initiates free radical reactions leading to oxidative decomposition of
the lipid. Light can also initiate lipid oxidation, by pathways that may or may
not involve free radical mechanisms. In this process, metals and other initiators trig-
ger the removal of a hydrogen from the allylic (i.e., adjacent to a double bond) posi-
tion of unsaturated fatty acids. The resulting fatty acid free radical can react with
oxygen to generate a peroxy radical that can react with other fatty acids, forming
hydroperoxides. Breakdown of these generates aldehydes, hydrocarbons, ketones,
and alcohols. In edible applications, these are perceived as rancid flavor. More
important to industrial applications, further reaction can lead to polymer formation
and deposition. To prevent such undesired degeneration, steps such as the removal
of air and blanketing with nitrogen are adopted to reduce fatty acid oxidation.
In the context of free radical oxidation mechanisms, antioxidants are compounds
able to quench lipid radicals, thereby terminating autooxidation (103). The most
popular natural antioxidants are the tocopherols (Vitamin E), which are commer-
cially available for use in this application. Their endogenous levels in animal fats
are low, generally two to three orders of magnitude, than in vegetable oils and fats,
even after refining and hydrogenation of the latter (104). The tocopherol contents of
tallow and lard range from 7 to 27 mg/kg (105). a Tocopherol is the prevalent isomer
present, representing 90.7%, 94.6%, and 69.8% of the total tocopherols in the body
fat of beef, lamb, and pork, respectively (104, 106). Increasing the level of toco-
pherols fed to meat animals has increased the tissue levels of tocopherol, and it
has provided some protection against fatty tissue lipid oxidation, even in pig fat,
CHARACTERISTICS OF ANIMAL FAT-BASED SHORTENINGS AND FRYING FATS 195
with its relatively elevated levels of unsaturated fatty acids compared with beef
tallow (10).
Lard and tallow respond well to the addition of antioxidants, and numerous stu-
dies have been conducted to evaluate the protection afforded to lard by various
antioxidants and metal sequestering agents, such as citric acid. These involved
either tocopherols or synthetic antioxidants. In the latter category, ethoxyquin,
butylated hydroxyanisole (BHA) or butylated hydroxytoluene (BHT) and tertiary-
butylhydroquinone (TBHQ) are effective in animal fats. The latter is approved for
use in the United States and Canada, but not in Japan, the EU, and elsewhere. In
lard, the tocopherols increase in effectiveness in the order: a, b, g, and d (107). A
concentration of 0.02% g tocopherol was reported to be more effective than the
same concentration of BHA or BHT in chicken, pork, and beef fat (108). The oxi-
dative stability of lard was improved by the addition of up to 250-ppm tocopherol,
whereas for BHA and BHT, the optimum effective dose was 200 ppm (66). Toco-
pherols also have the advantage of carry-through properties in baked and fried pro-
ducts prepared from lard, which will reduce oxidative instability in these products
(109). In addition, the protective effects of major synthetic antioxidants on lard, and
their carry-through effects in products made with lard, have been demonstrated
(63). The best combinations of antioxidants and chelating agents for use in animal
fats for particular applications has been reported (46). Natural antioxidants other
than tocopherols also afford protection to animal fats (111, 112).
9. CHARACTERISTICS OF ANIMAL FAT-BASED SHORTENINGS

AND FRYING FATS
Although their use in home baking and cooking in the United States has declined
greatly in recent years, animal fat-based shortenings are to some extent used in
industrial baking and in baking and deep fat frying elsewhere worldwide. Tradition-
ally, the solids content, crystal structure, and working characteristics of lard made it
the shortening of choice for pie crust. Vegetable shortenings can, however, be for-
mulated to have characteristics similar to those of lard, and these shortenings have
more favorable nutritional characteristics such as less saturated fat and no choles-
terol. Deodorized and stabilized lard and tallow are examples of the lowest cost
shortenings suitable for cookies (113).
The physical and textural characteristics of North American vegetable and ani-
mal fat shortenings have been compared (113). Selected data from this study are
presented in Table 11. There was not much difference in the vegetable and animal
fat shortenings as far as melting and crystallization temperatures were concerned,
but the polymorphic forms were different. The animal fat shortenings were mostly
in the b form, except for a tallow and a tallow-vegetable frying fat. The vegetable-
based shortenings were mostly in the b0 form. The texture of several of the meat-
vegetable blend shortenings was comparable with that of the vegetable-palm short-
enings, although the meat-vegetable samples had a higher solids content. The lards
had high values for degree of deformation at breaking force, whereas the tallows
196 ANIMAL FATS
TABLE 11. Selected Characteristics of Meat-Fat or Meat-Vegetable Fat Shorteningsa

(reprinted by permission of AOCS).
SFC (%) at
Hardness 20 C
Sample and Dropping Softening Polymorphic Index (Tempered at Air
Source Point C Point C Form (g/min) 30 C) Content
Shortening
Lard, U.S. 45.2 44.4 b 4.8 26.9 2.3
Lard, Canada 40.6 41.9 b > b0 5.1 26.2 6.2
Lard, Canada 38.2 37.7 b 4.5 25.3 10
Tallow-lard, 42.7 42.3 b b0 11.8 27 8.8
Canada
Tallow-lard, 42.3 43.7 b 14.6 28.7 9
Canada
Meat-vegetable, 44.6 45.8 b0 > b 5.4 25 19.5
U.S.
Meat-vegetable, 45.1 46 b0 b 5.9 26.9 21
U.S.
Vegetable-tallow, 50.6 50.7 b 8.1 26.6 4
Canada
Frying Fat
Tallow, Canada 45.8 46.4 b0 8.9
Tallow-vegetable 44.8 44.7 b0 6.9
Canada
a
See (113).
were not as pliable. Also, shortenings containing high levels of palm oil were able
to withstand large deformations without breakage. The tallows and tallow-lard
blends were very hard.
As in baking, animal fats were the item of choice for deep frying of foods in the
United States until the 1980s. Meat fats exhibit good stability and have generally
been economical to use. The flavors they impart have been considered desirable for
some foods, as in the flavor added to french-fried potatoes by tallow and to pie
crusts by lard. However, concerns regarding the relationship of dietary cholesterol
and saturated fatty acid to coronary heart disease have caused the replacement of
animal fats with unhydrogenated vegetable oils in U.S. deep fat frying. In other
countries, animal fats are still used in frying, such as in the United Kingdom where
the use of both tallow and lard is reported (114).
With the emphasis in some parts of the world on a reduction in the dietary con-
sumption of animal fats, the possibility of using blends of animal and vegetable fat
has been explored. It has been shown that such blends confer some beef-like flavor
notes on fried foods, and that foods fried in straight vegetable oils lack the charac-
teristic flavors imparted by beef tallow. It is presently unclear whether the use of
such animal-vegetable fat blends in frying will be widely adopted.
Attempts have been made to extract and concentrate beef fat volatiles using
supercritical carbon dioxide (115). Total volatiles were concentrated over controls
RECENT DEVELOPMENTS 197
by 10100 fold, with the lowest pressure extraction conditions yielding the highest
concentration of volatiles. Similarly, it has been shown that the flavor volatiles of
heated pork fat can be fractionated with supercritical carbon dioxide (116).
The cholesterol present in tallow used for frying undergoes oxidative changes
(117, 118), and the generated products are found in fried foods. Thus, the presence
of cholesterol oxides have been demonstrated in french-fried potatoes at concentra-
tions approximately four times as high as those that existed in the heated tallow
used for frying (119, 120), although some of the values may have included contri-
butions from oxidized plant sterols (121).
10. RECENT DEVELOPMENTS
Since publication of the most recent prior edition of this chapter in 1996 (122), glo-
bal events, scientific discoveries, and technological developments have impacted
the real and potential use of animal fats in a number of areas. These should continue
to affect lipid usage into the future. This section presents an introduction and over-
view of these developments.
10.1. Bovine Spongiform Encephalopathy (BSE)

In 1985, a new disease appeared in cattle, an unavoidably fatal neurological degen-
eration characterized by weight loss, a decline in milk yield, difficulty in walking,
and a nervous appearance. Anatomically, it was shown to result in the formation of
small holes throughout the brain, making it appear sponge-like, and leading to the
name bovine spongiform encephalopathy, or BSE (123, 124). Because of its symp-
toms, the affliction is known as mad cow disease. The disease was first detected in
the United Kingdom, and since the beginning of the outbreak, over 180,000 cattle
have contracted and died from it there (125). This constitutes over 99% of known
cases to date. The remaining occurrences have largely, but not exclusively, been
limited to Europe. Pigs and poultry are not infected.
Although other theories for the cause of mad cow disease have been proposed,
the most accepted one currently is that it is caused by a biochemically novel (and in
many ways astounding) infectious agent termed a proteinaceous infectious particle,
or prion (126, 127). Prions consist largely of protein, with minor amounts of sphin-
golipid and polysaccharide. They are not living organisms and in fact lack genetic
material. Among other prion-mediated diseases are scrapie in sheep and
CreutzfeldtJacob disease in humans. Prions are derivatives of natural body pro-
teins synthesized and located in the brain of the host animal and normally serving
their functions without incident. These proteins can undergo a change in their three-
dimensional conformation, assuming the prion configuration, which renders them
resistant to degradation and removal by normal body housekeeping machinery.
They thus accumulate in the brain, causing the degeneration characteristic of the
disease. This conformation change can occur spontaneously or as a result of genetic
damage, and thus it occurs at a low rate in older animals.
198 ANIMAL FATS
Prions also have the ability to trigger conformational change in their normal
counterpart proteins in the brain, causing these to also adopt the improper config-
uration. In this fashion, the prion can essentially replicate itself. Because the con-
formational structure assumed during prion formation is thermodynamically stable,
prions are exceedingly resistant to inactivation (128, 129). In addition, they can be
transmitted orally. The effect of these traits is that the disease is not solely depen-
dent on spontaneous generation to infect an individual or population, but it also can
be acquired by consumption of infected tissue.
These features, the fact that the disease typically takes 3 to 5 years to manifest
itself, and the industry practice of capturing the nutritional and economic value of
carcasses by feeding unused components, in the form of meat and bone meal, to
succeeding generations of cattle combined to lead to widespread infection within
the U.K. population when the disease did appear.
Before the development of full awareness of the BSE epidemic, infected animals
continued to enter the human food supply. The degree of urgency of efforts to
understand and control the epidemic was increased with the description, in 1996,
of a similarly debilitating and fatal human disease, termed variant Creutzfeldt
Jacob disease (vCJD) that appears to be the result of transmission of BSE to
humans (130 132). As of late 2002, 117 humans were known to have died from
the disease in the United Kingdom (133). Because of the threat not only to the
meat supply but also to human health, massive action was taken to halt the spread
of BSE, including its designation as a reportable animal disease, the destruction of
no less than 3 million head of cattle in the United Kingdom alone, and the adoption
of regulations forbidding the feeding of cattle- and sheep- derived material to rumi-
nants. In the 1990s, bans on the feeding of ruminant animal parts to ruminants were
enacted in the Unied Kingdom and to some degree in the other European countries.
In the United States and Canada, a ban restricting the feeding to ruminants of mate-
rials from any mammalian source (with some exceptions, including material of
porcine and equine origin) was enacted in 1997 (134). In December 2000, EC Reg-
ulation 2000/766 came into force, banning the use of meat and bone meal in all
feeds. This came about after BSE was found in Germany and Spain for the first
time. As a result of such regulatory changes, and new surveillance policies, the inci-
dence of BSE has fallen from a high of more than 37,000 reported cases in 1992, all
located in the United Kingdom, to a world total of 2179 cases (1144 of these in the
United Kingdom) in 2002 (135). Most incidents of BSE have been confined to
Europe. Single or double cases detected in Israel, Greece, Canada, the United States,
the Falklands, Oman, and Japan are in many cases attributable to the importation of
U.K. beef or beef products. In mid-2003, Canada reported its first case of the dis-
ease in a reportedly native-born animal, an 8-year-old cow. Given the lengthy incu-
bation period of the disease, it is possible that this animal was infected early in its
life by contaminated imported meat and bone meal. In December of that same year,
an infected animal was detected in the United States. Subsequent investigations
determined that this individual had been imported from Canada.
To further combat the spread of the disease, European regulations categorize ani-
mal-based raw material into three classes and establish the appropriate uses for each
class. Only materials fit for human consumption are allowed for food, feed, and
oleochemical uses. Nonedible material free of specific risk material (SRM: skull,
eyes, brain, and spinal cord of cattle, sheep, and goats) can be used for industrial
chemical technical uses. Finally, dead stock and SRM tissue may not be used, and
must be destroyed. In addition, the feeding of animal proteins to animals of the
same species is banned, and catering waste, which includes waste restaurant fats
and oils, cannot be used as animal feed (136). Furthermore, only edible raw mate-
rial can be used to produce tallow for food, animal feed, fertilizer, and cosmetic
products (137).
In an attempt to prevent the further international spread of the disease, importa-
tion of animals from infected countries has been banned. Thus, for example, the
United States in 1989 banned the importation of live ruminants and most ruminant
products from the United Kingdom and other countries having BSE. The ban was
extended to European products in 1997 after the discovery of the disease in some
countries there.
The causes of the BSE epidemic have not been established with certainty.
Scrapie is a similarly prion-mediated disease of sheep, known and present in
U.K. sheep for centuries. Factors contributing to the BSE outbreak in cattle are sus-
pected to be the inclusion of scrapie-infected sheep material in the rendering stream
and industry changes in the United Kingdom during the 1970s and 1980s that
reduced the rigor of heat treatment during rendering, thus (perhaps) allowing the
BSE causative agent to persist in the resulting meat and bone meal and infect
healthy cattle. In combating the disease, regulations were put in place in the United
Kingdom that call for the processing of meat and bone meal at 133 C and 3-atm
pressure for 20 minutes, which is believed to inactivate the BSE prion. It is unclear
how such treatment will affect the quality of the resulting tallow.
Meat and bone meal has been identified as a vector for transmission in the BSE
outbreak. Tallow is also an animal-derived product, is produced largely from cattle,
and is a coproduct of MBM production. Thus, concerns existed regarding its health
status. After an examination of existing data, the Scientific Steering Committee of
the European Commission concluded that normal industrial tallow production
processes result in a product that is free of detectable BSE infectivity, even if the
source material was highly infective (138). In another study, the rates of prion inac-
tivation during conventional oleochemical processing were determined and used to
estimate that the risks of human infection caused by consumption of oleochemical
products of bovine origin subjected to hydrogenation or high-temperature
high-pressure hydrolysis were less than the spontaneous rate of appearance of
CreutzfeldtJakob disease (139). The European Commission considers tallow and
its derivatives to be safe. The U.S. Food and Drug Administration has ruled that
tallow and other rendered fats are safe, and it specifically omitted them from reg-
ulations prohibiting rendered products in feeds for cattle and other ruminants (140).
The United Nations World Health Organization (WHO) has examined the issue and
has concluded that because prions are proteinaceous, they would partition with the
cellular residues of meat and bone during processing. The tallow fraction was there-
fore judged not a risk to human or animal health (141).
200 ANIMAL FATS
As animal fats are a potential feedstock for biodiesel production, Cummins et al.
(142) assessed the danger of a human contracting CJD as a result of the use of
tallow as a fuel in diesel engines. They concluded that the risk was several orders
of magnitude less than the rate of spontaneous appearance of CJD. Thus, scientific
analysis indicates that processed (i.e., rendered) animal fat is not an agent of trans-
mission of BSE. Nonetheless, especially in the United Kingdom, the public remains
skeptical. This has in some cases led to less use of animal fats in feed applications.
Especially in the United Kingdom, the BSE epidemic has reduced the amount of
domestically available tallow (because of condemnation) and increased the use
of other lipids in place of animal fats.
The discovery of single infected animals in Canada and in the United States in
2003 reinvigorated the discussion of whether there should be a total ban on the use
of mammalian products in animal feeds. Although the BSE situation seems to be
under control at this time, new outbreaks could considerably impact the availability
and the allowed uses of animal fats.
10.2. Use of Fats and Oils as Fuels

Sales restrictions by some petroleum exporting countries in the 1970s stimulated
research in the oil-importing countries into the development of non-petroleum fuels
for internal combustion engines. This led, in the case of compression ignition (die-
sel) engines, to a resurgence of interest in the use of vegetable oils and animal fats
as fuels. Although Rudolph Diesel himself demonstrated more than a century ago
that his engine would run on vegetable oils, the rise of the petroleum industry had
largely extinguished research on this topic. Intact fats and oils are generally unsui-
table as neat diesel engine fuels without fuel system modification, because they can
lead to engine failure. However, conversion of the fatty acids in fats and oils to their
simple akyl esters, now known as biodiesel, produces a fuel that functions well in
unmodified engines, while substantially reducing the undesirable emission of
unburned hydrocarbons, particulates, carbon monoxide, and sulfur dioxide (143).
In addition, biodiesel reduces net emissions of carbon dioxide (a greenhouse
gas responsible for global warming), reduces dependence on imported energy
sources, increases agricultural income, and increases fuel lubricity, thereby redu-
cing wear in close-tolerance fuel injection systems. The removal of sulfur from pet-
roleum diesel fuels will soon be implemented in some countries, including the
United States. This process removes lubricity agents naturally present in the fuel.
The enhanced lubricity properties of biodiesel may thus stimulate its greater use in
the future.
Biodiesel is presently making the transition in many countries of the world from
a research curiosity to an accepted alternative to petroleum-based fuel. Europe is
the leading region for the production and use of biodiesel, with an estimated
2001 output of 757-million L. Biodiesel production in the United States in 2001
was estimated at 79.5-million L. Production and consumption are rapidly increasing
worldwide, with estimates of combined US and European output in 2003 at around
1,628 million liters (144). Present and anticipated use constitutes but a fraction of
total fuel consumption, and it does not eliminate the use of petroleum-based fuels.
However, anticipated vigorous future growth in this area has the potential to
consume considerable volumes of fats and oils.
The predominant feedstocks for biodiesel production to date have been refined
vegetable oils (rapeseed in Europe, soybean in the United States), although it
appears that in Britain a subsidy program has had the effect of promoting the use
of tallow and greases over vegetable oils (144). The production of biodiesel from
animal fats has been reported (145148), and the suitability of tallow-based biodie-
sel as an engine fuel has been demonstrated (149151). Relative to biodiesel
produced from vegetable oils, animal fat-based biodiesel offers the advantages of
reduced raw material cost, increased cetane value (which improves engine perfor-
mance), and greater oxidative stability. As raw material costs can contribute over
70% to the cost of biodiesel, the former advantage could be significant. On the other
hand, the greater content of saturated fatty acids in animal fats raises the melting
points of biodiesels made from them. For example, soy methyl ester has pour and
cold points of 2 C and 0 C, respectively, whereas the corresponding values for
tallow methyl esters are 15 C and 17 C (147). In cold climates, this may lead to
engine inoperability because of the plugging of fuel lines and filters. For heavy
duty diesel engines, this is of concern only during nonoperational periods, because
once the engine is running, the fuel recirculation loops keep the fuel fluid by cir-
culating it through the warm engine. However, perhaps due to concerns regarding
cold weather performance, the use of animal fats as a biodiesel feedstock has been
negligible to date. As demand increases in the future, fat-based feedstocks may be
more widely adopted, with such approaches as the use of low blend rates in petro-
leum diesel, the addition of freezing point depressants, the production of the lower
melting esters of branched chain alcohols (147), or the blending of animal- and
vegetable-biodiesels being taken to avoid low-temperature performance problems.
Waste greases, largely consisting of spent deep fat frying oils and fats, can also
be used as feedstocks for biodiesel production (152156), and the engine emissions
of such fuels are comparable or superior with those of soy oil-based biodiesel (149,
153, 154). Acceptable engine wear and performance were obtained during extended
use of grease-based biodiesel blends with petrodiesel in a heavy duty truck (157).
As they are generally priced at about half the cost of refined edible oils, waste
greases are an attractive feedstock. However, greases contain a higher content of
free fatty acids than refined oils. These reduce the level of available alkaline cata-
lyst in conventional transesterification reactions via the production of fatty acid
soaps. Soaps can foster problematic emulsion formation during product washing.
Thus, either the soaps must be isolated and discarded, representing a loss of poten-
tial feedstock and catalyst, or an acid-catalyzed esterification of free fatty acids
must be adopted upstream of the conventional alkali-catalyzed acylglycerol
transesterfication reaction. Either way, a more complicated process is required to
produce biodiesel from waste greases than from refined oils (158).
As lipases are able both to esterify free fatty acids and catalyze the alcoholysis of
acylglycerols, they can in theory eliminate the need for multiple catalyst systems
during biodiesel production from waste greases. Several groups have investigated
202 ANIMAL FATS
lipase-catalysis for the production of simple fatty acid esters for use as biodiesel
from waste grease, as well as from animal fats (159161). In some of these reports,
esterification efficiencies in excess of 95% were described. Nonetheless, enzymatic
catalysis has yet to be developed to the stage where it is efficient and economical for
industrial use in biodiesel production.
As a consequence of increased biodiesel production, increasing amounts of
byproduct glycerol will enter the market. There is some concern that this will
depress glycerol prices. As glycerol is a valuable coproduct of the splitting of
fats and oils to produce free fatty acids, this could negatively impact the economics
of fat splitting. For this reason, glycerol utilization research is becoming a priority,
although the industrial and market impact of such work has yet to be realized.
Another potential expansion in the use of animal fats (and vegetable oils) as
fuels comes in their use to fire burners for steam and hot water production. In
this situation, it is not necessary to convert the lipids to their simple alkyl esters,
as must be done for use in internal combustion engines. Rather, the intact acylgly-
cerols are used as fuels. They can be acceptable in this application because indus-
trial boiler burners are both simpler and more robust than internal compression
engines, and they are also subject to less emissions monitoring and fewer emissions
regulations. During some winters near the recent turn of the century, agricultural
lipid prices in the United States were in some cases competitive with that of fuel
oil. This led to a number of unscientific, and largely unpublicized, tests of lipids as
industrial boiler fuels. At least one scientific report of such an investigation has
been released, demonstrating that tallow, white and yellow greases, and chicken
fat were technically and economically viable fuels, either neat or in blends with
U.S. No. 2 fuel oil, in an industrial boiler (162). The direct burning of lipids is
more common in Europe, and to the extent that it is adopted more widely could
represent a new lipid outlet, especially for low-value animal fats.
Increased consumption of agricultural lipids as fuels, either for boilers or diesel
engines, could increase their prices and curtail their availability, impacting their use
in traditional applications. All these changes are dependent on the future course
of petroleum fuel prices, on research advances to reduce biofuel production costs,
and/or on the establishment and continuation of government programs providing
financial incentives that foster the use of renewable fuels.
10.3. Trans-Fatty Acids

The latter half of the twentieth century saw the use of hydrogenation technology to
decrease the degree of unsaturation in vegetable oils. This converts the oils to fats
(solids at room temperature) and increases their oxidative stability. Hydrogenated
fats have been widely adopted in place of solid animal fats in such applications as
table spreads, as is illustrated by the increasing consumption of vegetable-based
lipids as margarines, as shortenings, and as frying fats for the production of deep
fried pastries, potatoes, corn chips, chicken, and so on. In addition, as yellow grease
is employed as a component of animal feeds, hydrogenated fatty acids may be
incorporated intact into meat, tallow, mutton, lard, poultry fat, and so on that
are subsequently consumed by humans. The degree of hydrogenation in these

products is typically less than complete, because desired performance is achieved
with partial saturation. One outcome of this partial saturation is the conversion of a
proportion of the double bonds of the fatty acids from their original cis-form to the
trans-configuration. It has become clear that trans-fatty acids are not metabolically
equivalent to their cis-counterparts, but rather exhibit effects more similar to those
of saturated fatty acids. Trans-fatty acids raise concentrations of total serum cho-
lesterol, triacylglycerols, LDL cholesterol, and lipoprotein a, and they may lower
HDL cholesterol (163165). All of these effects increase the risk of coronary
heart disease. Although ruminant fat also contains trans-fatty acids, the levels are
substantially lower than are present in industrially hydrogenated lipids. It has been
estimated that this increased consumption of trans-fatty acids results in 30,000
premature deaths annually in the United States alone (164).
In response to growing awareness in the scientific community and the populace
at large of the negative health consequences of trans-fatty acid consumption, indus-
try has refined hydrogenation catalysts and processes to reduce the production of
trans-double bonds (166). Some major lipid-producing firms have voluntarily
reduced or eliminated trans-fatty acids from their product lines. Following the
conclusion by the Danish Nutrition Council that trans-fatty acids contribute to car-
diovascular disease, to reduced fetal birthweight, and to the development of geria-
tric (type 2) diabetes, the Danish government as of June 2003 banned the sale of fats
and oils with anthropogenically generated trans-fatty acid contents greater than 2%.
This prohibition was extended to the oils and fats in processed foods in December
2003. In July 2002, the U.S. National Academy of Science issued a report conclud-
ing that the only safe intake of trans-fat is zero. Also in the United States, the
trans-fatty acid contents of foods are due to be listed on the ingredients label as of
January 2006. Actions such as these may renew interest and research in the devel-
opment and use of animal fats, either neat or after mixing or interesterficiation with
vegetable oils, in foods. This would trigger an increase in the consumption of
animal fats.
10.4. Conjugated Linoleic Acid

In general, multiple double bonds in the naturally occurring fatty acids are sepa-
rated by one or more carbon atoms that are not involved in a double bond. Thus,
for example, the two double bonds in the prevalent (a-) isomer of linoleic acid
occupy the 9, 10 and 12, 13 positions (numbering from the carboxyl terminus).
The no. 11 carbon, separating these unsaturations, is a methylene carbon (CH2),
being singly bonded to its neighboring carbons. It is also possible for two double
bonds to be adjacent to one another, with no intervening methylene carbon, a con-
dition known as conjugation (Figure 3). Such double bonds may be of the cis-or
trans-configuration.
Fatty acids possessing conjugated double bonds occur in biological lipids,
although their frequency is low. Conjugated linoleic acids (CLAs) are among the
most common of the naturally occurring conjugated fatty acids (167, 168). CLA
204 ANIMAL FATS
H H H H
C C C C
Conjugated
H H H H H
C C C C C
Non-conjugated
Figure 3. Conjugated and nonconjugated carboncarbon double bonds. (Used with the kind
permission of the Institute of Shortening and Edible Oils, Inc. Washington, D.C.)
is rare to nonexistent in the lipids of plants and most monogastric animals, gener-
ally constituting 0.010.08% of the total lipid on a weight basis (169). In ruminant
fats, however, CLA levels exceed these amounts by 10-fold (170). The most com-
mon of the ruminant CLAs is the cis-9, trans-11 isomer, known as rumenic acid,
which constitutes upward from 85% of the total CLA content. In ruminants, two
primary origins for CLA have been identified: as an escaped intermediate of the
biohydrogenation of dietary linoleic and linolenic acids, and via delta-9 desatura-
tion of vaccenic acid (trans-11 octadecenoic acid) generated in the rumen from
dietary linoleic and linolenic acids (171). In dairy cows, the desaturase pathway
is believed to be the predominant source of CLA.
Studies begun in the 1970s to identify carcinogenic chemicals in common foods
found instead that fried hamburger actually contained an antimutagenic agent
(172). This was identified as cis-9, trans-11 conjugated linoleic acid, or c-9, t-11
CLA, rumenic acid (173). Since then, extensive research has validated and extended
these observations in animals and humans (174). Minor amounts of other CLAs are
also found in ruminant lipids. The physiological effects of the trans-10, cis-12 iso-
mer have been studied and have shown to include the reversal of obesity in test
animals, and possibly in humans (175). Decreased atheroschlerosis, improved
hyperinsulinemia in prediabetic rats, and potentiation of the immune response
have also been observed as a consequence of the consumption of CLAs.
Given the current widespread interest in reducing cancer, obesity, and other
maladies, there is considerable interest in the use of the CLAs, either as a mixture
or in the form of individual isomers, as beneficial dietary adjuncts. Cows milk, beef
tallow, and products made from them are natural sources of CLA. However, CLA is
also readily synthesized in high yield in the laboratory from vegetable oils that are
rich in linoleic acid, such as sunflower and safflower. The resulting synthetic pro-
duct has CLA levels of about 80%, not the 0.30.5% (fat basis) found in beef tallow
and dairy products (176). As a result, except for studies of the specific effects of
foods containing CLA, vegetable oil is the typical source of CLA in contemporary
studies and in commercial dietary supplements. This trend will probably continue.
REFERENCES 205
Although it might trigger some increased consumption of dairy products and meats,
it seems unlikely that the continued and increasing interest in CLA will translate to
increased use of animal fats.
A continuously updated listing of publications relating to CLA can be found at
www.wisc.edu/fri/clarefs.htm.
ACKNOWLEDGMENTS
The author thanks Jane Love, Iowa State University, for providing essential back-
ground information, and Neville Chandler, Deborah Dempsey, Frank Gunstone,
Gary Pearl, Ray Rouse, John Starkey, and Fred Wellons for generously providing
information included in this chapter. In addition, Drs. Gunstone, Pearl, and Wellons
read drafts of the manuscript. Their comments are greatly appreciated. Mention of
trade names or commercial products in this publication is solely for the purpose of
providing specific information and does not imply recommendation or endorsement
by the U.S. Department of Agriculture.
REFERENCES
1. B. Rossell, ed., Animal Carcass Fats, vol. 2, Oils and Fats Series, Leatherhead Publish-
ing, Leatherhead, U.K., 2001.
2. F. D. Gunstone, and F. B. Padley, Lipid Technologies and Applications, Marcel Dekker,
New York, 1997.
3. C. C. Akoh and D. B. Min, eds., Food Lipids; Chemistry, Nutrition, and Biotechnology,
2nd ed., Marcel Dekker, New York, 2002.
4. M. Bockisch, Fats and Oils Handbook, AOCS Press, Champaign, IL, 1998.
5. Darling International, Irving, TX. (2003) http://www.darlingii.com.
6. Darling International. (2003). Available: http://www.darlingii.com/askross/Faq.htm
Fat8.
7. F. T. Orthoefer, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, 5th ed., Wiley,
New York, 1996, pp. 1943.
8. Institute of Shortening and Edible Oils. (1999). Food Fats and Oils, 8th ed. Available:
http://www.iseo.org.
9. M. Enser, in J. B. Rossell and J. L. R. Pritchard, eds., Analysis of Oilseeds, Fats and Fatty
Foods, Elsevier Applied Science, New York, 1991, pp. 329394
10. F. J. Monahan, D. J. Buckley, P. A. Morrissey, P. B. Lynch, and J. I. Gray, Meat Sci. 31,
229241 (1992).
11. L. C. St. John, C. R. Young, D. A. Knabe, L. D. Thompson, G. T. Schelling, S. M.
Grundy, and S. B. Smith, J. Animal Sci., 64, 14411447 (1987).
12. C. A. Morgan, R. C. Noble, M. Cocchi, and R. McCartnery, J. Sci. Food Agric., 58, 357
368 (1992).
13. D. K. Larick, B. E. Turner, W. P. Schoenherr, M. T. Coffey, and D. H. Pilkington, J. Anim.
Sci., 70, 13971403 (1992).
206 ANIMAL FATS
14. I. W. McDonald and T. W. Scott, World Rev. Nutr. Diet., 26, 144207 (1977).
15. D. C. Rule and D. C. Beitz, J. Amer. Oil Chem. Soc., 63, 14291436 (1986).
16. A. M. Pearson, J. D. Love, and F. B. Shoreland, in C. O. Chichester, E. M. Mrak, and
G. F. Stewart, eds., Advances in Food Research, vol. 23, Academic Press, New York,
1977, pp. 274.
17. J. D. Wood, in J. Wiseman, ed., Fats in Animal Nutrition, Butterworth, London, 1984,
pp. 407 435.
18. J. A. Marchello, D. A. Cramer, and L. J. Miller, J. Anim. Sci., 26, 294297 (1967).
19. M. Enser, Production, Properties and Uses of Edible Oils and Fats in the Food Industry.
Animal Fats: Chemical and Physical Properties and Applications/ Training Course
Notes, Course 011E1, Leatherhead Food RA, Leatherhead, U.K., 1996.
20. W. E. Connor and S. L. Connor, Current Atheroschlerosis Rep., 4, 425432 (2002).
21. Anonymous, Nutritional Aspects of Cardiovascular Disease, Report on Health and
Social Subjects, British Department of Health, London, U.K., 1994.
22. Anonymous, High Blood Cholesterol: Detection, Evaluation, Treatment, Third Report of
the (U.S.) National Cholesterol Education Program (NCEP) on the Detection, Evaluation
and Treatment of High Blood Cholesterol in Adults, National Cholesterol Education
Program, National Heart Lung and Blood Institute, National Institutes of Health,
Bethesda, Maryland. NIH Publication 025215. (2002). Available: http://www.nhlbi.
nih.gov/guidelines/cholesterol/atp3full.pdf
23. R. Sieber, Lebensmittel Wissenschaft und Technologie, 26, 375387 (1993).
24. R. D. Kiley, C. T. Massie, A. L. Bachman III, and V. P. Wagher, Lipid Technol., 8, 510
(1996).
25. R. R. Chao, S. J. Mulvaney, and H. Huang, J. Amer. Oi1 Chem. Soc., 70, 139143 (1993).
26. W. W. Christie, in R. J. Hamilton and J. B. Russell, eds., Analysis of Fats and Oils,
Elsevier Science, New York, 1986.
27. H. Brockerhoff, R. J. Hoyle, and N. Wolmark, Biochim. et Biophys. Acta, 116, 6772 (1966).
28. W. C. Byrdwell, E. A. Emken, W. E. Neff, and R. O. Adlof, Lipids, 31, 919935 (1996).
29. H. R. Mottram, Z. M. Crossman, and R. P. Evershed, Analyst, 126, 10181024 (2001).
30. H. Kallio, K. Yli-Jokipii, J.-P. Kurvinen, O. Sjovall, and R. Tahvonen, J. Agric. Food
Chem., 49, 33633369 (2001).
31. W. E. Neff, W. C. Byrdwell, K. R. Steidley, G. R. List, and G. Snowder, J. Liquid
Chromatogr. Related Technol., 25, 985998 (2002).
32. A. Thomas, in W. Gerhartz, ed., Ullmans Encyclopedia of Industrial Chemistry, 5th
ed., vol. A10, Ethanloamins to Fibers, VCH Publishers, New York, 1987, Table 22,
p. 215.
33. J. Wang, R. Berthollet, P. Ducret, and M. Fleith, U.S. Patent 6,034,130, (1998).
34. K. Kennedy, M. S. Rewtrell, R. Morley, R. Abbott, and P. T. Gyinlan, Am. J. Clin. Nutr.,
70, 920927 (1999).
35. T. Yang, X. Xu, C. He, and L. Li, Food Chem., 80, 473481 (2003).
36. P. J. Lawler and P. S. Dimick, in C. C. Akoh and D. B. Min, eds., Food Lipids; Chemistry,
Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002.
37. N. Garti and K. Sato, eds., Crystallization and Polymorphism of Fats and Fatty Acids,
Marcel Dekker, New York, 1988.
REFERENCES 207
38. Anonymous, Color; FAC Standard Color, AOCS Official Method Cc 13a-43, Official
Methods and Recommended Practices of the AOCS, 5th ed. and updates, American Oil
Chemists Society, Champaign, Illinois, 1997.
39. Anonymous, Color; Wesson Method Using Color Glasses Calibrated in Accordance with
the AOCS-Tintometer Color Scale, AOCS Official Method Cc13b-45, American Oil
Chemists Society, Champaign, Illinois, 1997.
40. Anonymous, Color; Lovibond Method Using Color Glasses Calibrated in Accordance
with the Lovibond Tintometer Color Scale, AOCS Official Method Cc13e-92, American
Oil Chemists Society, Champaign, Illinois, 1997.
41. Anonymous, Code of Federal Regulations, Part 319 Definitions and Standards of
Identity or Composition, Chapter III Food Safety and Inspection Service, Department
of Agriculture, Title 9 Animals and Animal Products, U.S. Government Printing
Office, Washington, D. C. (2003). Available: http://www.access.gpo.gov/nara/cfr/wai-
sidx_03/9cfr3319_03.html.
42. Anonymous, Codex Standard for Named Animal Fats, Codex-Stan 2111999. Food and
Agriculture Organization of the United Nations, World Health Organization, Rome, Italy
(1999). Available: http://www.codexalimentarius.net/standards_search.asp
43. C. Paquot and A. Hautfenne, in Determination of the Bomer Value (B.V.), Section 2.131,
Standard Methods for the Analysis of Oils, Fats and Derivatives, 7th ed. revised and
enlarged, Blackwell Scientific Publications, Oxford, 1987, pp. 4448.
44. R. S. Kirk and R. Sawyer, Pearsons Composition and Analysis of Foods, 9th ed.,
Langman Scientific and Technical, Essex, U.K., 1991, pp. 611612.
45. M. L. Meara, in R. I. Hamilton and A. Bhati, eds., Fats and Oils: Chemistry and
Technology, Applied Science, London, 1980, p. 196.
46. L. R. Dugan Jr., in J. F. Price and B. S. Schweigert, eds., The Science of Meat and Meat
Products, 3rd ed., Food and Nutrition Press, Inc., Westport, Connecticut, 1987, pp. 507
529.
47. American Fats & Oils Association, Available: http://www.afoaonline.org.
48. The Dupps Co. Available: http://www.dupps.com.
49. Anonymous, BS3919:1987 Specification for technical tallow and animal grease, BSI
British Standards Headquarters, London, U.K. (1987). Available: http://www.bsi-global.com
50. U.S. Government Printing Office. (2002). Agricultural Statistics. Available:http://
www.usda.gov/nass/pubs/agroz/oz_ch3.pdf, and previous years 1970, 1980, 1990,
1994, 1997.
51. INFORM, 5, 715 (1994).
52. R. B. Hu, M. J. Stampfer, J. E. Manson, A. Ascherio, G. A. Colditz, F. E. Speizer,
C. H. Hennekens, and W. C. Willett, Am. J. Clin. Nutr., 70, 100108 (1999).
53. S. Renaud and D. Lanzmann-Petithory, Curr. Atheroscl. Res., 4, 419424 (2002).
54. W. E. Connor and S. L. Connor, Curr. Atherosl. Res., 4, 425432 (2002).
55. Agricultural Statistics, U.S. Government Printing Office, Washington, D. C., 1994.
56. See Schnell Publishing, Chem. Mark. Rep., 257 (2000).
59. C. Watkins, Inform, 12, 580587 (2001).
208 ANIMAL FATS
ing, Leatherhead, U.K., 2001, pp. 79.
61. J. Starkey, unpublished information, U.S. Poultry and Egg Association. Available:
http://www.poultryegg.org
62. R. H. Rouse, in Multi-State Poultry Feeding and Nutrition Conference, Indianapolis,
Indiana, 1998.
63. M. Bockisch, Fats and Oils Handbook, AOCS Press, Champaign, IL, 1998, pp. 131138.
64. G. Pietroszek, The National Provisioner, June 27 p. 819 (1987).
65. N. O. V. Sonntag, in D. Swern, ed., Baileys Industrial Oil and Fat Products, vol. 2, 4th
ed., Wiley & New York, 1979, pp. 289 477.
66. H. B. W. Patterson, Bleaching and Purifying Fats and Oils: Theory and Practice, AOCS
Press, Champaign, Illinois, 1992.
67. INFORM, 1, 638644 (1990).
68. W. T. Koetsier, in F. D. Gunstone, and F. B. Padley, Lipid Technologies and Applications,
69. A. Rozendaal and A. R. Macrae, in F. D. Gunstone, and F. B. Padley, Lipid Technologies
and Applications, Marcel Dekker, New York, 1997.
70. D. Rousseau and A. G. Marangoni, in C. C. Akoh and D. B. Min, eds., Food Lipids;
Chemistry, Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002.
71. W. M. Willis and A. G. Marangoni, in C. C. Akoh and D. B. Min, eds., Food Lipids;
Chemistry, Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002.
72. B. Sreenivasan, J. Amer. Oil Chem. Soc., 55, 796805 (1978).
73. E. S. Lutton, M. F. Mallery, and J. Burgers, J. Amer. Oil Chem. Soc., 39, 233235 (1962).
74. D. Chobanov and R. Chobanova, J. Amer. Oil Chem. Soc., 54, 4750 (1977).
75. Y. C. Lo and A. P. Handel, J. Amer. Oil Chem. Soc., 60, 815818 (1983).
76. U. Bornscheuer, ed., Enzymes in Lipid Modification, Wiley-VCH, Weinheim, 2000.
77. P. Quinlan and S. Moore, INFORM, 4, 580585 (1993).
78. N. Matsuo, Lipid Technol., 13, 129133 (2001).
79. P. Forssell, R. Kervinen, M. Lappi, P. Linko, T. Suortti, and K. Poutanen, J. Amer. Oil
Chem. Soc., 69, 126129 (1992).
80. T. A. Foglia, K. Petruso, and S. H. Feairheller, J. Amer. Oil Chem. Soc., 70, 281285
(1993).
81. A. Rodriguez, E. Castro, M. C. Salinas, R. Lopez, and M. Miranda, J. Amer. Oil Chem.
Soc., 78, 431436 (2001).
82. S. Bhattacharyya, D. K. Bhattacharyya, and B. K. De, Eur. J. Lipid Sci. Technol., 323
328 (2000).
83. A. D. MacKenzie and D. E. Stevenson, Enz. Microb. Technol., 27, 302311 (2000).
84. T. Yang, X. Xu, C. He, and L. Li, Food Chem., 80, 473481 (2003).
85. F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker, New York, 2001.
86. K. T. Lee, T. A. Foglia, and M. J. Oh, J. Food Sci., 67, 10161020 (2002).
87. K.-T. Lee and T. A. Foglia, J. Food Sci., 65, 826831 (2000).
88. K.-T. Lee and T. A. Foglia, J. Amer. Oil Chem. Soc., 77, 10271034 (2000).
89. H. T. Osborn and C. C. Akoh, J. Food Sci., 67, 24802485 (2002).
REFERENCES 209
90. E. Deffense, in B. Rossell, ed., Animal Carcass Fats, vol. 2, Oils and Fats Series,
Leatherhead Publishing, Leatherhead, U.K., 2001.
91. R. E. Timms, in F. D. Gunstone, and F. B. Padley, Lipid Technologies and Applications,
92. W. Hamm, Trends Food Sci. Technol., 6, 121126 (1995).
93. M. A. Grompone, J. Amer. Oil Chem. Soc., 66, 253255 (1989).
94. M. A. Grompone, J. Amer. Oil Chem. Soc., 67, 980 (1990).
95. D. A. Glassner, E. A. Grulke, and J. I. Gray, J. Amer. Oil Chem. Soc., 61, 19191924
(1984).
96. J. P. Seuge, H. F. Vinconneau, Oleagineux, 30, 2530 (1975).
97. A. Tirtiaux, J. Amer. Oil Chem. Soc., 60, 473 (1983).
98. F. E. Luddy, J. W. Hampson, S. F. Herb, and H. L. Rothbart (to the U.S. Secretary of
Agriculture), U.S. Patent 3,944,585, 1976.
99. H. H. Taylor, F. E. Luddy, J. W. Hampson, and H. L. Rothbart, J. Amer. Oil Chem. Soc.,
53, 491495 (1976).
100. B. F. Haumann, J. Amer. Oil Chem. Soc., 64, 789795 (1987).
101. C. L. DeFouw, M. E. Zabik, and J. I. Gray, J. Food Sci., 46, 452456 (1981).
102. N. O. V. Sonntag, in R. W. Johnson and E. Fritz, eds., Fatty Acids in Industry; Processes,
Properties, Derivatives, Applications, Marcel Dekker, New York, 1989, pp. 2372.
103. D. B. Min and J. M. Boff, in C. C. Akoh and D. B. Min, eds., Food Lipids; Chemistry,
Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002, pp. 335363.
104. R. S. Parker, in J. E. Kinsella, ed., Advances in Food Research, vol. 33, Academic Press,
New York, 1989, pp. 157232.
105. R. J. Hamilton, in B. Rossell, ed., Oils and Fats Handbook, vol. 1: Vegetable Oils and
Fats, Leatherhead Food RA Publishing, Leatherhead, Surrey, U.K., 1999, pp. st.18.
106. H. Kanematsu et al. J. Jpn. Oil Chem. Soc. (Yukaguba), 32, 5153 (1983).
107. J. Griewahn and B. F. Daubert, J. Amer. Oil Chem. Soc., 25, 2627 (1948).
108. W. M. Cort, J. Amer. Oil Chem. Soc., 51, 321325 (1974).
109. L. R. Dugan, Jr., and H. R. Kraybill, J. Amer. Oil Chem. Soc., 33, 527528 (1956).
110. E. R. Sherwin and J. W. Thompson, Food Technol. 21, 912916 (1967).
111. C. M. Houlihan and C-T. Ho, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of
Fats and Oils, AOCS Press, Champaign, Illinois, 1985, pp. 117143.
112. C. Banias, V. Oreopoulou, and C. D. Thanopoulos, J. Amer. Oil Chem. Soc., 69, 520524
(1992).
113. L. deMan, J. M. deMan, and B. Blackman, J. Amer. Oil Chem. Soc., 68, 6369 (1991).
ing, Leatherhead, U.K., 2001, p. 185.
115. J. A. Merkle and D. K. Larick, J. Food Sci., 59, 478483 (1994).
116. M. E. Bailey, R. R. Chao, A. D. Clarke, K.-W. Um, and K. O. Gerhardt, in A. M. Spanier,
H. Okai, and M. Tamura, eds., Food Flavor and Safety: Molecular Analysis and Design,
American Chemical Society, Washington, D. C., 1993, pp. 117137.
117. S. W. Park and P. B. Addis, J. Agric. Food Chem., 34, 653659 (1986).
118. S. W. Park and P. B. Addis, J. Food Sci., 51, 13801381 (1986).
210 ANIMAL FATS
119. T. C. Ryan and J. I. Gray, J. Food Sci., 49, 13901391, 1393 (1984).
120. S. W. Park and P. B. Addis, J. Food Sci., 50, 14371441, 1444 (1985).
121. K. Lee, A. M. Herion, and N. A. Higley, J. Food Protect., 48, 158161 (1985).
122. J. A. Love, Animal Fats, in Y. H. Hui, ed., Baileys Industrial Oil & Fat Products, vol. 1,
Edible Oil & Fat Products: General Applications, Wiley, New York, 1996, pp. 118.
123. M. S. Brewer, J Muscle Foods, 10, 97117 (1999).
124. Recommended websites: www.aphis.usda.gov; www.bseinquiry.gov.uk; http://euro-
pa.eu.int/comm/food/fa/bse.
125. Anonymous, U. K. Department for Environment, Food and Rural Affairs. Available:
http://www.defra.gov.uk/animal/bse/indec.html.
126. S. B. Prusiner, Science, 216, 136144 (1982).
127. S. B. Prusiner, Proc. Natl. Acad. Sci. USA, 95, 1336313383 (1998).
128. A. Casolari, Food Microbiol., 15, 5963 (1998).
129. K. Danner, Pharm. Ind., 53, 614619 (1991).
130. R. G. Will, J. W. Ironside, M. Zeidler, S. N. Cousens, K. Estibeiro, A. Alperovitch,
S. Poser, M. Pocchiari, A. Hofman, and P. G. Smith, Lancet, 347, 921925 (1996).
131. P. Brown, Brit. Med. J., 322, 841844 (2001).
132. J. Collinge, Lancet, 354, 317323 (1999).
133. Anonymous, Render, the Bulletin of the National Renderers Association, 832, 2
(2002).
134. M. A. Friedman and D. E. Shalala, Federal Register 62(108), 3093530978 (1997).
135. Anonymous, Office International des Epizooties. Available: http://www.oie.int.
136. Anonymous, Regulation (EC) 999/2001 of the European Parliament and of the Council
of 22 May 2001, Laying down rules for the prevention, control, and eradication of certain
transmissible spongiform encephalopathies. Available: http://europa.eu.int/comm/food/
fs/afs/marktlab/marktlab14_en.pdf. See especially Annex V.
137. Anonymous, Regulation (EC) 1772/2002 of the European Parliament and of the Council
of 3 October 2002, Laying down health rules concerning animal by-products
not intended for human consumption. Available: http://www.eurits.org/pages/PDFs/
Oct02healthrules.pdf.
138. European Commission, 2000 Preliminary Report on Quantitative Risk Assessment on
the Use of the Vertebral Column for the Production of Gelatine and Tallow. Submitted to
the Scientific Steering Committee at its meeting of 1314 April, 2000, Brussels, Belgium.
139. T. R. Appel, D. Riesner, F. Von Rheinbaben, and M. Heinzel, Eur. J. Lipid Sci. Technol.,
103, 713721 (2001).
140. Anonymous, U.S. Food and Drug Administration, Title 21 Food and Drugs, Chapter 1
Food and Drug Administration, Department of Health and Human Services, Part 589
Substances Prohibited from Use in Animal Food or Feed Table of Contents,
Subpart B Listing of Specific Substances Prohibited from Use in Animal Food or
Feed, Sec. 589.2000 Animal Proteins Prohibited in Ruminant Feed. Available: http://
w3.aces.uiuc.edu/AnSci/BSE/BSE_FDA _Regul_Title_21_Part_589_2000.htm.
141. World Heath Organization, INFORM, 12, 588 (2001).
142. E. J. Cummins, S. F. Colgan, P. M. Grace, D. J. Fry, K. P. McDonnell, and S. M. Ward,
Human Ecologi. Risk Assess., 8, 11771192 (2002).
REFERENCES 211
143. M. S. Graboski and R. L. McCormick, Prog. Energy Combust. Sci., 24, 125164 (1998).
144. Anonymous, World Biodiesel Review, F. O. Lichts, World Ethanol and Biofuels Report,
Vol. 1, No. 10, (2003). Available: http://www.agra-net.com.
145. C. L. Peterson, D. L. Reece, B. L. Hammond, J. Thompson, and S. M. Beck, Appl. Eng.
Agric., 13, 7179 (1997).
146. D. Zheng and M. A. Hanna, Bioresource Technol., 57, 137142 (1996).
147. T. A. Foglia, L. A. Nelson, R. O. Dunn, and W. N. Marmer, J. Amer. Oil Chem Soc., 74,
951955 (1997).
148. K. T. Lee, T. A. Foglia, and K. S. Chang, J. Amer. Oil Chem. Soc., 79, 191195 (2002).
149. R. L. McCormick, M. Graboski, T. L. Alleman, and A. M. Herring, Environ. Sci.
Technol., 35, 17421747 (2001).
150. Y. Ali and M. A. Hanna, Trans. ASAE, 39, 793797 (1996).
151. C. L. Peterson, D. L. Reece, B. L. Hammond, J. Thompson, and S. M. Beck, Appl. Eng.
Agric., 13, 7179 (1997).
152. M. J. Nye, T. W. Williamson, S. Deshpande, J. H. Schrader, and W. H. Snively, J. Amer.
Oil Chem. Soc., 60, 15981601 (1983).
153. M. Mittelbach and P. Tritthart, J. Amer. Oil Chem. Soc., 65, 11851187 (1988).
154. W.-H. Wu, T. A. Foglia, W. N. Marmer, R. O. Dunn, C. E. Goering, and T. E. Briggs,
155. M. Mittelbach and H. Enzelsberger, J. Amer. Oil Chem. Soc., 76, 545550 (1999).
156. G. A. Lowe, C. L. Peterson, J. C. Thomopson, J. S. Taberski, P. T. Mann, and C. L. Chase,
ASAE Paper No. 986081, Am. Soc. Automotive Engineers, St. Joseph, Michigan, 1998.
157. C. L. Chase, C. L. Peterson, G. A. Lowe, P. Mann, J. A. Smith and N. Y. Kado, SAE
Technical Paper No. 200001-2647, Soc. Automotive Engineers, Warrendale,
Pennsylvania, 2000.
158. Y. Zhang, M. A. Dube, D. D. McLean, and M. Kates, Bioresource Technol., 89, 116 (2003).
159. M. J. Haas, G. J. Piazza, and T. A. Foglia, in T. M. Kuo and H. W. Gardner, eds., Lipid
Biotechnology, Marcel Dekker, New York, 2002, pp. 587598.
160. Y. Watanabe, Y. Shimada, A. Sugihara, and Y. Tominaga, J. Amer. Oil Chem. Soc., 78,
703707 (2001).
161. A. Hsu, K. Jones, T. A. Foglia, and W. N. Marmer, Biotechnol. Appl. Biochem., 36,
181186 (2002).
162. T. T. Adams, J. Walsh, M. Brown, J. Goodrum, J. Sellers, and K. Das, A Demonstration
of Fat and Grease as an Industrial Boiler Fuel, Engineering Outreach Service, University
of Georgia, Athens, GA. (2002). Available: http://www.engr.uga.edu/service/outreach/
biofuel.htm.
163. M. B. Katan, P. L. Zock, and R. P. Mensink, in D. B. McCormick, D. M. Bier, and
A. G. Goodridge, eds, Annu. Rev. Nutr., 15, 473 493.
164. A. Ascherio and W. C. Willett, Am. J. Clin. Nutr., 66, 1006S-1010S (1997).
165. A. H. Lichtenstein, Womens Health (Larchmt), 12, 109114 (2003).
166. J. M. Hasman, INFORM, 6, 12061213 (1995).
167. M. P. Yurawecz, M. M. Mossoba, J. K. G. Kramer, M. W. Pariza and G. J. Nelson, eds.,
Advances in Conjugated Linoleic Acid Research, vol. 1, AOCS Press, Champaign,
Illinois, 1999.
212 ANIMAL FATS
168. J.-L. Sebedio, W. W. Christie, and R. Adlof, eds., Advances in Conjugated Linoleic Acid
Research, vol. 2, AOCS Press, Champaign, Illinois, 2003.
169. J. Fritsche, R. Rickert, and H. Steinhart, in M. P. Yurawecz, M. M. Mossoba,
J. K. G. Kramer, M. W. Pariza and G. J. Nelson, eds., Advances in Conjugated Linoleic
Acid Research, vol. 1, AOCS Press, Champaign, Illinois, 1999.
170. J. Fritsche, R. Rickert, and H. Steinhart, in M. P. Yurawecz, M. M. Mossoba,
J. K. G. Kramer, M. W. Pariza and G. J. Nelson, eds., Advances in Conjugated Linoleic
Acid Research, vol. 1, AOCS Press, Champaign, Illinois, 1999.
171. J. M. Griinari and D. E. Bauman, in M. P. Yurawecz, M. M. Mossoba, J. K. G. Kramer,
M. W. Pariza and G. J. Nelson, eds., Advances in Conjugated Linoleic Acid Research,
vol. 1, AOCS Press, Champaign, Illinois, 1999.
172. M. W. Pariza, S. H. Ashoor, F. S. Chu, and D. B. Lund, Cancer Lett., 7, 6369 (1979).
173. Y. L. Ha, N. K. Grimm, and M. W. Pariza, Carcinogenesis, 8, 18811887 (1987).
174. M. W. Pariza, in M. P. Yurawecz, M. M. Mossoba, J. K. G. Kramer, M. W. Pariza and
G. J. Nelson, eds., Advances in Conjugated Linoleic Acid Research, vol. 1, AOCS Press,
Champaign, Illinois, 1999.
175. U. Riserus, A. Smedman, S. Basu, and B. Vessby, Lipids, 38, 133137 (2003).
176. S. F. Chin, W. Liu, J. M. Storkson, YU. L. Ha, and M. W. Pariza, J. Food Composit.
Analy., 5, 185197 (1992).
6
Vegetable Oils
Frank D. Gunstone
1. INTRODUCTION
This chapter is concerned with the major and minor vegetable oils. It includes a
brief account of the biosynthetic pathways for plant lipids and a description of min-
or, but important, components present in commercial vegetable oils. This is
followed by a description of the major and minor vegetable oils. The major oils
are discussed in more detail elsewhere in this work and in another recent book
(1). The natural oils do not always meet human dietary requirements and may
have to be modified. There is a discussion on what drives modification and of
the various ways in which this can be achieved. Finally, some production and trade
statistics are provided and discussed.
2. BIOSYNTHESIS
2.1. Introduction
This section provides a brief account of the biosynthetic pathways to triacylglycer-
ols in plants, but it requires a preliminary discussion of fatty acid biosynthesis.
The so-called acetate-malonate pathway leads to three different kinds of natural
products depending on the detailed pathway followed. Fatty acids result from a
reductive pathway to be described here, but acetate and malonate are also precur-
sors for the isoprenoids (terpenes and sterols) produced via mevalonic acid (C6) and
213
214 VEGETABLE OILS
for a wide range of phenolic compounds resulting from cyclization of polyacetate.

This illustrates the observation that nature is economical in the range of both
substrates and reactions employed in biosynthesis.
The major biosynthetic pathways to fatty in plants involve three stages (24):
de novo synthesis of palmitic (or other alkanoic) acid from acetate (C2, a
product of carbohydrate metabolism) by reaction with malonate (C3).
Further chain-elongation of saturated or unsaturated acids by two-carbon units.
Desaturation. Particularly of stearic acid, first to oleic acid and then to linoleic
and linolenic acids.
These changes take place in different parts of the cell, under the influence of spe-
cific enzymes or enzyme complexes, and they require the acids to be in appropriate
substrate form.
2.2. de novo Synthesis of Saturated Acids

In plant systems, de novo synthesis occurs in the plastid and results mainly in the
conversion of acetate to palmitate. All 16 carbon atoms in palmitic acid are derived
from acetatehalf from the methyl carbon and half from the acyl carbon. Two of
the carbon atoms (C-15 and C-16) come directly from acetate, and the other 14
come from acetate via the more reactive malonate. Production of malonate requires
the incorporation of an additional carbon atom into the acetyl group. This is sup-
plied as bicarbonate, and this same carbon atom is subsequently lost as carbon
dioxide. The acyl groups are attached to co-enzyme A (CoASH) during part of
the cycle and to acyl carrier protein (ACPSH) during another part. The abbreviated
symbols used for these co-enzymes emphasize the thiol groups (SH) to which the
acyl chains are attached.
In the de novo pathway, acetate and malonate react through a series of steps con-
verting acetate first to butanoate (C4), then to hexanoate (C6), and then sequentially,
two carbon atoms at a time, to palmitate (C16). At this stage, a thioesterase liberates
the acyl chain from ACP. The thioesterase is not completely specific, and acids of
other chain lengths may be produced. This is obviously true in the lauric oils where
the specificity of their thioesterases causes lauric acid (12:0) to be the major satu-
rated acid, accompanied by lower levels of caprylic (8:0), capric (10:0), myristic
(14:0), and palmitic acid (16:0).
Conversion of acetyl-CoA to malonyl-CoA with a biotin enzyme (acetyl-CoA

carboxylase)
CH3 COSCoA HCO
3 ATP ! HO2 CCH2 COSCoA PI ADP
Conversion of acetyl-ACP to butanoyl-ACP (four-step cycle)
CH3 COSACP HO2 CCH2 COSACP ! CH3 COCH2 COSACP CO2

ACPSH condensation
BIOSYNTHESIS 215
CH3 COCH2 COSACP NADPH H ! CH3 CHOHCH2 COSACP

NADP reduction
CH3 CHOHCH2 COSACP ! CH3 CH

CHCOSACP

H2 O dehydration

CH3 CH

CHCOSACP NADPH H ! CH3 CH2 CH2 COSACP

NADP reduction
This four-step cycle includes condensation of acetate and malonate to give ketobu-
tanoate with subsequent reduction to butanoate in three further steps. These are
reduction to the 3R hydroxy acid, dehydration to the 2t acid, and reduction again.
Reduction is affected by NADPH and a proton. The process is then repeated to add
further two-carbon units until a thioesterase liberates the free acid. This sequence
requires a fatty acid synthase, which contains the enzymes needed for each of the
four steps viz. b-ketoacyl-ACP synthase, b-ketoacyl-ACP reductase, b-ketoacyl-
ACP dehydrase, and enoyl-ACP reductase, respectively.
There are several minor modifications of the de novo process, but these are not
important for the major fatty acids occurring in vegetable oils. They are detailed in
more extensive accounts of this topic (24).
2.3. Desaturation to Monoene and Polyene Acids in Plant Systems

The first desaturation of a saturated acyl chain occurs in the plastid. The most com-
mon is the conversion of stearate to oleate and involves the removal of pro-(R)
hydrogen atoms from C-9 and C-10 to give a cis-olefinic bond under the influence
of a 9 desaturase. The system is oxygen-dependent and involves the reduced form
of ferredoxin.
Further desaturation, occurring in the cytoplasm, converts oleate, in the form of
its phosphatidylcholine, to linoleate (12 desaturase) and linoleate, as the mono-
galactosyldiacylglycerol derivative, to linolenate (15 desaturase). The additional
double bonds have cis-configuration and are in a methylene-interrupted relation to
each other. This 1,4-diene unit is characteristic of polyunsaturated fatty acids and is
to be distinguished from the 1,3 (conjugated) systems in carotenoids and the 1,5
system in terpenes. It is of interest that the enzymes converting oleate to the acet-
ylenic acid crepenynic (9c12a18:2) and to the epoxy acid vernolic (12,13-epoxyo-
leic) are very similar to 12 deaturase, which converts oleate to linoleate. A small
change in DNA sequence is sufficient to lead to these different fatty acids from the
same substrate. The same holds for the conversion of oleate to ricinoleate where a
small change alters a desaturase to a hydroxylase (5).
Linoleic acid and linolenic acid are essential fatty acids that cannot be made by
animals and must be obtained by dietary intake from plant sources. When metabo-
lized in animals, these two acids each give rise to a family of C18, C20, and C22 n-6
and n-3 polyunsaturated fatty acids; thus:
216 VEGETABLE OILS
n-6 acids based on linoleic acid 18 : 2 ! 18 : 3 ! 20 : 3 ! 20 : 4:

n-3 acids based on a-linolenic acid 18 : 3 ! 18 : 4 ! 20 : 4 ! 20 : 5 ! 22 : 5
! 24 : 5 ! 24 : 6 ! 22 : 6
Although common in animal systems, the 6 desaturase is less apparent in the

plant world, although it must operate in the production of g-linolenic acid (6, 9,
1218:3) from linoleate and of stearidonic acid (6, 9, 12, 1518:4) from a-linole-
nate. The C20 and C22 polyenes that characterize animal systems and particularly
fish lipids either do not exist in plant systems or are exceedingly rare. The produc-
tion of important acids such as arachidonic (5, 8, 11, 1420:4), eicosapentaenoic
(5, 8, 11, 14, 1720:5), and docosahexaenoic (4, 7, 10, 13, 16, 1922:6) in plant
systems is a challenge for plant geneticists (6).
2.4. Elongation
Elongation by two carbon atoms occurs commonly in fatty acid biosynthesis. It is a
variant of de novo chain-lengthening and occurs with acetyl or malonyl CoA or ACP
derivatives. The substrate is any preformed saturated or unsaturated acid. For exam-
ple, erucic (22:1) in high-erucic acid rapeseed oil and nervonic acid (24:1) in seed
oil are formed from oleic acid by two and three elongations, respectively:
18 : 19 ! 20 : 111 ! 22 : 113 ! 24 : 115

oleic cetoleic erucic nervonic
2.5. Formation of Triacylglycerols

The glycerol present in glycerolipids is derived from glycerol (1,2,3-trihydroxypro-
pane), glyceraldehyde (2,3-dihydroxypropanal), or dihydroxyacetone (1,3-dihy-
droxypropan-2-one), all of which are products of carbohydrate metabolism. The
triacylglycerols are made from phosphatidic acids or 1,2-diacylglycerols, which
are themselves interconvertible, and acylation is effected by fatty acids as their
CoA thiol esters. The esterification reactions are enzyme-controlled with marked
selectivity in the acylation of each hydroxyl group leading to a nonrandom distri-
bution of fatty acids in plant lipids. Saturated acids are selectively bound at the sn-1
position and unsaturated acids at the sn-2 position. The acyl transferase for the sn-3
position is less selective, and fatty acids in this position are largely controlled by the
composition of the acyl-CoA pool.
Although not shown in the following simplified scheme, the phospatidic acid/
1,2-diacylglycerol duo is also an intermediate in the pathway to the phospholipids
(phosphatidylcholines, phosphatidylethanolamines, phosphatidylinositols, phospha-
tidylinositols, and phosphatidylserines) and the mono- and digalactosyldiacylgly-
cerols.
MINOR COMPONENTS 217
sn-glycerol-3-phospate HOCH2 CHOHCH2 OPO3 H2

acylation at sn-1 #
acylation at sn-2 #
phosphatidic acid R1 COOCH2 CHOCOR2 CH2 OPO3 H2
#"
sn-1;2-diacylglycerol R1 COOCH2 CHOCOR2 CH2 OH
acylation at sn-3 #
triacylglycerol R1 COOCH2 CHOCOR2 CH2 OCOR3
3. MINOR COMPONENTS
Crude vegetable oils are mainly triacylglycerols (around 95%) along with some free
acids, monoacylglycerols, and diacylglycerols. They also contain variable amounts
of other components such as phospholipids, free and esterified sterols, triterpene
alcohols, tocopherols and tocotrienols, carotenes, chlorophylls and other coloring
matters, and hydrocarbons as well as traces of metals, oxidation products, undesir-
able flavors, and so on. These are discussed in detail elsewhere. Refining procedures
have been developed to convert the crude oil into a bland product that meets a
defined specification. Some of the minor components are valuable in their own right
and should be retained in the refined oil and/or trapped in a side stream for recovery
and further utilization.
Crude oils generally contain phospholipids that are removed during the degum-
ming stage of refining as a crude mixture (lecithin). This valuable product is the
basis of the phospholipid industry, and phospholipids are used extensively in
food products, in animal feeds, and in industrial processes. The major members
are phosphatidylcholines, phosphatidylethanolamines, and phosphatidylinositols
and are accompanied by smaller proportions of other phospholipids. Soybean oil
(3.2%), rapeseed oil (2.5%), and sunflower seed oil (1.5%) contain the proportions
of total phospholipids indicated in parentheses and are the main sources of commer-
cial lecithins, especially soya lecithin. Palm oil contains little or no phospholipids
(79).
Most vegetable oils contain 10005000 ppm (15 g/kg) of sterols, partly as free
sterols and partly as esterified sterols. Higher levels are present in rapeseed oil
(511 g/kg, mean 7.5) and in corn oil (822 g/kg, mean 14). Sitosterol is
generally the major phytosterol (5080% of total sterol) with campesterol, stigmas-
terol, and 5-avenasterol also frequently attaining significant levels. Brassicasterol
is virtually absent from the major seed oils except for rapeseed oil where it com-
prises 10% of the total sterol. Cholesterol is generally considered to be a zoosterol
and is not present in plant systems at any significant level. The normal value of
2050 ppm in vegetable oils compares with the much higher levels reported for ani-
mal fats (up to 1000 ppm), fish oils (up to 7000 ppm), dairy fats (20003000 ppm),
and egg yolks (12,500 ppm). Phytosterols (and other compounds) can be recovered
from deodorizer distillate and are used to produce pharmaceutical steroids (10).
218 VEGETABLE OILS
Tocol extracts are mixtures of up to eight compounds. There are four tocopherols
with a saturated, branched C16 side chain and four analogous tocotrienols with three
double bonds in the side chain. The tocotrienols, although significant in palm oil,
are generally less common than the tocopherols and much less is known about their
biological properties. The four tocopherols differ in the number of methyl groups
attached to the heterocyclic moiety. They are designated a (5,7,8-trimethyl), b (5,7-
dimethyl), g (7,8-dimethyl), and d (8-methyl). The tocols have two valuable proper-
ties: They show vitamin E activity, and they are powerful antioxidants. These two
properties are not identical. For vitamin E activity, the order is a (1.0) >b (0.5) > g
(0.1) > d (0.03) with total activity usually expressed in a-tocopherol equivalents.
For antioxidants, this order is reversed.
Natural tocopherol mixtures are used as antioxidants, usually at levels up to
500 ppm, along with ascorbyl palmitate to extend the antioxidant activity. At higher
levels (>1000 ppm), a-tocopherol is considered to act as a pro-oxidant. As vege-
table oils contain tocols at 200800 ppm, further additions show only a limited
effect. The tocols are very sensitive to oxidation and are more stable in esterified
form where the all-important hydroxyl group is not free. However such compounds
do not show antioxidant activity until they have been hydrolyzed in vivo to the free
phenolic form (11).
Hydrocarbons are very minor components of oils and fats but are of dietary and
legislative interest. They include alkanes, alkenes such as squalene and carotenes,
and polycyclicaromatic hydrocarbons. Squalene (C30H50) is a highly unsaturated
open-chain triterpene. It is used in the cosmetic industry after hydrogenation to
squalane (C30H62). The most abundant source of squalene is the liver oil of the
deep-sea dogfish (Squalus acanthushence the name squalene) and some other
marine species. Vegetable sources of potential interest include olive oil and amar-
anthus (Section 6).
Carotenes are minor components in many vegetable oils and particularly in palm
oil. They contain a long chain of conjugated unsaturation and are yellow/orange/red
in color. Crude palm oil normally contains 500700 ppm of carotenes. These are
mainly a-carotene (24 42% of total carotene) and b-carotene (5060%) along
with low levels of several other carotenes. Carotenes are also present in palm leaves
and in pressed fiber remaining when oil has been expressed from palm fruits.
Attempts have been made to retain these valuable materials in refined palm oil
(red palm oil) or to recover them in concentrated form (12). Carotenes can be
recovered from palm methyl esters. The esters are prepared by methanolysis of
palm oil and are produced in large quantities for use as biodiesel, as a solvent,
for conversion to alcohols, and so on. The carotenes can be recovered from the
esters by chromatography in an open column or by molecular distillation. The latter
gives a carotene concentrate (8%) which can be purified (>90%) by chromatogra-
phy (1315). The concentrates are used as food-dyes, as a vitamin additive, and by
the pharmaceutical and cosmetic industries.
Polycyclic aromatic hydrocarbons are present at levels up to about 150 mg/kg
(ppb) in most crude vegetable oils, although slightly less after refining (<80 ppb).
They are removed only to a small extent during bleaching and somewhat more
CLASSIFICATION OF VEGETABLE OILS 219
during deodorization. This holds more particularly for the more volatile tri- and tet-
racyclic compounds. The pentacyclic and other less volatile compounds are best
removed with activated charcoal added to the earth during bleaching. These low
values do not hold for crude coconut oil dried with combustion gases where values
around 3000 ppb are routinely recorded. Normal values are obtained after charcoal
treatment (16). Extracted oils may contain pesticides resulting from agricultural
practices, but these are usually removed during deodorization.
4. CLASSIFICATION OF VEGETABLE OILS
4.1. Classification by Source Type

One market analyst provides regular information on the production and trade of
17 commodity oils and fats. In Oil World Publications (ISTA Mielke GmbH, Ham-
burg, Germany), these are listed in the order: soybean, cotton, groundnut, sun-
flower, rapeseed, sesame, corn, olive, palm, palm kernel, coconut, butter, lard,
fish, linseed, castor, and tallow. The list includes 4 materials of animal origin and
13 of vegetable origin. This chapter will report on the 13 vegetable oils and some
others of plant origin.
The above list does not include cocoa butter nor minor oils such as rice bran oil
or safflower oil. Nor does it distinguish between oils from a common botanical
source with a modified fatty acid composition, such as canola oil and high-erucic
rape seed oil, linseed oil and linola, or the various types of sunflower oil.
The commodity vegetable oils can be classified in various ways. Some are
byproducts so that decisions regarding their production are largely controlled by
the nonoil component. Examples are corn oil and cottonseed oil. These are bypro-
ducts of cereal and fiber production, respectively. Also, rice bran oil is a byproduct
of rice production. As a consequence, oil production is not the main economic fac-
tor that influences the areas cultivated by these crops. It is sometimes argued that
soybean falls into this byproduct category and that the bean is grown as a source of
protein with the oil as byproduct. It is true that the bean produces only 18% of oil
against 79% of residual meal rich in protein, but the value of these two commodities
is evenly balanced and there are times when the demand is oil-led and other times
when it is meal-led.
Some commodity oils and fats such as palm, palm kernel, coconut, and olive are
tree crops. Once the trees mature, they continue to produce fruit for many years and
production levels cannot be greatly changed from season to season.
This leaves the annual crops of rape, sunflower, and linseed (often called soft
oils) where planting decisions are made on a year-to-year basis depending on per-
ceived benefit. Planting areas are decided on agricultural grounds (such as crop
rotation) and on economic grounds. Low prices and/or high stocks in one season
tend to lead to reduced plantings in the following season.
It is also possible to distinguish between oils from seeds, such as soybean and
rapeseed, and those coming from the fleshy part of a fruit such as palm and olive.
An important point here is that for oilseeds, exports and imports are as seeds as well
220 VEGETABLE OILS
as extracted oil, whereas for the other group, trade is confined to extracted oil.
When making comparisons, it is not sufficient to consider figures for traded oil
without also including the oil-equivalent of traded seeds.
The annual production of soybean oil exceeds that of palm oil, but it is claimed
that trade (imports/exports) in palm oil is larger. This view, based on the fact that
trade in palm oil far exceeds that in soybean oil, does not take into account the very
large exports of soybeans themselves. What are the numbers if exported beans are
also considered in terms of their oil-equivalent? The following figures for year
2000/01 are taken from Oil World Annual 2001. For soybean, 7.4 million tons1
were exported as oil and a further oil-equivalent of 7.8 million tons was exported
in the form of beans. This latter figure is based on exports of 50.0 million tons of
beans, of which 85.2% was crushed (world average) with an oil content of 18.3%.
On this basis, the full level of exported soybean oil is about 15.2 million tons. This
is still lower than the figure for palm oil (16.8 million tons), but the disparity is not
so large (17).
The term tropical oils is correctly applied to oils and fats produced in the tropics
and refers particularly to the (highly saturated) lauric oils (Sections 5.3 and 5.10)
and to palm oil (Section 5.9). This term is frequently and unfairly used in a dero-
gatory sense, partly through ignorance about the difference in fatty acid composi-
tion and use between lauric oils and palm oil and of the considerable nutritional
value of the latter.
Castor oil is classed as an industrial oil because it is used only for nonfood pur-
poses (Section 5.1). Linseed oil also is used almost entirely for industrial purposes.
In its limited use as an edible oil, it is generally known by its alternative name of
flaxseed oil (Section 5.7).
4.2. Classification by Fatty Acid Composition

The list of natural fatty acids exceeds 1000, but commercial interest is limited to a
smaller numberperhaps around 20. Ignoring the lipid membrane, rich in a-lino-
lenic acid and present in all green tissue, the three dominant acids in the plant king-
dom are palmitic, oleic, and linoleic, sometimes accompanied by stearic acid and
by linolenic acid. Others, occuring in specialty oils, include myristic, lauric, erucic,
hexadecenoic, petroselinic, g-linolenic acid, eleostearic and isomers, ricinoleic, and
vernolic (Table 1).
Although it is convenient to categorize oils by their fatty acid composition, it
must be remembered that this is not the only index of their nutritional value or
of their oxidative stability. Attention must also be given to the minor components
in the crude oil and to those remaining after refining (see Section 3).
1
The figures cited in tons in this chapter have been taken from publications using tonnes, and at the
editors request, the numbers have not been adjusted: 1 ton is equivalent to 0.984 tons, and the numbers
are not significantly different in the present context.
TABLE 1. Structures of the More Common Acids in Vegetable Oils.
Trivial Name Symbol Unsaturation (if any)

Saturated
Lauric 12:0
Myristic 14:0
Palmitic 16:0
Stearic 18:0
Monounsaturated
Oleic 18:1 9c
Petroselinic 18:1 6c
Erucic 22:1 13c
Polyunsaturated (non-conjugated)
Linoleic 18:2 9c12c
Linolenic (a) 18:3 9c12c15c
Linolenic (g) 18:3 6c9c12c
Polyunsaturated (conjugated)
Eleostearic 18:3 9c11t13t
Calendic 18:3 8t10t12c
Oxygenated
Ricinoleic 18:1 12-OH 9c
Vernolic 18:1 12,13-epoxy 9c
TABLE 2. (a) Typical Fatty Acid Composition (%wt).

Oil source 16:0 18:0 18:1 18:2 18:3
Cocoa butter 26 34 35
Corn 13 3 31 52 1
Cottonseed 27 2 18 51 Tr
Groundnut 13 3 38 41 Tr
Linseed 6 3 17 14 60
Olive 10 2 78 7 1
Palm 44 4 39 11 Tr
Palm olein 41 4 31 12 Tr
Palm stearin 4774 46 1637 310
Rape (high erucic)
3 1 16 14 10
Rape (low erucic) 4 2 56 26 10
Rice bran oil 20 2 42 32
Safflower 7 3 14 75
Safflower (high oleic) 6 2 74 16
Sesame 9 6 41 43
Soybean 11 4 22 53 8
Sunflower 6 5 20 60 Tr
Sunflower (Sunola) 4 5 81 8 Tr
Sunflower (NuSun) 4 5 65 26
Adapted from Gunstone (11).

tr trace (<1%).
20:1 6% and 22:1 5%.
TABLE 2. (b) Typical Fatty Acid Composition (%wt) of Lauric Oils.

Oil source 8:0 10:0 12:0 14:0 16:0 18:0 18:1 18:2
Coconut 8 7 48 16 9 2 7 2
Palm kernel 3 4 45 18 9 3 15 2
Adapted from Gunstone (11).

222 VEGETABLE OILS
TABLE 3. Vegetable Oils by Fatty Acid Type.
Acid(s) Vegetable Oil
Lauric coconut, palm kernel

Palmitic palm, cottonseed
Oleic/linoleic groundnut, safflower, sesame, sunflower, cottonseed, canola, soybean
High oleic olive, safflower, sunflower, canola, groundnut, soybean
Linolenic linseed, canola, soybean
Vegetable butters cocoa butter
Erucic acid HEAR
, crambe
Conjugated acid tung, calendula
Oxygenated acids castor, vernolic
HEAR high erucic acid rapeseed oil.
Most of these oils are described in more detail elsewhere in this work. Production
and trading figures are discussed in Section 10 (see Tables 2 and 3).
4.2.1. Lauric Oils There are two major lauric oilscoconut oil and palm-kernel
oil. Both are tropical oils, and both are tree crops. They differ from all other com-
modity oils in their higher level of medium chain acids, especially lauric, and
slightly from one another as shown in Table 2(b). They find limited use in food
products and are used extensively in the production of surface-active compounds.
For more information, see Sections 5.3 and 5.10.
4.2.2. Palmitic Acid Oils The commodity oil richest in palmitic acid is palm oil
(44%). This oil is also rich in oleic acid (37%), contains lower levels of linoleic acid
(10%), and is a valuable source of minor components, especially carotenes, toco-
pherols, and tocotrienols (Section 3). Palm oil is an important world commodity in
feeding the developing world. It is fractionated extensively to give a wider range of
uses as palm olein and palm stearin. The only other commodity oil with a signifi-
cant level of palmitic acid is cottonseed oil (27%).
To produce high-quality spreads, the solid portion should be in the b0 form and
this is most likely when the solid triacylglycerols contain acids of varying chain
lengthgenerally C16 and C18. For this reason, palm oil or cottonseed oil are
frequent components of the blend used to produce spreads.
4.2.3. The Oleic/Linoleic Acid Group This is the most common type of vege-
table oil and includes peanut or groundnut (38% oleic and 41% linoleic acid), saf-
flower (14% and 75%), sesame (38% and 45%), and sunflower (20% and 69%). The
sum of these two acids is generally 8090% so there can only be low levels of satu-
rated or other acids. At the present time, there is a demand for high oleic oils, so
variants of these oils enriched in oleic acid have been developed (Section 4.2.4).
Cottonseed (18% and 51%) differs from the others cited here in its higher level
of palmitic acid. Low-erucic rape/canola (56% and 26%) and soybean oil (22%
and 53%), while belonging to this oleic/linoleic group, also contain linolenic acid at
levels of around 10% and 8%, respectively.
4.2.4. High Oleic Oils Olive oil is an important high-oleic oil (78%). It is gen-
erally consumed in an unrefined state and therefore retains all of the natural unsa-
ponifiable material, including valuable antioxidants. Other high oleic oils have been
developed by traditional breeding methods or by genetic engineering. These include
variants of regular safflower (77% oleic acid), sunflower (8090%), canola (78%),
peanut (76%), and soybean (7986%) (See appropriate entries in Section 5).
4.2.5. Linolenic Acid Oils The most familiar high-linolenic acid oil is linseed
(5060%), but rape/canola and soybean are important commodity oils containing
linolenic acid at 10% and 8%, respectively. This triene acid has both positive
and negative connotations. It is very easily oxidized, and its oxidation products
have strong undesirable flavors. It therefore contributes to a shortening of shelf
life and is not favored by food processors for this reason. Hence, these last two
oils are frequently subjected to brush hydrogenation (at least) to reduce the level
of linolenic acid. Also, seed breeders are striving to produce low-linolenic forms
of rape/canola and soybean. However, nutritional scientists tell a different tale.
There is a growing awareness that the present ratio of dietary n-6 to n-3 acids at
between 5 and 10 to 1 is too high and should be changed to the lower end of
this range or below. This is particularly a problem for those countries whose diet
contains high levels of n-6 PUFA oils as in the United States (18, 19).
4.2.6. Vegetable Butters Most fats/oils derived from vegetable sources are
liquid, reflecting the unsaturated nature of most of their component acids. The
few that are solid (i.e., have melting points above ambient temperature) are known
as butters. The best-known and most important member of this class is cocoa butter
(Section 5.2), which is the major or only, fat component in chocolate. Others dis-
cussed in Section 6 include illipe butter (Borneo tallow), kokum butter, mango ker-
nel fat, sal fat, and shea butter, These along with palm oil are, in some countries,
permitted replacements, in part, for cocoa butter in chocolate (20, 21).
4.2.7. Erucic Acid Oils Traditional rapeseed oil (colza), which was used as an
illuminant as well as for food, was rich in erucic acid (22:1). Certain findings with
animals suggested that this was not a healthy oil, and although the matter was never
proved for humans, new varieties of rape (canola) were developed, principally in
Canada, with virtually no erucic acid. However, erucic acid and its oils have a num-
ber of oleochemical uses and there is a continuing if limited demand for erucic oils.
Crambe oil (Crambe abyssinica and C. hispanica) is being developed as an alter-
native erucic-rich oil. A major use of erucic acid involves the formation of eruca-
mide (RCONH2), which acts as a nonslip agent in polythene and polypropylene.
Some 30,000 tons of this is used each year in clingfilm and related materials.
Oleamide can be used as an inferior alternative (2225).
224 VEGETABLE OILS
4.2.8. Conjugated Acid Group Most of the important polyunsaturated fatty

acids such as linoleic and linolenic are 1,4-dienes with methylene-interrupted unsa-
turation, and this structural unit relates to many of the characteristic properties of
these acids. However, a smaller number of vegetable oils have acids with conju-
gated unsaturation. Most of these are C18 acids with three double bonds, although
some have four. The best known acid in this group is a-eleostearic acid (9c11t13t-
18:3). This is a major component of tung oil (china wood oil, Aleurites fordii and
A. montana), which contains 70% of eleostearic acid. The oil polymerizes readily.
Calendula oil (Calendula officinalis) is being developed as a source of calendic acid
(8t10t12c-18:3) (Section 6).
4.2.9. Hydroxy and Epoxy Acid Group Although a number of oils contain
acids with hydroxy, epoxy, or oxo (keto) functions, only one is readily available.
Castor oil contains over 90% of ricinoleic acid (12-hydroxyoleic acid) and about
1% of 9,10-dihydroxystearic acid (section 5.1).
Some contain vernolic acid (12,13-epoxyoleic acid). This has several potentially
useful properties, and attempts are being made to produce an economically viable
crop.
5. THE MAJOR VEGETABLE OILS AND FATS
There follows a brief account of the commodity oils. This includes the nature of the
oil and any special features, its production levels, and major areas of production.
Most of these oils are discussed in greater detail elsewhere in this series and in
Gunstone (1). Triacylglycerol composition is indicated by three-letter symbols
that include all of the isomeric triacylglycerols containing the acids designated
where P palmitic, St stearic, S saturated, O oleic, L linoleic, and Ln
linolenic.
5.1. Castor Oil

Castor oil is derived from the plant Ricinus communis grown mainly in India,
Brazil, and China at a world production level of about 0.5 million tons of oil.
This oil differs from all other commercial oils in being rich in ricinoleic acid
( 90%, 12-hydroxyoleic). Compared with the common vegetable oils, castor oil
is more viscous, less soluble in hexane, and more soluble in ethanol, all as a con-
sequence of the presence of the hydroxy acid. This hydroxy acid has several inter-
esting properties by which it can be converted to useful products.

CHCH2 7 COOH
CH3 CH2 5 CHOHCH2 CH

ricinoleic acid 12-hydroxyoleic acid
THE MAJOR VEGETABLE OILS AND FATS 225
Sulfation converts the hydroxyl group to a sulfate (OSO2OH) with improved

surfactant properties. Apart from soap, it is the earliest anionic surfactant
(dating back to 1874) and is still used in textile processing, leather treatment,
and as an additive for cutting oils and hydraulic fluids. The sulfated hydro-
genated oil has the consistency of an ointment and gives adjustable viscosity
to water-based formulations with excellent skin compatibility.
Dehydration of castor oil and of castor acids gives products enriched in diene
acids, some with conjugated unsaturation. These products are valuable
alternatives to drying oils such as tung oil.
Hydrogenated castor oil and hydrogenated castor acids, with higher melting
points than the nonhydrogenated material, are used in cosmetics, coatings, and
greases. Greases prepared from tallow are much improved when salts of
12-hydroxystearic acid are added.
Castor oil reacts with isocyanates to give polyurethanes that are much used for
wood preservation and have been developed as encapsulating materials.
Splitting ricinoleic acid with caustic soda gives C8 and C10 products. At 180
200 C with a 1:1 caustic/castor ratio, the major products are 2-octanone and
10-hydroxydecanoic acid. At 250275 C and a 2:1 ratio, the products are
2-octanol and sebacic (decanedioic) acid. The dibasic acid, when reacted
appropriately, gives a nylon (polyamide) and efficient lubricants (esters).
Splitting ricinoleic acid with steam gives C7 and C11 products. This splitting
process has been much improved by the development of a continuous steam
cracking process. Heptanal is used in perfumes, and 10-undecenoic acid can
be converted to a polyamide (Rilsan) while its salts show antifungal proper-
ties. Several new uses developed for these C7 and C11 compounds have been
described by Caupin (26).
5.2. Cocoa Butter

The cocoa bean (Theobroma cacao) is the source of two important ingredients of
chocolate: cocoa powder and a solid fat called cocoa butter. The usefulness of
cocoa butter for this purpose is related to its fatty acid and triacylglycerol composi-
tion. The major triacylglycerols are symmetrical disaturated oleic glycerol esters of
the type SOS and include POP (1823%), POSt (36 41%), and StOSt (2331%).
Cocoa butter generally commands a premium price, and cheaper alternatives have
been developed. These are known as cocoa butter alternatives, cocoa butter equiva-
lents, cocoa butter improvers, cocoa butter replacers, and cocoa butter substitutes.
They may or may not have similar chemical composition to cocoa butter, but they
must display similar melting behavior. The annual production of cocoa beans is
about 2.7 million tons containing 45 48% cocoa butter, i.e., 1.21.3 million tons
of fat. Most fats/oils derived from vegetable sources are liquid, reflecting the unsa-
turated nature of most of their component acids. Only a few are solid (i.e., have
melting points above ambient temperature). The best-known and most important
member of this class is cocoa butter. This is the major fat component in chocolate.
226 VEGETABLE OILS
To be a satisfactory ingredient in chocolate, the fat must have certain defined melt-
ing behavior. It must be hard and brittle at ambient temperature, have a steep melt-
ing curve, and be completely melted at mouth temperature. These properties give a
cooling sensation on the tongue. In addition, the chocolate should break with a
snap. This behavior is associated with triacylglycerols with the structure SOS. A
few less well-known fats have similar composition and properties. Other fats
with similar properties can be produced more economically. These can be used
in confectionery fats, but the composition of chocolate is now defined by law,
although not identically in all countries. The degree of substitution of cocoa butter
by other hard fats is limited both in the fats that can be used and in the level at
which they may be added. This varies between zero and a maximum of 5% of
the product. The permitted list is palm oil, illipe butter (Borneo tallow), kokum
butter, mango kernel fat, sal fat, and shea butter, which are detailed in Section 6
(20, 21, 27).
5.3. Coconut Oil

Coconut oil from the coconut palm (Cocus nucifera) is one of two important lauric
oils (see also palm-kernel oil, Section 5.10). Annual production exceeds 3 million
tons and comes mainly from Indonesia and the Philippines (Section 10). It is char-
acterized by its high level of lauric acid (12:0) accompanied by the 8:014:0 acids.
A detailed fatty acid composition of this oil is given in Table 2(b). The oil is used in
the food industry and in the nonfood industry. In the latter case, it is used mainly as
derivatives of the corresponding alcohols (dodecanol or coco alcohol).
Attempts are being made to develop oils of various cuphea species rich in one
or the other of the C8 to C14 acids. (Section 6). Genetically modified rapeseed oil
with lauric acid is also available but has not yet proved to be economically viable
(2830).
5.4. Corn Oil

Corn oil is a major vegetable oil with an annual production of around 2 million tons
obtained from corn or maize (Zea mays) by wet milling, particularly in the United
States. The major acids are palmitic (917%), oleic (2042%), and linoleic (39
63%), and the major triacylglycerols are typically LLL (15%), LLO (21%), LLS
(17%), LOO% (14%), LOS (17%), LSS (5%), OOO (6%), and OOS (4%). Despite
its high unsaturation, the oil has good oxidative stability. The refined oil is used as a
frying oil, a salad oil, and in the production of spreads after partial hydrogenation
(31).
5.5. Cottonseed Oil

The cotton plant is grown for its fiber with the oil being a byproduct and represent-
ing only about 1112% of the gross value of the product. Cottonseed oil was once
the major vegetable oil in competition with the more widely used animal fats.
Today it occupies ninth place in production tables after the four major vegetable
oils (soybean, palm, rape/canola, and sunflower), peanut oil, and three land animal
fats (tallow, lard, and butter). With an annual production of about 3.9 million tons, it
is grown mainly in China (1.1 million tons) and at lower levels of 0.40.2 million
tons in India, the United States, the ex-USSR, Pakistan, Brazil, and Turkey. The oil
is consumed mainly in the country of origin with only limited exports/imports. Cot-
tonseed oil is unusual among commodity vegetable oils in that it contains a rela-
tively high level of palmitic acid (27%) along with oleic (18%) and linoleic
acids (51%). Linolenic acid is virually absent. Low levels of malvalic and sterculic
acids (cyclopropene acids) are removed during refining. Gossypol present in the
crude oil gives it a strong yellow color (32).
5.6. Linseeed (Flaxseed, Linola)

Different varieties of flax (Linum usitatissimum) are grown for fiber and for oil. Lin-
seed oil is well known as one of the most unsaturated vegetable oils, resulting from
its high level of linolenic acid (5060%, Table 4). As a consequence of this, it oxi-
dizes and polymerizes very readily and is used in paints, varnishes, and inks, in the
production of linoleum, and as a sealant for concrete. These uses diminished with
the appearance of alternative petroleum-based products, but the natural oil is com-
ing back into favor on environmental grounds (33).
With recognition of the importance of n-3 acids in the diet, the oil and seed
under the name of flaxseedare being used increasingly in food products both for
humans (cereals and breads) and for animals. This is independent of the growing
use of linola oil (solin) discussed below.
Using chemical mutation, plant breeders in Australia developed a variety of lin-
seed with a low level of linolenic acid ( 2%) and a high level of linoleic acid. This
is called linola and is a linoleic-rich oil like sunflower. Solin is the generic name
given to a similar Canadian flaxseed oil with <5% of linolenic acid. To distinguish
these from traditional linseed oil, they must have yellow seed coats. They can be
grown in the same temperate zones as rapeseed (canola), and the oil is used as an
alternative to sunflower seed oil in the production of spreads rich in EFA. It is being
grown in Australia, Canada, and Europe (3438) (Table 4).
5.7. Olive Oil

Olive oil is a major vegetable oil obtained from the mesocarp of the fruits of
the olive tree (Oleo europaea). Annual production is about 2.5 million tons, and
commercial growth of the tree is confined almost entirely to the Mediterranean
TABLE 4. Fatty Acid Composition of Linseed and Linola Oils.
Oil source Saturated 18:1 18:2 18:3

Linseed 10 16 24 50
Linola 10 16 72 2
228 VEGETABLE OILS
countries of Italy, Greece, Spain, Turkey, and Tunisia. Virgin olive oil is produced
from the first pressing, and other grades of lower quality are produced subsequently.
The oil is characterized by a high level of oleic acid with Codex ranges of 820%
for palmitic acid, 5583% for oleic acid, and 421% for linoleic acid. The major
triacylglycerols are typically OOO (43%), LOO (11%), and POO (22%), and the oil
is characterized by a range of unsaponifiable constituents, that confer high oxidative
stability (60). The oil contains squalene at a higher level (150170 mg/100 ml) than
in other vegetable oils (550 mg/100 ml), and this can be recovered from deodor-
izer distillate (Section 3.5) (39).
5.8. Palm Oil

The oil palm (Elaies guinensis) produces two distinct oilspalm oil from the fleshy
endosperm and palm-kernel oil (Section 5.10) from the kernels. It grows in the tro-
pical regions of Asia, Africa, and America and predominantly in Malaysia and
Indonesia. At an average of about 4 tons per hectare of the two oils combined on
well-managed plantations, the oil palm outcrops all other oil crops. Fruit bunches of
420 kg contain 2002000 individual fruits that furnish palm oil (2024%) and
palm-kernel oil (24%). Through seed breeding, palm trees are being developed
with lower height, higher oil yields, more unsaturated oil, and a higher proportion
of kernel (40, 41).
The supply of palm oil has risen considerably since around 1980. It was almost
24 million tons per annum in 20012002 and is predicted to exceed the production
of soybean oil during the period 20112015 at around 37 million tons. The oil con-
tains almost equal proportions of saturated (palmitic 48% and stearic 4%) and
unsaturated acids (oleic 37% and linoleic 10%). The major triacylglycerols are
POP (3040%) and POO (2030%). The oil is used mainly for food purposes but
finds some nonfood uses. It is a source of valuable byproducts such as carotene and
tocopherols and tocotrienols (vitamin E). Red palm oil is a carefully prepared oil
that retains about 80% (500700 ppm) of the carotenes present in the crude oil and
is a valuable dietary source of these important compounds (42).
Palm oil, melting in the range 2127 C, can be fractionated to give solid (palm
stearin, 3035%, mp 4850 C) and liquid fractions (palm olein, 6570%, mp 18
20 C), thereby extending the range of usefulness of this oil. With improved filtra-
tion procedures, the yield of olein has been increased to 7178%. This olein has a
cloud point of 710 C and can be fractionated further to give even more unsaturated
oleins and palm mid fraction (Table 5). Palm olein is a high-quality, highly stable
TABLE 5. Fatty Acid Composition of Palm Oil and its Fractions.
Oil source 16:0 18:0 18:1 18:2

Palm oil 44 4 39 11
Palm olein 41 4 41 12
Palm stearin 4774 46 1637 310
Source: Lin (42).

frying oil, and the major export of palm oil from Malaysia is now in the form of
palm olein. Palm stearin is the less valuable commodity, but it can be used as a hard
fat in the production of spreads and as a vegetable alternative to tallow in the
oleochemical industry.
5.9. Palm-Kernel Oil

Palm-kernel oil is produced from the kernels of the oil palm, usually by solvent
extraction and is an important lauric oil (see also coconut oil, Section 5.3). Its fatty
acid composition is detailed in Table 2(b). Annual production is about 2.3 million
tons. The kernels originate mainly in the oil palm growing areas of Malaysia and
Indonesia and are crushed almost entirely in the country of origin (28, 29).
5.10. Peanut Oil (Groundnut Oil)

Peanut oil (groundnut), obtained from the legume Arachis hypogea, is also known
as monkeynut oil and arachis oil. The plant is grown widely and especially in India,
China, and the United States. A lot of the nuts are consumed as snacks, but crushing
still yields about 4.6 million tons of oil each year. Its major acids are palmitic
(814%), oleic (3667%), and linoleic (1444%) along with 58% (total) of C20,
C22, and C24 saturated and monoene acids. The major triacylglycerols in one sam-
ple of oil were reported to be LLL (6%), LLO (26%), LLS (8%), LOO (21%), LOS
(13%), OOO (5%), OOS (16%), and other (5%). The oil shows high oxidative sta-
bility and is considered to have a desirable nutty flavor. There is also a high-oleic
variety with 76% of oleic acid (Section 4.2.4) (43).
5.11. Rapeseed Oil

The seed oil of Brassica napus or B. campestris was typically rich in erucic acid
(22:1), and the seed meal had an undesirably high level of glucosinolates. These
components reduced the value of the oil and the protein meal but both have been
bred out of modern rapeseed, known as double zero or canola. Rapeseed (of all
kinds) is now the third largest source of oil at about 14.1 million tons a year
(2000/01) after soybean oil and palm oil. It is grown mainly in Western Europe,
China, India, and Canada (where the canola varieties were developed). Typically
it contains palmitic (4%), stearic (3%), oleic (56%), linoleic (26%), and linolenic
(10%) acids with less saturated acids than any other commodity oil. In one exam-
ple, its major triacylglycerols were LLL (5%), LLO and LnOO (19%), LOO (27%),
and OOO (41%). Rapeseed oil lends itself to genetic modification, and several rape-
seed varieties giving oils with modified fatty acid composition have been devel-
oped. It is still not clear how many of these will be economically viable.
Rapeseed oils containing less linolenic acid, or enhanced levels of lauric acid, stea-
ric acid, oleic acid, or containing unusual acids such as g-linolenic acid, ricinoleic
acid, or vernolic acid have all been developed with a view to commercial exploita-
tion (see Section 6 borage oil). An oleic-rich variety developed in Australia, called
230 VEGETABLE OILS
Monola, contains about 78% oleic acid (Section 4.2.4). Efforts to modify rapeseed
oil by conventional breeding and by genetic engineering are detailed in Sections 9.3
and 9.4 (44).
5.12. Rice Bran Oil

Rice (Oryza sativa) is an important cereal with an annual production of over 500
800 million tons. To produce white rice, the hull is removed and the bran layer is
abraded giving 810% of the rice grain. The bran contains the testa, cross cells,
aleurone cells, part of the aleurone layer, and the germ and includes almost all of
the oil of the rice coreopsis. Gopala Krishna (45) considers that there is a potential
for over 5 million tons of rice bran oil per annum, but present production is only
about 0.7 million tons and not all of this is of food grade. India (0.50 million tons),
China (0.12 million tons), and Japan (0.08 million tons) are the major countries
producing rice bran oil.
Lipases liberated from the testa and the cross cells promote rapid hydrolysis of
the oil, and therefore, it should be extracted within hours of milling. Attempts have
been made to upgrade oil with 30% free acid by reaction with glycerol and the
enzyme Lipozyme (Mucor miehei lipase) followed by neutralization. The major
acids in rice bran oil are palmitic (1218%, typically 16%) oleic (4050%, typically
42%), and linoleic acid (2942%, typically 37%). The oil contains phospholipids
( 5%), a wax that may be removed and finds industrial use, and unsaponifiable
matter including sterols, 4-methylsterols, triterpene alcohols, tocopherols, and
squalene among others.
Refined rice bran oil is an excellent salad oil and frying oil with high oxidative
stability resulting from its high level of tocopherols and from the presence of the
oryzanols (ferulic acid esters of sterols and triterpene alcohols). The oxidative
stability of this oil is exploited in Good Fry Oil. This is a frying oil based on
oleic-rich sunflower oil to which is added up to 6% of rice bran and/or sesame
oil to confer high oxidative stability. Rice bran oil also finds several nonfood
uses (45).
Rice bran oil is reported to lower serum cholesterol by reducing LDL and VLDL
without changing the level of HDL. This effect seems not to be related to fatty acid
or triacylglycerol composition but to the unsaponifiable fraction and probably to the
oryzanols (1.52.0% of the oil). These can be isolated in concentrated form from
rice bran oil soapstock but have not yet been accepted for food use (4651).
5.13. Safflower Oil

Safflower seed oil is a minor oil obtained from the seed of Carthamus tinctorius,
grown particularly in India as a source of a valuable red-yellow or orange dye.
Annual production of seed varies between 600,000 and 800,000. Normally it is a
linoleic-rich oil ( 75% linoleic acid) with LLL (47%), LLO (19%), and LLS (18%)
as the major triacylglycerols. An oleic-rich variety ( 74% oleic acid) has been
developed and designated saffola (52).
5.14. Sesame Oil

Sesame oil comes from the plant Sesamum indicum. This is grown mainly in India
and China and in Myanmar (Burma), Sudan, Mexico, and Egypt with a total annual
production of oil of 0.8 million tons. The seed has 4060% of oil with almost
equal levels of oleic (range 3554, average 46%) and linoleic acids (range 3959,
average 46%). The oil contains sesamin (0.51.1%) and sesamolin (0.30.6%),
which give the oil high oxidative stability (51, 53). It may be added to other
oils to enhance oxidative stability as in the preparation of Good-Fry Oil
(Section 5.11) (50).
5.15. Soybean Oil

The seed of Glycine max is grown as a source of high-grade protein (79%) used in
many human foods and the most dominant protein in animal feed. It is also the
source of a high-quality oil ( 18%). The oil contains palmitic acid (typically
11%, range 714%), oleic acid (typically 23%, range 1930%), linoleic acid (typi-
cally 53%, range 4462%), and linolenic acid (typically 7%, range 411%). Tri-
acylglycerols exceeding 5% include LeLL (7%), LeLO (5%), LLL (15%), LLO
(16%), LLS (13%), LOO (8%), LOS (8%), OOS (5%), and other (19%) at the
approximate levels shown.
Soybean oil is produced in a larger amount than any other vegetable oil and is
grown primarily in the United States, Brazil, Argentina, and China (see Section 10).
It is a very significant part of the diet in the first three of these countries (perhaps
over 80% of dietary fat intake in the United States). Because it contains linolenic
acid, it is usually subjected to a light hydrogenation before being used as a frying
oil or as a salad oil and is more extensively hydrogenated for the production of
spreads and cooking fats (54). Nonfood applications are based mainly on the
high unsaturation of this oil and include surface coatings, dimer acids, and epoxi-
dized oil.
Soybean oil is richly endowed with several minor components that can be recov-
ered during the refining process. These include phospholipids recovered as lecithin,
mixed sterols that serve as a starting material for production of valuable pharma-
ceutical products, and tocopherols (vitamin E) (10).
Attempts are being made to modify the fatty acid composition of soybean oil to
enhance its usefulness. Oils with less or more saturated acid, with less linolenic
acid, and with high levels of oleic acid are in various stages of development (55).
5.16. Sunflower Oil

Sunflower seed oil is obtained from Helianthus annus grown mainly in the USSR,
Argentina, Western and Eastern Europe, China, and the United States. The oil nor-
mally contains 6075% of linoleic acid, >90% of oleic and linoleic acids com-
bined, and virtually no linolenic acid. Its major triacylglycerols are typically
LLL (14%), LLO (39%), LLS (14%), LOO (19%), LOS (11%), and other (3%).
232 VEGETABLE OILS
TABLE 6. Fatty Acid Composition of Tall Oil.
Source Sat (a) 18:1 18:2 (b) (c) (d)

American 2.5 46 36 2 9 1-5
Scandinavian 2.5 30 45 9 5 1-5
(a) 16:0 18:0; (b) pinolenic acid, 5c9c12c-18:3; (c) conjugated diene acids; (d) rosin acids and
unsaponifiable.
It is widely used as a cooking oil and is valued as an important component of soft

spreads. High oleic varieties have been developed. Sunola or Highsun has about
85% oleic acid (some samples reach 90%). These are used to meet the growing
demand for high oleic oils. NuSun with 60% oleic acid has been developed in
the United States [Table 2(a)]. It is hoped that it will replace regular sunflower oil in
the United States, and by 2001, it represented 32% of all sunflower grown there
(56).
5.17. Tall Oil

The term tall oil comes from the Swedish word for pine oil (tallolja). It is a mixture
of fatty acids and some neutral material. Tall oil fatty acids are a byproduct of the
wood pulp industry and result when pine wood chips are digested, under pressure,
with an alkaline solution of sodium sulfate or an acidic solution of sodium sulfite.
Tall oil is produced mainly in North America ( 250,000 tons) and Scandinavia
( 90,000 tons), but the products from these two sources differ in composition
because of the differences in wood species being pulped. The crude extract is dis-
tilled to separate fatty acids (with less than 2% of resin acids) from resin acids (with
less than 2% of fatty acids). The former is a good and cheap source of an oleic
linoleic acid mixture (7580%). However, tall oil fatty acids contain sulfur com-
pounds that interfere with catalytic processes, so the acids are not usually converted
to alcohols or to nitrogen-containing compounds. They are used instead to prepare
dimer acids, alkyds and coatings, detergents, and lubricants and are being examined
for use as solvent, in inks, and for biodiesel production (57). Tall oil pitch is a
valuable source of sterols. These are hydrogenated and acylated for use in
cholesterol-lowering spreads (58) (see Table 6).
6. SPECIALITY AND MINOR OILS
The term speciality oil is a vague term used to describe oils that usually have been
carefully refined to retain the special qualities of color and flavor normally
associated with the oil. For the most part, they are used as food ingredients and
in cosmetic and pharmaceutical products.
These oils are generally available in only limited quantities, and it is essential to
ensure that the sources located will provide a reliable and adequate supply of good
SPECIALITY AND MINOR OILS 233
quality material. As the oils are to be used as dietary supplements, health foods, or
gourmet oils, it is important that the seeds be handled, transported, and stored under
conditions that will maintain quality. It may be necessary to consider growing crops
under conditions that minimize the level of pesticides.
Many fruits are now being processed in large amounts at centralized facilities.
This means that larger quantities of waste products are available at one center
and can be more easily treated to recover oil and other valuable byproducts. This
is particularly relevant in the fruit industry where pips, stones, and kernels are
available in large supply.
Extraction can be carried out in several ways, including cold-pressing at tem-
peratures not exceeding 45 C, pressing at higher temperatures, and/or solvent
extraction. Solvent extraction is not favored for high-quality gourmet oils. Super-
critical fluid extraction with carbon dioxide is an acceptable possibility, but there is
no evidence that this technique is widely used for this purpose. A further possibility
is to use enzymes to break down cell walls followed by extraction under the mildest
possible conditions.
Some speciality oils such as walnut, virgin olive, hazelnut, pistachio, and sesame
can be used as expressed, merely after filtering, but for others, some refining is gen-
erally necessary. On the other hand, if the oil has a characteristic flavor of its own, it
may be desirable to retain this and high-temperature deodorization must then be
excluded or reduced to a minimum. Once obtained in its final form, the oil must
be protected from deteriorationparticularly by oxidation. This requirement neces-
sitates the use of stainless steel equipment, blanketing with nitrogen, and avoiding
unnecessary exposure to heat and light. At the request of the customer, natural and/
or synthetic antioxidant can be added to provide further protection (59).
There follows a description of many minor oils and for convenience these are
presented in alphabetical order (60, 61). Useful fatty acid data for many oils are
given in an AOCS publication (62) and in a book by Ucciani (63).
Aceituno oil (Simarouba glauca). This tree grows in Central and South Amer-
ica. Its seeds produce oil (about 30%), which is rich in oleic acid ( 58%), and con-
tains significant levels of stearic ( 28%) and palmitic (12%) acids. (64).
Almond (Prunus dulcis, P. amygdalis, Amygdalis communis). Almond oil is
generally considered as an oleic-rich oil (6570%), but its fatty acid composition
can vary widely. The triacylglycerol composition of the oil has also been reported.
Low-saturated, high-monounsaturated oils show high oxidative and cold weather
stability (i.e., they are slow to deposit crystals) (65, 66).
Amaranthus (Amaranthus cruentus). Amaranthus or amaranth is a grain con-
taining low levels (69%) of oil. A study of 21 accessions gave the following
results: oil content 58% (mean 6.5), palmitic 822% (mean 19), stearic 14%
(mean 3), oleic 1625% (mean 22), linoleic 4161% (mean 45), and tocopherols
2.87.8 mg/100 g (mean 4.9). Amaranthus oil is unusual in that it has a relatively
high level (68%) of squalene, and this concentration can be raised 10-fold by
short-path high-vacuum distillation. There is no other convenient vegetable source
of squalene apart from olive oil, which has a squalene level of 0.30.7% rising to
1030% in deodorizer distillate (6770).
234 VEGETABLE OILS
Apricot (Prunus armeniaca). Apricot seed oil is used in cosmetics and is avail-
able as a speciality oil for food use. It contains oleic (5874%) and linoleic acids
(2034%). One study gives values of palmitic 5%, stearic 1%, oleic 66%, and lino-
leic acid 29%. With its low content of saturated acids, it shows excellent cold
weather stability (71, 72). The fatty acid composition of the phospholipids has
been reported (73).
Avocado (Persea americana). The avocado grows in tropical and subtropical
countries between 40 N and 40 S and is available particularly from California,
Florida, Israel, New Zealand, and South Africa. Like the palm and the olive, lipid
is concentrated in the fruit pulp (425%) from which it can be pressed. There is
very little oil in the seed (2%). The oil is used widely in cosmetic products as it
is easily absorbed by the skin, and its unsaponifiable material is reported to provide
some protection from the sun. It is also available as a high-oleic speciality oil for
food use. It is rich in chlorophyll, making it green before processing. It contains
16:0 (1020%), 18:1 (6070%), and 18:2 (1015%) as its major fatty acids.
Its unsaponifiable matter, total sterol, and tocopherol levels have been reported
(7478).
Babassu (Orbignya martiana and O. oleifera). This palm, grown in South and
Central America, contains a lauric oil in its kernel. Annual production is small and
uncertain (100300 kt), but Codex values have been established. In line with other
lauric oils, it contains 8:0 (6%), 10:0 (4%), 12:0 (45%), 14:0 (17%), 16:0 (9%),
18:0 (3%), 18:1 (13%), and 18:2 (3%) acid (79).
Blackcurrant (Ribes niger) see Borage.
Borage (Borago officinalis). g-Linolenic acid (6,9,1218:3, GLA) is now recog-
nized as an interesting material with beneficial health properties. Claims have been
made for its use in the treatment of multiple sclerosis, arthritis, eczema, premenstr-
ual syndrome, and other diseases. It is a biological intermediate in the conversion of
freely available linoleic acid to the important but less readily available arachidonic
acid. This change is a three-step process involving 6-desaturation, elongation, and
5-desaturation, of which the first step is rate-determining.
9; 12-18 : 2 ! 6; 9; 12-18 : 3 ! 8; 11; 14-20 : 3 ! 5; 8; 11; 14-20 : 4
GLA is commercially available in three seed oils: blackcurrant, borage, and evening
primrose. The production and use of these oils has been reviewed by Clough
(80, 81). See also references 8285 and Section 2.3 (Table 7).
TABLE 7. Component Acids of Oils Containing c-Linolenic Acid (Typical Results,% wt).
Oil source 16:0 18:0 18:1 18:2 g-18:3 other

Evening primrose 6 2 9 72 10 1
Borage 10 4 16 38 23 9 (a)
Blackcurrant 7 2 11 47 17 16 (b)
(a) 20:1 (4.5), 22:1 (2.5), and 24:1 (1.5).

(b) a-18:3 (13) and n-3 18:4 (3).
Borneo tallow (Shorea stenoptera). This solid fat, also known as illipe butter,
contains palmitic (18%), stearic (46%), and oleic acid (35%). It is one of six per-
mitted fats (palm oil, illipe butter, kokum butter, sal fat, shea butter, and mango
kernel fat), which, in some countries at least, can partially replace cocoa butter
in chocolate (86, 87).
Buffalo gourd (Cucurbita foetidissima). The buffalo gourd is a vine-like plant
that grows in semiarid regions of the United States, Mexico, Lebanon, and India.
The seed contains good quality oil (3239%) and protein. The oil is very variable in
fatty acid composition, thus lending itself to seed breeding. A typical sample con-
tains 16:0 (9%), 18:0 (2%), 18:1 (25%), and 18:2 (62%) (88).
Calendula see Marigold.
Camelina see Gold of Pleasure.
Candlenut (lumbang, kemiri, kukui, Aleurites moluccana). This is a tropical
tree whose nuts contain a very unsaturated oil: 16:0 (68%), 18:0 (23%), 18:1
(1725%), 18:2 (3845%), and 18:3 (2530%). Its iodine value, however, is not
as high as that of linseed oil. It is used for cosmetic purposes and has been recom-
mended for the treatment of burns (89).
Caraway (Carum carvii). This is one of a group of plants whose seed oils con-
tain petroselinic acid (618:1). This acid reaches levels of 3543% in caraway, 66
73% in carrot, 3175% in coriander, and 80% in parsley. This isomer of oleic acid
has some potential use as a source of lauric and adipic acids, produced by oxidative
cleavage. The latter, an important component of many polyamides (nylons), is
usually made from cyclohexane by a reaction that is reported to be environmentally
unfriendly (90).
Carrot (Daucus carta). See caraway.
Cashew (Anacardium occidentale). Toschi et al. (91) have given details of the
fatty acids, triacylglycerols, sterols, and tocopherols in cashew nut oil. The major
fatty acids are palmitic (914%), stearic (612%), oleic (5765%), and linoleic
(1618%), and the major triacylglycerols are OOO, POO, OOSt, OOL, and POL.
Cherry (Prunus cerasus). Obtained by cold pressing and filtering, this oil is sold
in the unrefined state for use as a speciality oil for salad dressings, baking, and shal-
low frying and in the production of skin-care products. Its fatty acid composition is
unusual in that in addition to oleic (3040%) and linoleic acids (4050%), it also
contains a-eleostearic acid (612%, 9c11t13t-18:3). Some of these potential uses
are perhaps surprising for an oil containing a conjugated triene acid (9295). The
fatty acid composition of the phospholipids has been reported (96).
Chia (Salvia hispanica). Chia seeds contain 3238% of a highly unsaturated oil
(97). The fatty acid composition for five samples from Argentina have saturated
acids (911%), oleic (78%), linoleic (2021%), and linolenic acid (5263%).
Chinese vegetable tallow and stillingia oil (Sapium sebiferum, Stillingia sebi-
fera). This seed is unusual in that it yields lipid from its outer seed coating (Chinese
vegetable tallow, 2030%) and from its kernel (stillingia oil, 1017%), which are
very different (96). The former, with 75% palmitic acid and 2025% oleic acid, is
mainly a mixture of PPP ( 70%) and POP (2025%) triacylglycerols and is a
potential confectionery fat. However, it is difficult to obtain the fat free of stillingia
236 VEGETABLE OILS
oil (the kernel oil), which is considered to be nutritionally unacceptable. Stillingia

oil is different, with oleic (13%), linoleic (23%), and linolenic acids (47%) and
novel C8 (hydroxy allenic) and C10 (conjugated dienoic) acids that occur together
as a C18 estolide attached to glycerol at the sn-3 position, thus,
Glyc-OCOCH CHCH2 4 OCOCH

C CHCH2 4 CH3
CHCH
Coriander (Coriandrum sativum). See caraway. Attempts are being made both
to develop coriander as an agricultural crop and to transfer the necessary -6 desa-
turase to the rape plant (98).
Crambe (Crambe abyssinica, C. hispanica). Present interest in this oil, particu-
larly in North Dakota and in Holland, depends on the fact that it is a potential
source of erucic acid (5055%) that finds several industrial uses. This was once
the major acid in rapeseed oil, but modern varieties of this seed produce a low-
erucic oil (such as canola) suitable for food use. High-erucic rapeseed oil is still
grown for industrial purposes, and attempts are being made to increase the level
of this C22 acid from around 50% to over 65% and even to 90% by genetic engi-
neering (2223, 44, 99102).
Cuphea. Cuphea plants furnish seeds with oils that may be rich in C8, C10, C12,
or C14 acids. They generally contain >30% of oil and are expected to produce a
commercial crop in the period 20052010. Problems of seed dormancy and seed
shattering have already been solved. As markets for lauric oils already exist, there
should be no difficulty in substituting cuphea oils. More recently, it has been
reported that cuphea will be used as a commercial source of lauric acid from
2003 onward (30, 102, 103). Pandey et al. (104) have described the oil (1729%)
from Cupea procumbens containing 8995% of decanoic acid. See also Section 9.2.
Dimorphotheca The seed of Dimorphotheca pluvialis is not very rich in oil (13
28%, typically about 20%), but it contains an unusual C18 hydroxy fatty acid
( 60%) with hydroxyl group adjacent (allylic) to a conjugated diene system.
This is very unstable and easily dehydrates to a mixture of conjugated 18:3 acids
(105).

CHCH
CH3 CH2 4 CH

CHCHOHCH2 7 COOH

Dimorphecolic acid 9-OH10t12c-18 : 2
Evening Primrose (Oenothera biennis, O. lamarckiana, and O. parviflora). see

Borage.
Gold of Pleasure (Camelina sativa, also called false flax). In addition to its
interesting fatty acid composition, this plant attracts attention because it grows
well with lower inputs of fertilizers and pesticides than traditional crops like rape-
seed and linseed. The plant can also be grown on poorer soils and shows better
gross margins than the other two plants after allowing for direct costs and (EU) sub-
sidy payments. The seed yield is in the range 1.53.0 t/ha, and the oil content is
between 36% and 47%. The oil has an unusual fatty acid composition. It contains
significant levels of oleic acid (1020%), linoleic acid (1624%), linolenic acid
(3040%), and C20 and C22 acids, especially 20:1 (1523%). Another publication
reports 3038% oil containing oleic (1420%), linoleic (1924%), linolenic (27
35%), eicosenoic (1215%), and other acids (1220%) along with a range of tocols
(522, mean 17 mg/100 g). Despite its high level of unsaturation, the oil shows rea-
sonable oxidative stability. Attempts are being made to optimize the agronomy. Its
use in paints, varnishes, inks, cosmetics, and even as a food oil is being examined
and developed. Permission for food use has been granted in France and Britain
(106110).
Grapeseed (Vitis vinifera). These seeds produce variable levels of oil (620%),
now available as a gourmet oil and for which Codex values have been reported. The
oil is rich in linoleic acid (6076%) and contains palmitic (68%), stearic (36%),
and oleic acids (1225%). In common with other oils rich in linoleic, it is reported
to have a beneficial effect on the skin (79). Moret et al. (111) have described the
effect of processing on the content of polycyclic aromatic hydrocarbons in this oil.
Hazelnut (Corylus avellana, also called filberts). The oil is rich in oleic acid
(6575% or even higher) and contains linoleic acid (16 22%). Its levels of satu-
rated acids are low. Grown in Turkey and New Zealand, the nuts produced 55
63% of oil with saturated acids (68%), monoene acids (7480%), and linoleic
acid (68%). A recent study indicates the presence of several monoene acids in
the C16 C22 range, although this may refer to a different species (79, 112115).
Hemp (Marijuana, Cannabis sativa). Hemp seed oil has an interesting fatty acid
composition. One report gives the followimg values: palmitic (49%), stearic (2
4%), oleic (815%), linoleic (5360%), a-linolenic (1525%), g-linolenic (05%),
and stearidonic acid (03%). The oil is being used in cosmetic formulations (116).
Evidence from a study in Finland indicates that dietary consumption of hemp seed
oil leads to increased levels of g-linolenic acid in blood serum (117). The growing
of hemp is banned in the United States, and therefore, hemp seed oil must be
imported into that country (118119).
Honesty (Lunaria annua). This seed oil contains significant levels of erucic
(22:1, 41%) and nervonic acids (24:1, 22%) and is being studied as a new crop
because it is a good source of the latter acid, which may be useful in the treatment
of demyelinating dizease (120).
Illipe (Shorea stenoptera, also called Borneo tallow). This is one of a group of
tropical fats that are often confused with one another. They generally resemble
cocoa butter in their proportions of palmitic, stearic, and oleic acids and therefore
have similar triacylglycerol composition and display similar melting behavior.
Values of 18%, 46%, and 35% have been reported for palmitic, stearic, and oleic
acids and POP (7%), POSt (34%), and StOSt (47%) for the major triacylglycerols
(121125). It is one of six permitted fats (palm oil, illipe butter, kokum butter,
sal fat, shea butter, and mango kernel fat), which, in some countries at least, can
partially replace cocoa butter in chocolate (8687).
Kapok (Bombax malabaricum, Ceiba pentandra). This name is applied to a
number of tropical trees of the bombax family. The oil is a byproduct of kapok fiber
production. Its major component acids are palmitic (22%), oleic (21%), and linoleic
238 VEGETABLE OILS
TABLE 8. Fatty Acids and triacylglycerols of Kokum and Madhua Fats.
Fatty Acids Triacylglycerols
Oil source 16:0 18:0 18:1 18:2 StOSt POSt POP

Kokum 2.0 49.0 49.0 0 72.3 7.4 0.5
Madhua 23.5 20.0 39.0 16.7 10.6 22.2 18.9
Stearin
15.7 37.8 35.5 11.1 46.2 15.0 9.7
Obtained by dry fractionation of a 1:1 mixture of the two oils.
(37%), but it also contains about 13% of cyclopropene acids (malvalic and stercu-
lic), which make it unsuitable for food use.
Kokum (Garcinia indica). Both kokum and mahua fats are rich in saturated and
oleic acid and contain high levels of SOS triacylglycerols. They can be fractioned
separately or as blends of the two oils to produce stearins that can be used as cocoa
butter extenders (Table 8) (125). Kokum butter is one of six permitted fats (palm
oil, illipe butter, kokum butter, sal fat, shea butter, and mango kernel fat), which, in
some countries at least, can partially replace cocoa butter in chocolate (90).
Lesquerella. The only oil of significance with a hydroxy acid is castor oil (Sec-
tion 5.1) but among the new crops being seriously developed are two containing
hydroxy acids. Lesquerella oils have some resemblance to castor oil, but Dimor-
photheca pluvialis seed oil contains a different kind of hydroxy acid.
Plants of the Lesquerella species are characterized by the presence of the C20
bis-homologue of ricinoleic acidlesquerolic acidsometimes accompanied by
other acids of the same type at lower levels:
Ricinoleic acid 12-OH 9-18 : 1

Densipolic acid 12-OH 9; 15-18 : 2
Lesquerolic acid 14-OH 11-20 : 1
Auricolic acid 14-OH 11; 17-20 : 2
A typical analysis of L. fendleri seed oil showed the presence of 16:0 (1%), 18:0
(2%), 18:1 (15%), 18:2 (7%), 18:3 (14%), lesquerolic (54%), and auricolic (4%)
acids. As lesquerolic acid is the C20 homologue of ricinoleic with the same
b-hydroxy alkene unit, it undergoes similar chemical reactions but produces
(some) different products. For example, pyrolysis should give heptanal and 13-tri-
decenoic acid (in place of 11-undecenoic acid). This could be converted to 13-ami-
notridecanoic acid, the monomer required to make nylon-13. Similarly, alkali-
fusion will give 2-octanol and dodecanedioic acid in place of decanedioic (sebacic)
acid. This C12 dibasic acid is already available from petrochemical products and has
a number of applications. A recent account of the status of this oil is available
(126).
Macadamia (Macadonia integrifolia, M. tetraphylla). The nuts are used as a
snack food. They are rich in oil (6070%), which is used in cosmetics and is avail-
able as a gourmet oil. It is characterized by its high level of monoene acids [total
80%, 16:1 1623%, 18:1 5565%, 20:1 13%] and is a convenient source of the
relatively uncommon palmitoleic acid. Its high level of monoene acids makes it
good for skin care, but low levels of tocopherols limit its oxidative stability
(127128).
Mahua (Madhuca latifolia). see Kokum fat and Mango kernel fat.
Mango (Mangifer indica). Mango is consumed in large quantities as fruit. The
kernel contains 712% of lipid with palmitic (318%), stearic (2457%), oleic (34
56%), and linoleic acid (113%). In a typical case, these values were 10.3%, 35.4%,
49.3%, and 4.9%, respectively. It is fractionated to give an olein that is lower melt-
ing than mango fat and has excellent emollient properties and a stearin. The stearin
can serve as a cocoa butter equivalent (POP 1%, POSt 12%, StOSt 56%) (124, 129,
130) and as a component with fractioned mahua fat of a trans-free bakery shorten-
ing. It is one of six permitted fats (palm oil, illipe butter, kokum butter, sal fat, shea
butter, and mango kernel fat), which, in some countries at least, can partially
replace cocoa butter in chocolate (86).
Marigold (Calendula officinalis). Interest in this seed oil is based on the fact that
it contains significant levels (5362%) of calendic acid along with linoleic acid
(2834%). Calendic acid (8t,10t,12c-18:3) is a conjugated trienoic acid, and this
makes the oil an effective drying agent. Its alkyl esters can be used as a reactive
diluent in alkyd paints replacing volatile organic compounds. The crop is being stu-
died particularly in Europe (131133).
Meadowfoam (Limnanthes alba). This oil is unusual in that over 95% of its
component acids are C20 or C22 compounds and include 520:1 (6367%), 5
22:1 (24%), 1322:1 (1618%), and 5,1322:2 (59%). It is being grown in the
United States, and its potential uses are being thoroughly examined. Winter culti-
vars now being developed are expected to improve the suitability of the crop to con-
ditions in Northern Europe. Potential uses of this oil include cosmetic applications,
production of dimer acid, as a lubricant, and via a wide range of novel derivatives
based on reaction at the 5 double bond (134138).
Melon (Citrullus colocythis and C. vulgaris). This seed oil has been examined in
terms of its fatty acids and phospholipids by Akoh and Nwosu (139). The major
fatty acids in the total lipids are palmitic (11% and 12%), stearic (7% and 11%),
oleic (10% and 14%), and linoleic acid (71% and 63%) for two samples.
Mowrah (Madhuca latifolia, M.longifolia, M.indica). This is mainly an Indian
product where the fat is used for edible and industrial purposes. The nuts contain
46% of oil with variable levels of palmitic (1532%), stearic (1626%), oleic (32
45%), and linoleic acid (1418%) (140).
Mustard (Brassica alba, B. hirta, B. nigra, B. juncea, B. carinata). The seeds
contain 2440% of oil characterized by the presence of erucic acid. Typical values
are oleic 23%, linoleic 9%, linolenic 10%, eicosenoic 8%, and erucic acid 43%
(141, 142). The plant is grown extensively in India (59, 79).
Canadian investigators have bred Brassica juncea (orienal mustard) from an
Australian line with low erucic acid and low glucosinolate so that it has a
fatty acid composition (palmitic 3%, stearic 2%, oleic 64%, linoleic 17%, and
linolenic acid 10%) similar to that of canola oil from B. napus and B. rapa. This
makes it possible to expand the canola growing area of Western Canada (143).
240 VEGETABLE OILS
Neem (Azadirachta indica). This interesting seed oil contains chemicals used to
control 200 species of insects. The oil prevents some insect species from maturing
past the larval stage (144).
Nigella (Nigella sativa, black cumin). Typically, nigella oil contains palmitic
(10%), oleic (35%), and linoleic acid (45%). Related species (N. arvensis and
N. damascena) give similar oils with less oleic and more linoleic acid. The presence
of low levels of 20:1 (11c, 0.51.0%) and higher levels of 20:2 (11c14c, 3.6 4.7%)
in all of these oils may be of taxonomic significance. In one analysis, the oil con-
tained the following major triacylglycerols: LLL 25%, LLO 20%, LLP 17%, LOP
13%, and LOO 10% reflecting the high level of linoleic acid. The seeds appear to
contain an active lipase, and the oil quickly develops high levels of free acid. The
oil is reported to be a good source of thymoquinone and to assist in the treatment of
prostate problems (145148).
Niger (Guizotia abyssinica). This oil comes mainly from Ethiopia. The seeds
contain 2939% of oil rich in linoleic acid (7179%) along with palmitic, stearic,
and oleic acids, each at levels of 611%. It is used for both edible and industrial
purposes. It is rich in a-tocopherol and is therefore a good source of vitamin E
(149).
Nutmeg (Myristica malabarica and other M. species). Not surprisingly, consid-
ering its botanical name, seeds of the Myristica species are rich in myristic acid
( 40%). Higher levels (6072%) were quoted in earlier work (150).
Oats (Avena sativa). This grain seed contains 48% of lipid, although somewhat
more in certain strains. The major component acids are palmitic (1328%), oleic
(1953%), linoleic (2453%), and linolenic acid (15%) The oil contains triacyl-
glycerols (51%), di- and monoacylglycerols (7%), free acids (7%), sterols and
sterol esters (each 3%), glycolipids (8%), and phospholipids (20%). The special
features of this oil are used in various ways. It is reported to show cholesterolemic
and antithrombotic activity, it is present in Olibra used as an appetite-suppressant, it
is used in cosmetics by virtue of its glycolipids (151153), and it can be used in
baking at levels as low as 0.5% to increase loaf volume. Oat lipids are the subject
of recent reviews (154, 155).
Oiticica (Licania rigida). The kernel oil obtained from this Brazilian tree is
characterized by its high level ( 78%) of licanic acid (4-oxo-9c11t13t-octadeca-
trienoic acid)a keto derivative of the more familiar eleostearic acid. The oil
shows drying properties but does not dry as quickly as tung oil (156).
Parsley (Petroselinium sativum). See carrot.
Passionfruit (Passiflora edulis). This popular fruit contains about 20% of oil in
its seed and is available as a gourmet oil for use in speciality foods and salad dres-
sings. It is a linoleic-rich (6575%) but also contains palmitic (812%) and oleic
acids (1320%). Its high level of linoleic acid makes the oil good for skin care
(157).
Perilla (Perilla frutescens). Perilla is a linolenic-rich oil (5764%) used as a dry-
ing oil. It also contains oleic (1315%) and linoleic acids (1418%) and comes
mainly from Korea or India. Recent descriptions of this oil come from these two
countries (158160).
Pistachio (Pistachio vera). Pistachio nuts, produced mainly in Iran, are widely
consumed as shelled nuts. They contain about 60% of an oil that may be used for
cooking and frying. Mean fatty acid values for five varieties are given as palmitic
(10%), stearic (3%), oleic (69%), and linoleic (17%). Triacylglycerol composition
has been suggested as a method of determining the country of origin of pistachio
nuts (161163).
Poppy (Papaver somniferium). Opium is obtained from unripe capsules and
from the straw of the poppy plant. The narcotic is not present in the seed, which
is much used for birdseed. It contains 4070% of a semi-drying oil used by artists
and as an edible oil. Rich in linoleic acid (72%), it also contains palmitic (10%),
oleic (11%), and linolenic acids (5%) (79, 164).
Purslane (Portulaca oleracea). The plant (leaves, stem, and whole plant) is
reported to be the richest vegetable source of n-3 acids, including low levels of
the 20:5, 22:5, and 22:6 members. This is such a surprising result that it should
be confirmed. These acids have not been identified in the seed oil, which contains
palmitic (15%), stearic (4%), oleic (18%), linoleic (33%), and linolenic acids (26%)
(165).
Sal fat (Shorea robusta). This tree, which grows in Northern India, is felled
for timber. Its seed oil is rich in stearic acid, and it can be used as a cocoa butter
equivalent (CBE). The major acids are palmitic (28%), stearic (3548%), oleic
(3542%), linoleic (23%), and arachidic acid (611%). Its major triacylglycerols
are of the SUS type required of a cocoa butter equivalent. Sal olein is an excellent
emollient, and sal stearin, with POP 1%, POSt 13%, and StOSt 60%, is a superior
cocoa butter equivalent (122124). It is one of the six permitted fats (palm oil, illipe
butter, kokum butter, sal fat, shea butter, and mango kernel fat), which, in some
countries at least, can partially replace cocoa butter in chocolate (86).
Sea buckthorn (Hippophae rhamnoides). This is a hardy bush growing wild in
several parts of Asia and Europe and now cultivated in Europe, North America, and
Japan. It is resistant to cold, drought, salt, and alkali. Different oils are available
from the seeds and from the pulp/peel, but these are not always kept separate. Sev-
eral health benefits are claimed for this oil, which is now available in encapsulated
form and is being incorporated into functional foods. The oil is rich in sterols, car-
otenoids, and tocopherols. The seed oil is rich in 18:1, 18:2, and 18:3, but the berry
oil contains significant levels of 16:1 (1622%) (166169).
Shea (Butyrospermum parkii, shea butter, karite butter). This fat comes from
trees grown mainly in West Africa and contains an unusually high level of unsapo-
nifiable material ( 11%), including polyisoprene hydrocarbons. It is rich in stearic
acid, but its fatty acid composition varies with its geographical source. It contains
palmitic (48%), stearic (2358%), oleic (3368%), and linoleic acid (48%). It can
be fractionated to give a stearin (POP 1%, POSt 8%, and StOSt 68%), which can be
used as a cocoa butter equivalent (79, 122124). It is one of the six permitted fats
(palm oil, illipe butter, kokum butter, sal fat, shea butter, and mango kernel fat),
which, in some countries at least, can partially replace cocoa butter in chocolate (86).
Tobacco. Tobacco seeds contain an oil rich in linoleic acid (>70%) but with vir-
tually no linolenic acid. After refining, it can be used for edible purposes or as a
242 VEGETABLE OILS
TABLE 9. Minor Oils Rich in Particular Fatty Acids.
Fatty Acid Sources
Lauric and myristic Babassu, nutmeg

Stearic Aceituno, illipe, mango, mowrah, sal, shea
Petroselinic Caraway, carrot, coriander, parsley
Oleic Aceituno, almond, avocado, apricot, cashew, hazelnut, pistachio
Linoleic Amarantus, buffalo gourd, grape seed, hemp, melon, nigella, niger,
passionflower, poppy, tobacco, walnut, wheatgerm
Linolenic Candlenut, flax, gold of pleasure, hemp, mustard, perilla
Conjugated triene acids Cherry, marigold, oiticica, tung,
Long-chain monoene
acids C20C24 Crambe, gold of pleasure, honesty, meadowfoam, mustard
nonyellowing drying oil. In one sample of the oil that was analyzed, the major
triacylglycerols were LLL (38%), LLO (24%), and LLS (20%) (79).
Tung oil (Aleurites fordii). This oil comes mainly from China, which explains its
alternative name of China wood oil. It is characterized by the presence of a conju-
gated triene acid (a-eleostearic, 9c11t13t-18:3, 69%). The oil dries more quickly
than linseed with its nonconjugated triene acid, but oxidized tung oil contains less
oxygen (5%) than does oxidized linseed oil (12%). Put another way, tung oil hard-
ens at a lower level of oxygen-uptake than linseed oil. This oil is exported mainly
from China (3040,000 tons) and is imported mainly by Japan, South Korea,
Taiwan, and the United States (each 60007000 tons). Starting in 1993, attempts
have been made to develop this crop in Mississippi. It is planned to have 15,000
acres planted by 2006 producing 30,000 tons of oil (79, 170).
Walnut (Juglans regia). Walnut oil is an unsaturated oil containing both linoleic
(5060%) and linolenic acids (1315%) and rich in tocopherols ( 1500 mg/kg of
oil). It is used as a gourmet oil in Japan, France, and other countries. A recent paper
gives the detailed composition (fatty acids, triacylglycerols sterols, and tocophe-
rols) of oil extracted with hexane and with supercritical carbon dioxide (171).
Wheatgerm (Triticum aestivum). This oil is highly unsaturated with linoleic
( 60%) and some linolenic acid ( 5%). It is valued for its high tocopherol levels
( 2500 mg/kg of oil) (172173).
Oils rich in a particular fatty acid are listed in Table 9.
7. MODIFICATION OF OILS AND FATS
The oils and fats provided by nature are not always ideal for their ultimate use,
whether for food or nonfood purposes, and scientists and technologists have
devized procedures for changing the natural oils. The major reasons for modifica-
tion are nutritional, physical, and economic.
MODIFICATION OF OILS AND FATS 243
7.1. Nutritional Reasons for Modifying a Fat

Although ideas about the nutritional values of oils and fats change over time and
will probably continue to do so, present consensus is based on the lipid hypothesis
that is conveniently expressed in the following four statements:
Diets with high contents of fat, of saturated fatty acids (SFA), and of
cholesterol lead to high concentrations of cholesterol in blood and especially
in low-density lipoprotein (LDL).
This leads to high morbidity and mortality from coronary heart disease
(CHD).
Reducing the amount of fat/SFA/cholesterol in the diet reduces blood
cholesterol and especially LDL cholesterol.
This reduction leads to a lower risk of CHD and eventually to lower morbidity
and mortality from the disease.
Perhaps this is oversimplified. The following points certainly need to be

considered:
All saturated acids do not behave identically. Short- and medium-chain acids
up to 10:0 are rapidly metabolized by a different pathway from the longer
chain acids and have no effect on cholesterol level. Stearic acid appears to
lead only to a marginal cholesterol rise. This leaves lauric acid (12:0), myristic
acid (14:0), and palmitic acid (16:0), among which myristic acid has the
greatest cholesterol-raising effect. Some authorities have argued that even this
is only a problem if there is an inadequate intake of linoleic acid.
Monounsaturated acids (mainly oleic acid) are in great favor at the present
time, but this relates only to the natural cis-isomers, and there is concern that
acids with trans-unsaturation behave like the saturated acids in their effect on
blood cholesterol levels. Dietary acids with trans-unsaturation come mainly
from three sources: (1) Partially hydrogenated vegetable oils resulting mainly
from heterogeneous catalytic reduction (nickel catalyst) of linoleic acyl
groups. The product is a complex mixture of cis- and trans-18:1 esters. The
major trans-acids in this mixture are 812, but others are also present. (2)
Animal fats in ruminant meat and in dairy products contain trans-acids
formed by enzymatically controlled biohydrogenation of linoleic acid. These
are mainly 18:1 acids with the 11t isomer (vaccenic) dominant, but they also
contain some diene acids referred to as conjugated linoleic acid (CLA) and
mainly rumenic acid (9c11t-18:2). CLA is considered to have some positive
health benefits. (3) Trans-isomers of PUFA result from high-temperature
isomerization during deodorization ( 250 C). The detailed composition of
dietary trans-18:1 depends mainly on the ratio of ruminant fats to partially
hydrogenated vegetable oils. The difference between French diets (rich in
dairy fats) and U.S. diets (rich in partially hydrogenated soybean oil) accounts
for the differing dietary intake of trans-acids in these two countries (174).
244 VEGETABLE OILS
So far as polyunsaturated fatty acids are concerned, it is now considered that

n-6 acids (linoleic and its metabolites) should not exceed present levels of
around 6% of energy and that the n-6/n-3 ratio should be 5:1 or lower. There
have even been serious claims that linoleic acid should not exceed 3% of
energy (175). Some countries, including the United States, have a very low
intake of n-3 acids and possibly an excessive intake of n-6 acids leading to a
high n-6/n-3 ratio (18, 19).
Unsaturated fat intake should always be accompanied by an adequate supply
of antioxidants from fruit and vegetables.
7.2. Physical Reasons for Modifying a Fat

The important physical properties are most commonly associated with crystalliza-
tion, crystal form, and melting behavior. For example, frying oils and lubricants
ideally do not contain crystals and should therefore be free of those triacylglycerols
that crystallize readily and promote crystallization. For spreads, where solids are
needed, it is desirable that these be in the b0 -crystal form and remain in this
form. b0 -Crystals are relatively small and can incorporate large volumes of oil.
They give the product a glossy surface and a smooth luster. b-Crystals, on the other
hand, although initially small, grow into needle-like agglomerates that produce a
grainy texture and are less able to incorporate liquid. Oils with mixed chain-lengths
(usually C16 and C18) are more likely to exist in the b0 crystalline form, whereas
those comprising almost entirely of C18 acids are known to be b-tending (176).
7.3. Economic Reasons for Selecting a Fat

Through changes that can be produced in natural oils and fats by application of
appropriate technologies (Section 8), there is a high measure of interchangeability
among these materials. Food processors in different countries use different recipes
to produce very similar properties. This flexibility means that economic factors can
also be part of the basis of selection. See Section 8.1.
8. TECHNOLOGICAL PROCEDURES USED FOR LIPID

MODIFICATION
Although it may be necessary to modify natural oils to achieve desired functionality

and properties, there is an economic cost for all of these processes and they will not
be undertaken unnecessarily. Most of these procedures are discussed in detail else-
where in this series so that only a brief outline will be given here. The objective is to
give an overall view of the range of procedures (177) (see Table 10).
8.1. Blending
The mixing of oils and fats to produce blends with improved nutritional or physical
properties has a long history. This method continues to find favor and is illustrated
TECHNOLOGICAL PROCEDURES USED FOR LIPID MODIFICATION 245
TABLE 10. Methods Employed to Extend the Usefulness and Improve the Properties
of Oils.
Blending mixing of two or more oils

Fractionation separating oils into two or more fractions
Partial hydrogenation saturation of some double bonds accompanied
by double-bond isomerization
Interesterification by chemical reorganization of fatty acids among
or enzymatic catalyst triacylglycerol molecules
Domestication of wild crops conversion of wild crops to crops that can be
cultivated commercially
Seed-breeding by traditional methods interspecies crossing using irradiation
or mutagenesis if necessary
Seed-breeding by genetic modification crossing between species
in the production of Good-Fry oil (see below). Most spreads contain blends of two
or more oils to combine desirable nutritional and essential physical properties.
Interesterification is usually carried out on oil blends. Oils are also blended to
obtain the desired mixture at minimum cost and computer programs to give the
best solution have been developed (178).
Good-Fry is a blend of high-oleic vegetable oil such as sunflower mixed with up
to 6% of sesame and/or rice bran oil, both of which show high oxidative stability by
virtue of the antioxidants among their minor components. It is of interest that some
of these antioxidants are particularly active at frying temperatures. The nature of
the bulk oil (with its low levels of linoleic acid) and of the minor oils (with their
high oxidative stability) combine to produce a very stable frying oil. Good-Fry can
therefore be used longer than other frying oils. This makes it safer because of its
reduced levels of oxidized and polymerized products and more economical because
it does not have to be replaced so frequently (50).
8.2. Fractionation
Fractionation is a procedure for separating oils and fats into two or more compo-
nents depending on their solubility and melting point. This topic has been reviewed
by Timms (179) and Gibon and Tirtiaux (180). The less-soluble, higher melting
fractions are called stearins, and the more-soluble, lower melting fractions are
called oleins. The two products extend the range of use of the original oil or
fat. Sometimes both fractions have added value, but on other occasions, only one
fraction is of enhanced value and efforts have to be made to find a use for the less
valuable fraction. Fractionation can be repeated to give even more refined fractions,
but this is only commercially practicable when high-value products are obtained,
such as cocoa butter replacers. Although other procedures have been employed
in the past, this process is now usually carried out through dry fractionation
(179182).
246 VEGETABLE OILS
Dry fractionation is a two-step operation involving crystallization that should be

allowed to proceed slowly to the equilibrium state, followed by filtration of the
solid from the liquid phase. Crystallization occurs over several hours and requires
good temperature control. The temperature must be lowered at a fixed rate to the
selected value, and this must be combined with efficient but slow agitation. Good
filtrationaiming at complete separation of solid and liquidis important and may
be carried out under reduced pressure using a Florentine filter or under pressures up
to 5 Mpa (50 bar) with a membrane filter. Fractionation is applied mainly to palm
oil but also to lauric oils (coconut and palmkernel), butter oil, beef tallow, hardened
soybean, and cottonseed oil.
Palm oil is fractionated more than any other oil. A single fractionation converts
palm oil (IV 5153) to palm olein (IV 5659) and to hard stearin (IV 3236). Each
of these can be fractionated a second or a third time to give a range of products,
including superolein (IV 6466), topolein (IV 7072), soft stearin (IV 4042),
super stearin (IV 1721), soft palm mid-fraction (IV 4248), and hard palm
mid-fraction (IV 3236). These materials have a wide range of food and nonfood
uses and extend considerably the use of palm oil.
Palm-kernel oil (IV 18) is fractionated to give a stearin of IV 7, which can be
used as a cocoa butter substitute and as an olein of IV 25. These fractions are also
useful after complete hydrogenation.
8.3. Hydrogenation
Just over 100 years ago (1897), Sabatier and Senderens demonstrated that olefinic
compounds could be reduced with hydrogen in the presence of nickel or other
metallic catalyst. Shortly after, the German chemist Normann applied the process
to unsaturated fatty materials. Partial hydrogenation has since developed into a
much-used process for modifying liquid oils from oilseeds or from fish (178).
In 1990, it was claimed that among all edible fats, one-third was hydrogenated
and only one-tenth was fractionated or interesterified. These proportions are now
probably different because of the increasing volumes of palm oil available for
fractionation and of concern about trans-acids formed during partial hydrogenation.
Hydrogenation is appropriate for highly unsaturated oils such as soybean, rapeseed,
and cottonseed, and for fish oils, whereas fractionation is better applied to palm oil
and other more saturated oils. The following changes take place when an oil is par-
tially hydrogenated:
There is a change in the melting behavior of the oil as a consequence of the

increased proportion of saturated and/or trans-monoene acids and this affects
spreadability, oral response, and baking performance.
There is an improvement in stability toward atmospheric oxidation resulting
from reduced levels of the methylene-interrupted polyene acids that are so
easily oxidized.
There is a reduction in the nutritional value of the product, related to lower
levels of essential fatty acids (a-linolenic and linoleic acids) and enhanced
levels of both trans-monoene and saturated acids. It is possible to add

back essential fatty acids into the final product by blending with appropriate
oils.
Through reaction with hydrogen in the presence of a heterogeneous catalyst,
the unsaturated centers in the oil being hydrogenated may suffer one of three
fates. (1) The double bond can react with hydrogen and become saturated; as a
consequence of this, diene acids are reduced to monoenes and monoene acids
become saturated. (2) The double bond may change configuration and the
natural cis-isomers may become largely trans: Such acids have a higher
melting point than the cis-isomers, so stereomutation leads to a rise in melting
point without any uptake in hydrogen or change in iodine value. (3)
Interaction among double bond, catalyst, and hydrogen can lead to double
bond migration.
When partially hydrogenated, linoleic acid might be expected to give only 9 and
12 C18 monoenes. Each of these may then react further, but under conditions of
extended selective hydrogenation, the product is more complex and the C18 mono-
ene esters may include the cis- and trans-isomers from 5 through to 15 (i.e., 22
isomers).
Partial hydrogenation is a flexible process and can produce different products
depending on the:
nature of the starting material

extent of hydrogenation
selectivity of the catalyst, which influences the proportion of cis- and trans-
monoenes and of saturated acyl chains.
The nature of the hydrogenation products is controlled by the process conditions.

Important factors are catalyst properties (pore diameter, pore length, activity level,
and amount), reaction temperature, hydrogen pressure, and the degree of agitation
(which affects the transfer of hydrogen and oil to and from the catalyst surface).
Hydrogenation (reduction) is favored by a high concentration of hydrogen on the
catalyst through increased pressure or increased stirring. Isomerization is favored
by factors such as increased temperature, more catalyst, a more active catalyst,
or a more highly unsaturated oil, all of which lead to an increased hydrogen demand
that cannot be completely satisfied.
Dikstra (177, 184) has proposed a modification of the HoriutiPolanyi mechan-
ism to explain the changes that occur during partial hydrogenation of fatty acids and
their esters. In the sequence given below, the horizontal line shows the conversion
of diene (D) to monoene (M) and of monoene to saturated acid/ester via the half
hydrogenated states DH and MH. The steps shown vertically are the reverse pro-
cesses whereby DH reverts to diene and MH reverts to monoene. It is during these
reverse processes that trans- and positional isomers are formed. There are six stages
altogether, and it is important to know the relative rates of these.
248 VEGETABLE OILS
In the conversion of D to M, the first step is rate-determining and the second

step is fast. The conversion of DH back to D is slow and is only important in
the unusual situation that hydrogen is present in very low concentration.
In the conversion of M to S, the final stage is slow and rate-determining, thus
making it more likely that there will be considerable recycling of M and MH
leading to formation stereochemical and positional isomers.
D ! DH ! M ! MH ! S
# #
D M
The only useful commercial catalyst now used is nickel, available at a 1725%
level on a support and suspended in hardened edible oil or tallow. This preserves
the activity of the nickel in a form in which it can be safely and easily handled.
Catalyst can be recovered and reused but will be less active. Reaction is usually
effected at temperatures between 180 C and 200 C and at a pressure of about
0.3 MPa (3 bar). The catalyst is quickly poisoned by fatty acids, soaps, phospho-
lipids, oxidized acids, sulfur compounds, halogen compounds, carbon monoxide,
oxygen, and water. As a consequence, both the oil and the hydrogen should be
as pure as possible.
Catalysts are continually being improved. Hastert reports (185) that nickel load-
ing has fallen continuously from 0.25% (prior to 1960) to 0.1% (by 1970) and 0.05
0.1% (by 1990), and that 0.0250.05% is now normal. This is partly a consequence
of improved plant design, but catalyst surface area increased from 70 m2/g in 1970
to 180 m2/g in 1993, and there is increasing recognition of the importance of using
pure hydrogen and highly refined oil. Interesting developments now taking place
involve the use of precious metals (platinum and palladium), which although
more expensive, offer higher reaction rates at lower temperatures with formation
of less trans-isomer. In these ways, hydrogenation will probably continue as a
useful processing technique for many years to come.
8.4. Interesterification with a Chemical Catalyst

The production of fat spreads as an alternative to butter led to an increased demand
for solid fats. For the most part, this demand has been met by the use of partially
hydrogenated vegetable oils (Section 8.3), but concern about the health effects of
trans-unsaturated acids has raised interest in an alternative way of producing fats
with the required melting behavior. This can be achieved by interesterification of
blends of natural or fractionated fats. Products obtained in this way will probably
contain more saturated acids than their partially hydrogenated equivalents, but they
will have no trans-acids. This section is devoted to interesterification carried out
under the influence of a chemical catalyst (177, 186, 187). Similar reactions with
enzymes are discussed in the following section.
Interesterification is generally effected in 1015 ton batches at 8090 over 30

60 minutes at a cost not very different from that for partial hydrogenation. It does
not require expensive equipment nor use explosive gases. To get a product with the
desired properties, a soft oil is interesterified with a hard stock, which may be a
fractionated stearin, a lauric oil, or a fully hydrogenated seed oil. This last is a
scientifically acceptable choice but has the disadvantage that the word hydroge-
nated will have to appear on the label. The average customer does not appreciate
the difference between partially hydrogenated (with trans-acids) and fully hydroge-
nated (without unsaturated acids).
The catalyst normally employed is an alkali metal at a level of 0.10.2% or a
sodium alcoholate (usually sodium methoxide) at a 0.20.3% level. The true cata-
lyst is believed to be a diacylglycerol anion resulting from interaction of alkali and
triacylglycerol. The catalyst is easily destroyed by acid, water, or peroxides, so the
feedstock oil should be free of these impurities (188).
Natural oils and fractionated oils usually have their acyl chains organized in a
nonrandom manner, but they become randomized after interesterification with a
chemical catalyst. There is no change in fatty acid composition, only in triacylgly-
cerol composition, but this leads to a modification of the physical properties. More
selective interesterification can be achieved with enzymic catalysts (Section 8.5).
The following are typical applications of interesterification:
Lard, with an unusually high level of palmitic acid in the b-position, crystal-
lizes naturally in the b form. When randomized, the content of 2-palmito-
glycerol esters is reduced from around 64% to 24% and the interesterified
product crystallizes in the b0 form with consequent improvement in shortening
properties.
The crystal structures of margarines based on sunflower or canola oil (rape-
seed) along with hydrogenated oil are stabilized in the b0 form by interester-
ification leading to randomization of the glycerol esters.
Solid fats with about 60% of essential fatty acids can be obtained by (reduced
temperature) interesterification of sunflower oil and about 5% of hard fat.
Margarine made, for example, by interesterification of palm stearin and
sunflower oil (1:1), contains no hydrogenated fat and therefore no trans-acids.
Chemical interesterification is used in the production of caprenin, salatrim,
and olestra.
8.5. Interesterification with an Enzymic Catalyst

Interesterification can also be catalysed by enzymes, many of which show useful
specificities. The 1,3-specific lipases, such as those derived from Aspergillus niger,
Mucor javanicus, M. miehei, Rhizopus arrhizus, R. delemar, and R. niveus, are par-
ticularly useful for interesterification. They are used to effect acyl exchange at the
sn-1 and 3 positions while leaving acyl groups at the sn-2 position unchanged.
250 VEGETABLE OILS
Many interesting changes of this type have been affected on a bench scale, but as
yet only a few have been commercialized and then only for products of high value
(186 190).
Unilever developed a method for upgrading palm mid-fraction (PMF) as a cocoa
butter equivalent. The PMF is too rich in palmitic acid and has too little stearic acid,
but this deficiency can be repaired by enzyme-catalysed acidolysis with stearic acid.
Reaction is confined to the exchange of palmitic acid by stearic acid at the sn-1 and
3 positions with no movement of oleic acid from the sn-2 position. A similar pro-
duct is produced enzymatically by acidolysis of high-oleic sunflower oil (rich in
triolein) and stearic acid.
POP St ! POSt StOSt

OOO St ! StOO StOSt
Chemical interesterification would lead to randomization of all of the acyl chains,

and the products would have different melting behavior from that required by a
cocoa butter equivalent (188).
Another product manufactured by Loders-Croklaan (Unilever) and named Beta-
pol consists mainly of triacylglycerols of the type UPU. This is used as a constituent
of infant formulas (191). Human milkfat is unusual in that it contains a significant
proportion of its palmitic acid in the sn-2 position. Although this is true also for lard
(pig fat), it is not a feature of vegetable oils. Betapol is made from tripalmitin and
oleic acid using the lipase from Mucor miehei to promote 1,3-acyl exchange. These
materials are expensive, and in practice, fractionated palm oil, rich in tripalmitin, is
reacted with acids from high-oleic sunflower or safflower or from olive oil.
Bohenin (BOB) is the name given to glycerol 1,3-behenate 2-oleate, which
inhibits fat bloom when added to chocolate. It is produced in Japan by enzymic
interesterification of triolein and behenic (22:0) acid or ester in the presence of a
1,3 stereospecific lipase.
Diacylglycerols are being produced and used in growing quantities because of
their beneficial effects in the management of obesity. Cooking oil containing at
least 80% of diacylglycerols (mainly the 1,3- isomer) has been marketed in Japan
since 1999 by the Kao company and thereafter in other countries, including the
United States by ADM.
There are many reports showing how, with an appropriate enzyme (Mucor mie-
hei and Candida antarctica are frequently used), long-chain polyunsaturated fatty
acids such as eicosapentaenoic acid and/or docosahexaenoic acid can be introduced
into vegetable oils or synthetic glycerides to give products with enhanced nutri-
tional value. In a similar way, C8 and C10 acyl chains can be introduced into vege-
table oils or fish oils with a consequent change in nutritional properties and energy
values. The products are triacylglycerols with either one long and two short chains
(LS2) or two long and one short chain (L2S). This reaction can be used to produce
triacylglycerols with easily metabolizable short- and medium-chain acids at the
sn-1 and 3 positions and an essential fatty acid at the crucial sn-2 position. This
topic is the subject of some recent reviews (190, 192, 193).
BIOLOGICAL METHODS OF LIPID MODIFICATION 251
9. BIOLOGICAL METHODS OF LIPID MODIFICATION
9.1. Introduction
New sources of oils and fats develop from plants in three different ways. One pos-
sibility is to take a wild plant that produces oil with an interesting fatty acid and/or
triacylglycerol profile and to make it suitable for commercial growing and harvest-
ing. This is generally a slow process requiring many years. Traits developed over an
evolutionary time scale to maintain the plant in the wild are not always appropriate
in domesticated plants and have to be bred out. This approach is being pursued,
particularly in North America and in Europe, for a number of species identified
as promising, and they are at differing stages of development (see Section 9.2).
A second approach, when a diverse gene pool is available, is to interbreed spe-
cies with appropriate traits by standard seed-breeding processes. This has been done
very effectively with species of brassica to yield the modern oilseed rape (canola).
If necessary, the gene pool can be extended by mutation resulting from chemical
treatment or from irradiation. This may produce novel varieties with interesting
traits and is the basis of the low-linolenic lines from linseed, and other examples
are described in Section 9.3.
Finally, genes required for particular aspects of fatty acid and triacylglycerol
biosynthesis can be identified in appropriate sources, cloned, and transferred to
other plants. Rapeseed has proved to be particularly flexible in this respect, and
its fatty acid composition has been modified in several ways, some of which
have now reached or are very close to commercial application (Section 9.4). Genetic
modification procedures are also applied to soybean and other oilseed crops.
The commercial introduction of a new lipid source is not a trivial matter. Unless
the oil has some specific and novel property (like oils containing g-linolenic acid,
for example), it will have to compete with existing oils available in bulk at com-
modity prices. This exerts a number of constraints.
The new crop should be easily cultivated, harvested, processed, and marketed.
Additional costs may result from the need to have separate and distinct
harvesting, storage, processing, and marketing facilities.
The new crop must quickly become available in good and reliable quantities at
acceptable prices.
The demand for and interest in some new crops may come more from the
oleochemical industry than from the food industry, but traditionally the
oleochemical industry has used lower grade and cheaper oils than the food
industry.
Because the supply of new oil must start small and grow with demand, it is
useful to find some low-volume, high-value products that will support the crop
through its early years of development until the supply is adequate to be used
for high-volume, low-value products.
The demand must be market-led (at least after the first few years). At present,
there is an interest in new crops that produce oils with high levels of a single
252 VEGETABLE OILS
acid such as lauric, oleic, petroselinic, erucic, or acids with hydroxy or epoxy
groups.
Although agronomists must help to produce oils meeting these requirement,
chemists and technologists must assist in the substitution of existing oils by
new oils and in the development of new uses for new oils.
In addition, regulatory requirements will have to be met when the seed and/or
its products are novel. This topic has been reviewed (194, 195).
9.2. Domestication of Wild Crops

Serious attempts are being made in Europe and North America to develop a range
of wild oilseed crops, generally with a high content of one particular acid. The fol-
lowing account shows the variety of species being examined. Reference has already
been made to some of these in Sections 4 and 6.
Cuphea oils are interesting because they come from annual plants producing gly-
cerol esters based mainly on capric (10:0) or lauric acid (12:0) or occasionally on
caprylic (8:0) or myristic acid (14:0). Cuphea plants exhibit several wild plant char-
acteristics that need to be bred out. These include dormancy, nonuniform germina-
tion, indeterminate flowering, seed maturation over a broad time period (six weeks),
extreme dehiscence (pod shattering), and the presence of viscid hairs on stem,
leaves, flowers, and fruit. Several species are being studied in attempts to make
them commercial (30).
Oleic oils. Oleic acid is an important source material for the oleochemical industry,
and as the most common monoene acid, it has a good rating on dietary grounds. For
many oleochemical purposes, the presence of some saturated acid (palmitic, stearic)
is not significant but levels of linoleic and linolenic acid should be as low as pos-
sible because they promote undesirable oxidation. All oils contain oleic acid, and
frequently, it is the dominant component. For example, the oils of rapeseed (56%),
macadamia (56% along with a further 22% of 16:1), almond (61%), high-oleic
safflower (74%), olive (78%), and high-oleic sunflower (82%) contain the levels
of oleic acid indicated. Of these, almond and rape also contain 2530% of linoleic
acid. Euphorbia lathyris (caper spurge) is a Mediterranean annual containing about
50% of oil in its seed and 8090% of oleic acid in the oil. This would make it an
excellent source of oleic acid. At present, it suffers from a number of deficiencies,
especially seed shattering and the presence of a cocarcinogenic milky sap. But it is
hoped that these problems will be overcome through plant breeding. Other good
sources of oleic acidboth existing and potentialare discussed in Sections 9.3
and 9.4.
Petroselinic acid oils. Petroselinic acid is the 6 isomer of the more common oleic
acid (9). Although they share many properties, these acids display an interesting
difference in melting point. Petroselinic acid and its glycerol ester melt at 33 C and
28 C, respectively. The corresponding figures for oleic acid and its glycerol ester
are lower at 12 and 16 (two forms) and 5 . Attempts are being made to develop
a better source of petroselinic acid by improved cultivation of coriander ( 80%
petroselinic acid) or by transferring the appropriate genes from this plant to rape
(Section 9.4) (196).
Oils containing C18 polyene acids: Calendula officinalis seed oil. Calendula oil
(from marigold) is of interest because it contains about 58% of calendic acid
(8t10t12c-18:3). This unusual acid is an isomer of a-eleostearic (9c11t13t-18:3)
present in tung oil, and calendula oil should also be a good drying oil. The presence
of linoleic acid (30%) will add to the unsaturated nature of this oil (131133).
Camelina sativa seed oil. This plant is also known as gold of pleasure or false
flax. In addition to its interesting fatty acid composition, it attracts attention because
it grows well with lower inputs of fertilizers and pesticides than more traditional
crops like rape and linseed. The plant can also be grown on poorer soils and shows
better gross margins than the other two plants after allowing for direct costs and
subsidy payments. The seed yield is in the range 1.53.0 t/ha and the oil content
between 36% and 47%. The oil has an unusual fatty acid composition. It contains
significant levels of linolenic acid (31 41%) and of C20 and C22 acids, especially
20:1 (1523%). Despite its high level of unsaturation, it shows reasonable oxidative
stability. Attempts are being made to optimize the agronomy. Its use in paints,
varnishes, and inks, in cosmetics, and even as a food oil is being examined and
developed (108110).
Oils containing erucic and other long-chain monounsaturated acids. In 1994, it was
reported (22) that the demand for erucic acid-based oleochemicals was about 20 kt
of compounds derived from 5560 kt of high-erucic oil. These oleochemicals
include materials obtained from erucic acid (22:1) or from behenic acid (22:0) and
brassylic acid (the C13 dibasic acid resulting on ozonolysis). The demand is mainly
for erucamide (7 kt), other erucic acid nitrogen compounds (2.7 kt), erucic esters
(1.82.3 kt), erucyl alcohol (4.5 kt), behenyl alcohol (2.7 kt), and glycerol tribehe-
nate (1.11.4 kt). High-erucic oils are reported to have a growth rate of about 6%.
The traditional source of erucic acid was rapeseed oil before this acid was bred
out of that oil because of its reported adverse health effects. Most rapeseed oil now
contains less than 2% of erucic acid. The two major sources of erucic acid are high-
erucic rapeseed oil (HEAR) containing about 50% of erucic acid and crambe oil
with 5560% of erucic acid. As will be reported later (Section 9.4), attempts to
produce a still higher erucic rapeseed oil are being made by genetic engineering.
Crambe oil (from Crambe abyssinica) is grown most extensively in North Dakota
and to a lesser extent in Holland.
Meadowfoam oil from Limnanthes alba seed oil is unusual in that over 95% of
its component acids are C20 or C22 and include 520:1 (6367%), 522:1 (2 4%),
1322:1 (1618%), and 5,1322:2 (59%). It is being grown in the United States,
and its potential uses thoroughly examined. The crop yields 10001500 kg of seed
per hectare and contains 25% oil. Potential uses of this oil include cosmetic appli-
cations, production of dimer acid, as a lubricant, and via a wide range of novel deri-
vatives based on reaction at the 5 double bond (134138).
Simmondsia chinensis seed oil. Jojoba oil is another source of C20 and C22
compounds that has already been developed as a marketable product but in limited
supply (195). It is produced by a drought-resistant plant that withstands desert heat.
254 VEGETABLE OILS
It takes 57 years to first harvest, 1017 years to full yield, and has a life span of
around 100 years. It is being grown in the Southwestern United States and Mexico
mainly, but also in Latin America, Israel, South Africa, and Australia. Yields are
reported to be about 2.5 ton of oil/hectare.
Jojoba oil is not a triacylglycerol but a mixture of wax esters based mainly on
20:1 and 22:1 acids and alcohols. It contains C40, C42, and C44 esters with two iso-
lated double bonds (one in the acyl chain and one in the alkyl chain). The oil serves
as replacement for sperm whale oil, which is proscribed in most countries because
the sperm whale is an endangered species. At present, jojoba oil is a high-priced oil
used mainly in cosmetics, but it has excellent lubricating properties and could be
used extensively for this purpose if available in sufficient quantity at an appropriate
price.
The oil is fairly pure as extracted, has a light color, and because the double bonds
are well separated, it is resistant to oxidation. The oil can be chemically modified by
reaction of the double bonds (hydrogenation, stereomutation, epoxidation, sulfo-
chlorination) (197).
Honesty seed oil (Lunaria biennis) is characterized by its high levels of monoene
acids, including 18:1 (23%), 22:1 (46%), and 24:1 (23%). It is being developed as a
commercial crop for nutritional research based on its significant level of nervonic
acid (24:1) (120).
Oils containing hydroxy acids: The only oil of significance containing a hydroxy
acid is castor oil, but among the new crops being seriously developed are two that
contain hydroxy acids. Lesquerella oils have some resemblance to castor oil, but
Dimorphotheca pluvialis seed oil contains a different kind of hydroxy acid (see
Section 6).
Oils containing epoxy acids: Several natural epoxy acids are known, but vernolic
acid (12,13-epoxyoleic) is the most common and occurs at high levels in several
seed oils. Of these, serious attempts are now being made to develop Vernonia
galamensis (7378% vernolic acid) and Euphorbia lagascae (5762% vernolic
acid) as commercial crops (198). Several potential uses of this acid and the seed
oils in which it occurs are being explored.
9.3. Oils Modified by Conventional Seed Breeding

Seed breeding of industrial crops is a continuous activity. Much of this is concerned
with agronomical factors and is not of concern here. More significant for the present
purpose are those developments leading to new or improved seed oils. Many of the
changes are small and incremental and only become apparent after many years.
Others are more dramatic. Some examples that are already well established will
be discussed.
Rapeseed oil: Low-erucic rapeseed oil is now the third largest source of oil after
soybean and palm. The seed contains over 40% of oil and this represents about
80% of the seeds commercial value. Seed breeders have developed seeds which
produce oil low in erucic acid (<2%) and meal low in the undesirable sulfur-
containing glucosinolates (i.e. double low varieties). This low-erucic oil (LEAR)
finds many food uses and also some non-food uses (biodiesel, lubricants). The crude
oil is rich in phospholipids ( 3.5%) though these are reduced to 10300 ppm (phos-
phorus) after refining and are themselves a useful by-product (Section 3). The plant
grows in cooler agricultural regions including China, Northern Europe, and Canada
as well as in the Indian sub-continent. In common with other Brassica species rape-
seed oil contains brassicasterol at much higher levels ( 600 ppm) than is observed
in other seed oils. Low-erucic rapeseed oil has a very low level of saturated acids
and a high level of oleic acid: palmitic 4%, stearic 2%, oleic 56%, linoleic 26%,
linolenic 10%, and others 2% (194, 199).
Linseed oil: Linseed oil is well known as one of the most unsaturated vegetable oils
with a high level of linolenic acid ( 50%). As a consequence of this it oxidizes and
polymerizes very readily and is used in paints, varnishes, inks, linoleum, and as a
sealant for concrete. Using chemical mutation, plant breeders in Australia (33)
developed a variety of linseed with a low level of linolenic acid ( 2%) and a
high level of linoleic acid. This is called linola and is a linoleic-rich oil like sun-
flower (Table 4). The oil has GRAS status in USA.
High-oleic sunflower and safflower oils: By taking advantage of the wide range of
natural sunflower and safflower varieties seed breeders have developed lines which,
in place of the normal high levels of linoleic acid, have high levels of oleic acid
(Table 2a). These are commercially available as Sunola ( 85% oleic acid) and
Saffola ( 75% oleic acid) (49). They are used in Good-Fry (Section 2) and as
an alternative to triolein in some enzymic processes (Section 8.5). A third type
of sunflower oil (Nu-Sun) with an intermediate level of oleic acid (65%) and
reduced levels of saturated acids is now available.
The oil palm is already the most productive source of vegetable oil at an average
level of over 3t/ha/yr. Seed breeding through the last 25 years has led to palms
which in the best environments can produce 10 tons per hectare. According to
Jalani et al (38, 39) further objectives being pursued include the following.
High palm plants present harvesting problems. With shorter plants, there is
easier harvesting and a longer planting cycle because the trees do not need to
be replanted so often. Plants that grow only 1525 cm/yr are now available in
place of the usual 4575 cm/yr.
Oils with higher iodine value (normally 53, raised to 63) contain less palmitic
acid and more oleic acid. When fractionated, they produce more of the
valuable olein fraction (Section 5).
Kernels are normally about 6% of the fruit, but palms with 12% kernel have
now been developed. This is advantageous because palm-kernel oil commands
a higher price than palm oil.
9.4. Oils Produced Through Genetic Engineering

Genetically modified seed oils: Recent years have witnessed great strides in the
understanding and application of genetic engineering. This has been applied to
256 VEGETABLE OILS
oilseed plants to produce mainly agronomic benefits such as resistance to herbicides

and to pests, shorter times between sowing and harvest, increased yield, and so on.
But these techniques have also been applied to changing fatty acid composition and
hence triacylglycerol composition. A few products are commercially available at
the present time, others are at the stage of field trials, and yet more have been
obtained in the laboratory. The number of available products should increase
rapidly in the next few years (199 201). The number of known fatty acids exceeds
1000, although only a small number of theseperhaps around 25are of common
concern. Most vegetable oils contain only palmitic, oleic, and linoleic acids as
major components in varying proportions. These are accompanied by stearic and
linolenic acid in some seed oils. Others that become major components in selected
seed oils include 8:0, 10:0, 12:0, 14:0, 18:1(6c), 20:1, 22:1, 24:1, ricinoleic, verno-
lic, and GLA. Some of these are related biosynthetically as already discussed
(Section 2).
This range shows that plant lipids as a whole can produce a wide variety of fatty
acids, sometimes at very high levels. It follows that the enzymes necessary to pro-
duce these less common acids are available somewhere within the plant kingdom. It
is then possible to identify these, clone them, and introduce them into species that
are already cultivated on a large scale such as rapeseed, soybean, maize (corn), sun-
flower, and linseed among others (143, 202204).
The rape plant seems to lend itself to genetic manipulation, and the first geneti-
cally modified oilseed with changed fatty acid profile was canola oil containing
lauric acid. This was developed by Calgene, and the crop is grown in the United
States, although successful field trials have been conducted elsewhere. To obtain
this new oil, Calgene scientists isolated the thioesterase, which produces lauric
acid in the Californian Bay tree and transferred it to the rape plant. With this, de
novo synthesis stops mainly at the C12 level rather than the more usual C16 acid.
When introduced into rapeseed, the oil contained more than 50% of lauric acid,
although this was somewhat reduced in the commercial crop. To go beyond this
TABLE 11. Fatty Acid Composition of Commodity Canola Oil and Some Genetically
Modified Oils Based on it.
Oil source 12:0 14:0 16:0 18:0 18:1 18:2 18:3 Other
Rape 2 2 13 12 9 62
Canola 4 2 62 20 9 3
High in
16:0 10 1 51 19 13 6
16:0/18:0 9 10 57 14 4 6
16:0 29 2 31 22 13 3
18:0 4 34 22 18 18 4
12:0 40 4 3 1 29 12 8 3
14:0 40 3 1 29 10 7 10
18:1 4 1 84 5 3 3
18:2 4 2 33 49 7 5
20:1 (7%) and 22:1 (54%).

Based on Huang and Ziboh (84).
TABLE 12. Fatty Acid Composition of Commodity Soybean Oil and Some Genetically
Soybean oil 16:0 18:0 18:1 18:2 18:3 other

Commodity 10 4 23 52 8 2
Saturated (L) 4 3 23 60 10
Linoleic (H) 11 6 28 52 3
Palmitic (H) 24 4 15 44 11 2
Stearic (H) 8 25 17 39 8 3
Saturated (H) 22 18 9 38 11 2
Sat/len (L) 4 3 28 61 3 1
P/len (H) 19 4 23 48 3 3
H high, L low.
level, it is necessary to introduce a further gene (lysophosphatidic acid acyl trans-

ferase, LPAT), which will promote the acylation of the sn-2 position with lauric
acid. There is some concern about the economic viability of this project.
Other oils at various stages of development include the following (200):
Rapeseed oils still higher in lauric acid, high in erucic, palmitic, oleic, or
linoleic acid, or containing C8 and C10 acids, myristic, stearic, petroselinic,
ricinoleic, vernolic, or g-linolenic acid, and wax esters in place of the normal
triacylglycerols (Table 11).
Soybean oils with lower saturated acids, lower linolenic acid, and higher
stearic acid as well as seeds producing meal of enhanced nutritional value
(Table 12).
Sunflower oil with high palmitic, stearic, oleic, or linoleic acid (Table 13).
Corn oil with high oleic acid.
The level of linolenic acid is being reduced because its oxidation leads to unde-
sirable flavors. Saturated acids are being increased to produce oils that can be used
to make spreads without partial hydrogenation (see Section 6).
TABLE 13. Fatty Acid Composition of Commodity Sunflower Oil and Some Genetically
Sunflower oil 16:0 16:1 18:0 18:1 18:2 Other

Commodity 7 5 28 59 1
High oleic 3 2 92 2 1
High linoleic 8 2 13 76 1
High stearic/oleic 5 11 79 2 3
High palmitic/oleic 25 6 3 60 4 2
High palmitic/linoleic 27 4 3 17 47 2

258 VEGETABLE OILS
10. PRODUCTION AND TRADE STATISTICS
Table 14 contains production and export data for 13 vegetable oils. The figures are
given as annual average values for the four 5-year periods from 19911995 to
20052010 and thus cover a 20-year period. Readers who prefer information for
individual years can take the four columns of figures to be close to values reported
or expected for the midyear in each quinquennium. viz. 1993, 1998, 2003, and
2008. These figures are taken from The Revized Oil World 2020Supply, Demand
and Prices produced by ISTA Mielke GmbH of Hamburg in 2002 (205). This
company has been producing and interpreting data for oilseeds, oils and fats, and
seed meals since 1958. The 2002 publication contains much more relevant Informa-
tion. Attention is drawn to the following points.
During the 15 years (19932008) covered in Table 14, production of oils and
fats is expected to rise 69% from 86.8 to 146.7 million tons and exports are
expected to double, rising from 25.3 to 50.8 million tons. Comparisons
between production and exports of oils and fats are sometimes complicated
by the fact that for oilseeds, there is trade in oilseeds as well as in the
extracted oils. This does not apply to palm oil traded only as oil.
Three oilseeds and four oils dominate production and export and have become
more dominant with the passage of time. These are soybean oil (produced
mainly in the United States, Brazil, Argentina, and China), palm oil (Malaysia
and Indonesia), rape/canola oil (China, EU-15, India, and Canada), and
TABLE 14. Annual Average Production (million tons) of Oils and Fats for the 5-year
Periods 19911995 to 20062010.
Production Exports
Oil 9195 9600 0105 0610 9195 9600 0105 0610

World total 86.82 105.06 126.47 146.73 25.29 32.93 41.88 50.82
Soybean 17.90 23.14 29.56 33.60 4.12 6.81 8.88 10.63
Cotton 3.94 4.00 4.60 5.35 0.25 0.22 0.26 0.32
Peanut 4.09 4.56 5.45 5.72 0.27 0.25 0.29 0.31
Sunflower 7.96 9.11 9.83 12.43 2.16 2.93 2.83 4.17
Rape/canola 9.66 12.64 15.34 17.72 1.59 1.90 1.85 2.48
Sesame 0.62 0.70 0.75 0.85 0.02 0.02 0.03 0.03
Corn 1.67 1.91 2.18 2.49 0.49 0.71 0.86 1.04
Olive 1.96 2.47 2.58 2.75 0.38 0.47 0.57 0.62
Palm 13.34 18.72 25.24 31.43 9.55 12.76 18.55 22.66
Palmkernel 1.72 2.34 3.11 3.84 0.84 1.11 1.49 1.81
Coconut 3.03 3.01 3.33 3.70 1.50 1.64 1.91 2.00
Linseed 0.64 0.70 0.69 0.81 0.14 0.13 0.12 0.16
Castor 0.46 0.46 0.54 0.62 0.20 0.25 0.28 0.32
Anmal fats
19.80 21.30 23.26 25.42 3.77 3.69 3.94 4.26
Butter, lard, tallow, fish oil.

Adapted from Mielke (205).
REFERENCES 259
TABLE 15. Production and Export of Three Major Oilseeds and Four Major Oils. Figures
are Percent of Total Production and Exports Based on Average Annual Levels (Million
Tons) for the 5 year Period 20012005.
Oilseeds Oils and Fats
Production% Exports% Production% Exports%
Soybean 55.2 79.5 23.4 21.2

Rape/canola 12.6 10.4 12.1 4.4
Sunflower 8.4 3.8 7.8 6.8
Palm 19.9 44.3
Adapted from Mielke (205).
sunflower oil (ex-USSR, EU-15, and Argentina). These are produced mainly
in the countries listed (Table 14 and 15).
Cottonseed and groundnut (peanut) oils and some of the minor oils are used
almost entirely in the country of origin and exports amount to very little.
11. CONCLUSION
The vegetable oils are important materials, significant for agriculture, the refining
and processing industries, and for the food industry. Their production and use is
based on a wide range of supporting sciences (physics, chemistry, biochemistry,
agriculture, seed breeding, molecular biology, engineering, food science, nutrition,
and medicine among others). At present, we are particularly dependent on five oils
from four sources: soybean, the oil palm producing two different oils, rapeseed
(canola), and sunflower. These, and the minor vegetable oils, are produced and
used at increasing levels each year and are essential components of the human
diet. Although some individuals consume too much, others have too little and
demand is expected to grow for many years yet. As the land available to grow these
materials is limited, it is essential to increase yields. The vegetable oils do not
always have ideal physical and nutritional properties, so methods of modifying
the oils have been developed. For the most part, these have been technological in
the past, but increasingly in the future, they are expected to be biological. In the
second half of the twentieth century, soybean oil and palm oil have risen to promi-
nence. Will there be new major oils developed in the next half century? This seems
unlikely, but we may yet be surprised.
REFERENCES
1. F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell Publishing, Oxford,

U.K., 2002.
2. J. L. Harwood, Biochim. Biophys. Acta, 1301, 756 (1996).
260 VEGETABLE OILS
3. J. L. Harwood, in F. D. Gunstone, ed., Lipid Synthesis and Manufacture, Sheffield

4. K. L. Parkin, in C. C. Akoh and D. B. Min, eds., Food LipidsChemistry, Nutrition, and
Biochemistry, 2nd ed., Marcel Dekker, New York, 2002, pp. 909965.
5. P. Broun, J. Shanklin, E. Whittle, and C. Somerville, Science, 282, 13151317 (1998).
6. E. Heinz, Eur. J. Lipid Sci. Technol., 104, 12 (2002).
7. B. F. Szuhaj, Lecithins: Sources, Manufacture, and Uses, AOCS Press, Champaign,
Illinois, 1989.
8. M. Schneider, in F. D. Gunstone and F. B. Padley eds., Lipid Technologies and
9. F. D. Gunstone, in F. D. Gunstone, ed., Structured and Modified Lipids, Dekker,
New York, 2001, pp. 241250.
10. J. P. Clark, Lipid Technol., 8, 111114 (1996).
11. F. D. Gunstone, in W. Hamm and R. J. Hamilton, eds., Edible Oil Processing, Sheffield
12. Y. B. Che Man and C. P. Tan, in F. D. Gunstone, ed., Lipids for Functional Foods and
Nutraceuticals, Oily Press, Bridgwater, U.K., 2552 (2003).
13. C. K. Ooi, Y. M. Choo, S. C. Yap, Y. Basiron, and A. S. H. Ong, J. Amer. Oil Chem. Soc.,
71, 423 426 (1994).
14. B. S. Baharin, K. A. Rahman, M. I. A. Karim, T. Oyaizu, K. Tanaka, Y. Tanaka, and
S. Takagi, J. Amer. Oil Chem. Soc., 75, 399404 (1998).
15. N. V. Yaniishlieva, K. Aitzetmuller, and V. G. Raneva, Fett/Lipid, 100, 444 462
(1998).
16. B. K. Larsson, A. T. Eriksson, and M. Cervenka, J. Amer. Oil Chem. Soc., 64, 365370
(1987).
17. F. D. Gunstone, Inform, 13, 6568 (2002).
18. H. Okuyama, T. Kobayashi, and S. Watanabe, Prog. Lipid Res., 35, 409457 (1997).
19. H. Okuyama, Lipid Technol. Newslet., 6, 128132 (2000).
20. I. M. Stewart and R. E. Timms, in K. K. Rajah, ed., Fats in Food Technology, Sheffield
21. R. E. Timms, Confectionery Fats Handbook, The Oily Press, Bridgwater, U.K., 2003.
22. C. Leonard, Lipid Technol., 6, 7983 (1994).
23. C. E. Temple-Heald, Lipid Technol. Newslett., 6, 5963 (2000).
24. C. E. Temple-Heald, in F. D. Gunstone, ed. Rapeseed and Canola OilProduction,
Processing, Properties and Uses, Blackwell Publishing, Oxford, U.K., 2004, pp. 111130.
25. R. Przbylski and T. Mag, in F. D. Gunstone, ed., Vegetable Oils in Food Technology,
Blackwell Publishing, Oxford, U.K., 2002, pp. 98127.
26. H-J. Caupin, in F. D. Gunstone and F. B. Padley, eds, Lipid Technologies and App-
lications, Marcel Dekker, New York, 1997, pp. 787795.
27. C. Crews, in M. Jee, ed., Oils and Fats Authentication, Blackwell Publishing, Oxford,
U.K., 2002, pp. 6694.
28. Anon., Inform, 3, 10801090 (1992).
29. T. P. Pantzaris and Y. Basiron, in F. D. Gunstone, ed., Vegetable Oils in Food Technology,
Blackwell Publishing, Oxford, U.K., 2002, pp. 157202.
REFERENCES 261
30. T. A. Isbell, Lipid Technol., 14, 7780 (2002).

31. R. A. Moreau, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
Publishing, Oxford, U.K., 2002, pp. 278296.
32. R. D. OBrien, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
33. B. E. Prentice and M. D. Hildebrand, Lipid Technol., 3, 8389 (1991).
34. S. C. Cunnane and L. U. Thompson, eds., Flaxseed in Human Nutrition, AOCS Press,
35. A. G. Green, and J. C. P. Dribnenki, Lipid Technol., 6, 2933 (1994).
36. Anon., Lipid Technol. Newsletter, 1, 8081 (1995).
37. Advisory Committee on Novel Foods and Processes (UK), Annual Report, 1997,
p. 8.
38. Anon. Lipid Technol. Newslett., 4, 1315 (1998).
39. D. Boskou, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
40. B. S. Jalani and N. Rajanaidu, Lipid Technol., 12, 58 (2000).
41. B. S. Jalani, S. C. Cheah, N. Rajanaidu, and A. Darus, J. Amer. Oil Chem. Soc., 74, 1451
1455 (1997).
42. S. W. Lin, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
43. Sanders, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell Publish-
ing, Oxford, U.K., 2002, pp. 231243.
44. F. D. Gunstone, ed., Rapeseed and Canola OilProduction, Processing, Properties and
Uses, Blackwell Publishing, Oxford, U.K., 2004.
45. A. G. Gopala Krishna, Lipid Technol., 14, 8084 (2002).
46. B. Sayre and R. Saunders. Lipid Technol., 2, 7276 (1990).
47. Anon., J. Amer. Oil Chem. Soc., 66, 615618 (1989).
48. H. P. Gupta, J. Amer. Oil Chem. Soc., 66, 620623 (1989).
49. S. Bhattacharyya and D. K. Bhattacharyya, J. Amer. Oil Chem. Soc., 66, 14691471
(1989).
50. Anon., Inform, 8, 968 (1997).
51. S. P. Kochar, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
52. J. R. Smith, Safflower, AOCS Press, Champaign, Illinois, 1996.
53. S. P. Kochar, Lipid Technol. Newslett., 6, 3539 (2000).
54. T. Wang, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
55. F. D. Gunstone, J. Oleo Sci., 50, 269279 (2001).
56. M. K. Gupta, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, Blackwell
57. A. Hase and S. Pajakkala, Lipid Technol., 6, 110114 (1994).
58. Anon., Lipid Technol., 10, 98 (1998).
59. G. Webster, Lipid Technol. Newslett., 4, 38 43 (1998).
262 VEGETABLE OILS
60. F. D. Gunstone, Fett/Lipid, 101, 124131 (1999).

61. F. D. Gunstone, in F. D. Gunstone, ed., Structured and Modified Lipids, Dekker,
New York, 2001, pp. 185208.
62. D. Firestone, Official Methods and Recommended Practices of the AOCS, Physical and
Chemical Characteristics of Oils, Fats and Waxes, Champaign, Illinois, 1997.
63. E. Ucciani, Nouveau Dictionnaire des Huiles Vegetales, Paris, France, 1995.
64. F. B. Padley, in F. D. Gunstone, J. L. Harwood, and F. B. Padley, eds., The Lipid
Handbook, 2nd ed., Chapman and Hall, London, 1994, p 53.
65. K. Aitzetmuller and M. Ihrig, Fat Sci. Technol., 90, 464470 (1988).
66. M. S. Prats-Moya, N Grane-Teruel, V. Berenguer-Navarro, and M. l. Martin-Carattalla,
67. C. K. Lyon and R Becker, J. Amer. Oil Chem. Soc., 64, 233236 (1987).
68. J. T. Budin, W. M. Breene, and D. H. Putnam, J. Amer. Oil Chem. Soc., 73, 475 481
(1996).
69. H. Sun, D. Wiesenborn, K. Tostenson, J. Gillespie, and P. Rayas-Duarte, J. Amer. Oil
Chem. Soc., 74, 413418 (1997).
70. F. Jahaniaval, Y. Kakuda, and M. F. Marcone. J. Amer. Oil Chem. Soc., 77, 847852 (2000).
71. B. S. Kamel and Y Kakuda, J. Amer. Oil Chem. Soc., 69, 492494 (1992).
72. Advisory Committee on Novel Foods and Processes (UK), Annual Report, 1993.
73. M. Zlatanov and I. Janakieva, Fett/Lipid, 100, 312315 (1998).
74. M. J. Weerman and L. Neeman, J. Amer. Oil Chem. Soc., 64, 229232 (1987).
75. H. E. Swisher, J. Amer. Oil Chem. Soc., 65, 17041706 (1988).
76. Y. F. Lozano, C. D. Mayer, C. Bannon, and E. M. Gaydouu J. Amer. Oil Chem. Soc., 70,
561565 (1993).
77. L. Eyres, N. Sherpa, and G. Hendricks, Lipid Technol., 13, 8488 (2001).
78. J. Birkbeck, Food New Zealand, April/May, 40 42 (2002).
79. J. L. R. Pritchard in J. B. Rossell and J. L. R. Pritchard, eds., Analysis of Oilseeds, Fats,
and Fatty Foods, Elsevier, London, 1991, pp. 7797.
80. P. M. Clough, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker,
New York, 2001, pp. 75117.
81. P. M. Clough, Lipid Technol., 13, 912 (2001).
82. Y-S. Huang and D. E. Mills, g-Linolenic Acid, AOCS Press, Champaign, Illinois,
1996.
83. S. P. J. N. Senanayake and F. Shahidi, J. Amer. Oil Chem. Soc., 77, 5561 (2000).
84. Y-S. Huang and V. A. Ziboh, g-Linolenic Acid: Recent Advances in Biotechnology and
Clinical Applications, AOCS Press, Champaign, Illinois, 2001.
85. Y-S. Ju and T-C. Chen, J. Amer. Oil Chem. Soc., 79, 2932 (2002).
86. I. M. Stewart and R. E. Timms, in K. K Rajah, ed., Fats in Food Technology, Sheffield
87. I. Campbell (2002). Available: http://www.britanniafood.com/invite.
88. J. A. Vasceoncellos, W. P. Bemis, J. W. Berry, and C. W. Weber, in E. H. Pryde, L. H.
Princen, and K. D. Mukherjee, eds., New Sources of Fats and Oils, AOCS Press,
REFERENCES 263

Handbook, 2nd ed., Chapman and Hall, London, 1994, pp. 5455.
90. F. D. Gunstone, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker,
New York, 2001, p. 191.
91. T. G. Toschi, M. F. Caboni, G. Pennazzi, G. Lercker, and P. Capella, J. Amer. Oil Chem.
Soc., 70, 10171020, 1993.
92. B. S. Kamel and Y. Kakuda, J. Amer. Oil Chem. Soc., 69, 492494 (1992).
93. F. Comes, M. Farines, A. Aumelas, and J Soulier, J. Amer. Oil Chem. Soc., 69, 1224
1227 (1992).
94. Advisory Committee on Novel Foods and Processes (UK), Annual Report (1993).
95. Anon, Lipid Technol., 5, 53 (1993).
96. B. S. J. Jeffrey and F. B, Padley, J. Amer. Oil Chem. Soc., 68, 123127 (1991).
97. R. Ayerza, J. Amer. Oil Chem. Soc., 72, 10791081 (1995).
New York, 2001, p. 192.
99. K. J. Lessman and W. P. Andersom, in E. H. Pryde, L. H. Princen, and K. D. Mukherjee,
eds., New Sources of Fats and Oils, AOCS Press, Champaign, Illinois, 1981, pp. 223
246.
100. M. A. Reuber, L. A. Johnson, and L. R. Watkins, J. Amer. Oil Chem. Soc., 78, 661664
(2001).
101. G. Steinke and K. D. Mukherjee, J. Amer. Oil Chem. Soc., 77, 361366 and 367371
(2000).
102. Anon, Inform, 12, 921 (2001).
103. D. P. Geller and J. W. Goodrum, J. Amer. Oil Chem. Soc., 77, 111114 (2000).
104. V. Pandey, R. Banerji, B. S. Dixit, M. Singh, S. Shukla, and S. P. Singh, Eur. J. Lipid Sci.
Technol., 102, 463466 (2000).
105. F. D. Gunstone, in F. D. Gunstone, J. L. Harwood, and F. B. Padley, eds., The Lipid
Handbook, 2nd ed., Chapman and Hall, London, 1994, pp. 15 and 109.
106. J. T. Budin, W. M. Breene, and D. H. Putnam, J. Amer. Oil Chem. Soc., 72, 309315
(1995).
107. A. Bonjean, B. Monteuuis, and A. Messean, Oleagineux Corps gras Lipides, 2, 97100
(1995).
108. J. Zubr, Ind. Crops Products, 6, 113119 (1997).
109. A. Hebard, Lipid Technol., 10, 8183 (1998).
110. E. C Leonard, Inform, 9, 830838 (1998).
111. S. Moret, A. Dudine, and L. S. Conte, J. Amer. Oil Chem. Soc., 77, 12891292
(2000).
112. G. P. Savage, D. L. McNeil, and P. C. Dutta, J. Amer. Oil Chem. Soc., 74, 755759
(1997).
113. C. Bertoli, M. Stancanelli, D. Gumy, and P. Lambelet, J. Amer. Oil Chem. Soc., 75, 1037
1040 (1998).
114. M. H. Gordon, C. Covell, and N. Kirsch, J. Amer. Oil Chem. Soc., 78, 621624
(2001).
264 VEGETABLE OILS
115. M. L. Ruiz del Castillo, M. Herraiz, M. D. Molero, and A Herrera, J. Amer. Oil Chem.
Soc., 78, 12611265 (2001).
116. Anon., Lipid Technol. Newslett., 4, 2526 and 48 (1998).
117. Anon., Inform, 12, 1213 (2001).
118. Anon., Inform, 11, 404405 (2000).
119. R. Przybylski and K. C. Fitzpatrick, Inform, 13, 489 491 (2002).
120. J. Sargent and K. Coupland, Lipid Technol., 6, 1014 (1994).
121. V. K. S. Shukla and U. Blicher-Mathiesen, Fat Sci. Technol., 95, 367369 (1993).
122. V. K. S. Shukla, in R. J. Hamilton, ed., Developments in Oils and Fats. Blackie, London,
1995, pp. 6694.
123. M. Lipp. and E. Anklam, Food Chem., 62, 7397 (1998).
124. J. Birkett, in B. Rossell, ed., Oils and Fats Handbook, Vol 1, Vegetable Oils and Fats,
Leatherhead Food RA, Leatherhead, U.K., 1999, pp. Ch. 122.
125. Y. Jeyarani and S, Yella Reddy, J. Amer. Oil Chem. Soc., 76, 14311436 (1999).
126. T, Abbott, Inform, 8, 11691175 (1997).
127. N. Macfarlane and R. V. Harris, in E. H. Pryde, L. H. Princen, and K. D. Mukherjee, eds.,
New Sources of Fats and Oils, AOCS Press, Champaign, Illinois, 1981, pp. 103108.
128. C Croy, Inform, 5, 970971 (1994).
129. S. M. Osman, in E. H. Pryde, L. H. Princen, and K. D. Mukherjee, eds., New Sources of
130. S. Yella Reddy and T. Jeyarani, J. Amer. Oil Chem. Soc., 78, 635640 (2001).
131. R. J. Janssens, Lipid Technol. Newslett., 7, 133137 (1999).
132. R. J. Janssens and W. P. Vernooij, Inform, 12, 468477 (2000).
133. R. J. Janssens, Lipid Technol., 12, 5358 (2000).
134. G. D. Jolliff in E. H. Pryde, L. H. Princen, and K. D. Mukherjee, eds., New Sources of
135. T. A. Isbell, Lipid Technol., 9, 140144 (1997).
136. T. A. Isbell, in F. D. Gunstone, ed., Lipid Synthesis and Manufacture, Sheffield
Academic Press, Sheffield, U.K., 1998, pp. 401 421.
137. H. B. Frykman, T. Isbell, and S. C. Cermak, J. Surfactants Detergents, 3, 179183
(2000).
138. T. A. Isbell and S. C. Cermak, J. Amer. Oil Chem. Soc., 78, 527531 (2001).
139. C. C Akoh and C. V. Nwosu, J. Amer. Oil Chem. Soc., 69, 314316 (1992).
Handbook, 2nd ed., Chapman and Hall, London, 1994, pp. 7778.
141. J-W. Kao, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 75, 11031107
(1998).
142. F. Bondioli, C. Mariani, A. Lanzani, E. Fideli, and A. Muller, J. Amer. Oil Chem. Soc.,
70, 763766 (1993).
143. F. D. Gunstone, J. Oleo Science, 50, 269279 (2000).
144. Anon., Inform, 5, 713 (1994).
145. G. Ustun, L. Kent, N. Cekin, and H. Civelekoglu, J. Amer. Oil Chem. Soc., 67, 958960
(1990).
REFERENCES 265
146. M. A. M. Zeitou and W. E. Neff, Oleagineux Corps gras Lipides, 2, 245248 (1995).
147. K. Aitzetmuller, G. Werner, and S. A. Ivanov, Oleagineux Corps gras Lipides, 4, 385
388 (1997).
148. H. R. H. Takruri and M. A. F. Dameh, J. Sci. Food Agric., 76, 404 410 (1998).
149. P. C. Dutta, S. Helmersson. E. Kebedu, G. Alemaw, and L. A. Appelqvist, J. Amer. Oil
Chem. Soc., 71, 839845 (1994).
150. T. P. Hilditch and P. N. Williams, The Chemical Constitution of Natural Fats, 4th ed.,
Chapman and Hall, London, 1964, p. 336.
151. Anon., Lipid Technol. Newslett., 4, 105106 (1998).
152. J. M. Holmbach, A. A. Karlsson, and K. C. Arnoldsson, Lipids, 36, 153165 (2001).
153. B. G. Herslof, Lipid Technol., 12, 125128 (2000).
154. M. Zhou, K. Robards, M. Glennie-Holmes, and S. Helliwell, J. Amer. Oil Chem. Soc., 76,
159169 (1999).
155. D. M. Peterson, Lipid Technol., 14, 5659 (2002).
Handbook, 2nd ed., Chapman and Hall, London, 1994, p. 79.
157. Advisory Committee in Novel Foods and Procedures (UK), Annual Report, 1991.
158. T. Longvah and Y. G. Deosthale, J. Amer. Oil Chem. Soc., 68, 781784 (1991).
159. H-S Shin and S-W Kim, J. Amer. Oil Chem. Soc., 71, 619622 (1994).
160. I-H. Kim, H. Kim, K-T. Lee, S-H. Chung, and S-M. Ko, J. Amer. Oil Chem. Soc., 79,
363367 (2002).
161. S. M. Dyszel and B. C. Pettit, J. Amer. Oil Chem. Soc., 67, 947951 (1990).
162. M. Maskan and S. Karatas, J. Sci Food Agric., 77, 334340 (1998).
163. M. Yildiz, S. T. Gurcan, and M. Ozdemir, Fett/Lipid, 100, 8486 (1998).
164. S. P. Singh, S. Shukla, K. R. Khanna, B. S. Dixit, and R. O. Banerji, J. Sci. Food Agric.,
76, 168172 (1998).
165. T. R. Omara-Alwala, T. Mebrahtu, D. E. Prior, and M. O. Ezekwe, J. Amer. Oil Chem.
Soc., 68, 198199 (1991).
166. B Yang and H. Kallio, J. Agric. Food Chem., 49, 19391947 (2001).
167. H. Kallio, B, Yang, P. Peippo, R. Tahvonen, and R. Pan, J. Agric. Food Chem., 50, 3004
3009 (2002).
168. B. Yang, R. Karlsson, P. Oksman, and H. Kallio, J. Agric. Food Chem., 49, 56205629
(2001).
169. Yang Baoru, F. D. Gunstone, and H. Kallio, in F. D. Gunstone, ed., Lipids for Functional
Foods and Nutraceuticals, Oily Press, Bridgwater, U.K., 2003, pp. 263290.
170. Anon., Lipid Technol. Newslett., 2, 100 (1996).
171. R. Oliveira, M. F. Rodrigues, and M. G. Bernardo-Gil, J. Amer. Oil Chem. Soc., 79, 225
230 (2002).
172. P. J. Barnes and P. W. Taylor, J. Sci. Food Chem., 31, 9971006 (1980).
173. P. J. Barnes, Fat Sci. Technol., 84, 256269 (1982).
174. R. L. Wolff, N. A. Combe, F. Destaillats, C. Boue, D. Precht, J. Molkentin, and
B. Enstressangles, Lipids, 35, 825825 (2000).
175. E. Vos, Abstracts, AOCS Meetring, Montreal, QE, Canada, 2002.
266 VEGETABLE OILS
176. T. Koyano and K. Sato, in K. K. Rajah, ed., Fats in Food Technology, Sheffield Academic
Press, Shefield, U.K., 2002, pp. 129.
177. A. J, Dikstra, in K. K. Rajah, ed., Fats in Food Technology, Sheffield Academic Press,
Shefield, U.K., 2002, pp. 123158.
178. J. M. Block, D. Barrera-Arellano, M. F. Figueiredo, and F. A. C. Gomide, J. Amer. Oil
Chem. Soc., 74, 15371541 (1997).
179. R. Timms, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and App.
180. V. Gibon and A. Tirtiaux, Lipid Technol., 14, 3336 (2002).
181. T. Willner and K. Weber, Lipid Technol., 6, 5760 (1994).
182. E. Deffense, Lipid Technol., 7, 3438 (1995).
183. T. Koetsier, in F. D. Gunstone and F. B. Padley, eds., Lipid Technologies and App.
184. A. J. Dikstra, Inform, 8, 11501158 (1997).
185. R. C. Hastert, Lipid Technol., 10, 101105 (1998).
186. D. Allen, Lipid Technol., 10, 2933 and 5357 (1998).
187. D. Rousseau and A. G. Maringoni, in C. C. Akoh and D. B. Min, eds., Food Lipids:
Chemistry Nutrition, and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002,
pp. 301333.
188. A, Rosendahl and A. R, Macrae, in F. D. Gunstone and F. B. Padley, eds., Lipid
Technologies and App. lications, Marcel Dekker, New York, 1997, pp. 245259.
189. K. D. Mukherjee, in C. C. Akoh and D. B. Min, eds., Food Lipids: Chemistry Nutrition,
and Biotechnology, 2nd ed., Marcel Dekker, New York, 2002, pp. 751812.
190. W. M. Willis and A. G. Maringoni, in C. C. Akoh and D. B. Min, eds., Food Lipids:
pp. 839876.
191. A. R. Kavanagh. Oleagineux Corps gras Lipides, 4, 165168 (1997).
192. C. C. Akoh, in C. C. Akoh and D. B. Min, eds., Food Lipids: Chemistry Nutrition, and
Biotechnology, 2nd ed., Marcel Dekker, New York, 2002, pp. 877908.
193. U. T. Bornscheuer, M Adamczak, and M. M. Soumanou, in F. D. Gunstone, ed., Lipids
for Functional Foods and Nutraceuticals, Oily Press, Bridgwater, U.K., 2003, 149182.
194. Anon., Inform, 2, 678692 (1991).
195. L. J. M. van Soest and F. Mulder, Lipid Technol., 5, 6065 (1993).
New York, 2001, p. 192.
197. J. Wisniak, ed., The Chemistry and Technology of Jojoba Oil, AOCS Press, Champaign,
Illinois, 1987.
198. L. G. Angelini, E. Moschemi, G. Collonna, B. Polloni, and E. Bonari. Ind. Crops
Products, 6, 313323 (1997).
199. R. K. Downey and D. C. Taylor, Oleagineux Corps gras Lipides, 3, 913 (1996).
200. A. J. Kinney, Lipid Technol. Newslett., 5, 3639 (1999).
201. W. Friedt and W. Luhs, Fett/Lipid, 100, 219226 (1998).
202. J. Reske, J. Siebrecht, and J. Hazebroek, J. Amer. Oil Chem. Soc., 74, 989998
(1997).
REFERENCES 267
203. F. D. Gunstone and R. M. Pollard, in F. D. Gunstone ed., Structured and Modified Lipids,
Dekker, New York, 2001, pp. 155184.
204. V. C. Knauf and A. J. DelVecchio, in C. C. Akoh and D. B. Min, eds., Food Lipids:
pp. 967996.
205. S. Mielke, The Revised Oil World 2020Supply, Demand and Prices, ISTA Mielke
GmbH, Hamburg, Germany, 2002.
7
Lipid Oxidation:
Theoretical Aspects
K. M. Schaich
Rutgers University,
New Brunswick, New Jersey
1. INTRODUCTION
Many excellent chapters and books have been written on lipid oxidation (111).
Studies of lipid oxidation are provided differently by different authors: each scien-
tist studying lipid oxidation focuses on a different single aspect, such as following
early kinetics by oxygen uptake or LOOH production, determining volatile pro-
ducts by gas chromatography (GC) or nonvolatile products by high-performance
liquid chromatography (HPLC), or analyzing specific catalyst or antioxidant effects
on oxidation; oxidation mechanisms are then interpreted in that context. There have
been few attempts to integrate multiple stages or approaches to lipid oxidation, and
as a result, descriptions of lipid oxidation have been disparate and totally dependent
on the individual aspect being studied. This can be quite confusing to anyone not
deeply immersed in the field. That is not to say that any of the published informa-
tion is incorrect. Much of it, however, has been presented in too narrow of a context
to provide an accurate overall picture of complex lipid oxidation reactions.
Part of the problem stems from considering lipid oxidation as precisely follow-
ing classic free radical chain reactions. To be sure, lipids do oxidize by a radical
chain mechanism, and they show initiation, propagation, and termination stages
269
270 LIPID OXIDATION: THEORETICAL ASPECTS
CLASSIC FREE RADICAL CHAIN REACTION MECHANISM OF LIPID OXIDATION
Initiation (formation of ab initio lipid free radical)

ki
L1H L1 (1)
Propagation
Free radical chain reaction established
ko
L1 + O2 L1OO (2)
k
kp1
L1OO + L2H L1OOH + L2 (3)
kp1
L2OO + L3H L2OOH + L3 etc. LnOOH (4)
Free radical chain branching (initiation of new chains)

kd1
LnOOH LnO + OH (reducing metals) (5)
kd2
LnOOH LnOO + H (oxidizing metals) (6)
kd3
LnOOH LnO + OH (heat and uv) (7)
LnO kp2 LnOH (8a)

LnOO + L4H LnOOH + L4 (8b)
kp1
HO kp3 HOH (8c)
kp4
L1OO + LnOOH L1OOH + LnOO (9)
kp5
L1O + LnOOH L1OH + LnOO (10)
Termination (formation of non-radical products)

Radical recombinations
Ln Ln kt1 (11a)
LnO + LnO polymers, non-radical monomer products
kt2 (ketones, ethers, alkanes, aldehydes, etc.) (11b)
LnOO LnOO kt3 (11c)
Radical scissions
kts1 (12a)
LOO non-radical products
LO kts2 (aldehydes, ketones, alcohols, alkanes, etc.) (12b)
i - initiation; o - oxygenation; - O2 scission; p - propagation; d - dissociation; t - termination;

ts - termination/scission
Figure 1. Classic free radical chain reaction mechanism of lipid oxidation with propagation by a
series of hydrogen abstractions.
INTRODUCTION 271
as is normally depicted (Figure 1). However, the generalized reactions of the classic
free radical chain reaction scheme are very much oversimplified and, because they
do not portray the wide range of competing side reactions that contribute to the
great complexities of lipid oxidation, they are often inconsistent with observed
oxidation kinetics and product mixes.
Thus, this chapter presents lipid oxidation from a broad systems perspective to
make the overall process logical, reconcile some common inconsistencies in pro-
posed mechanisms, address some of the complexities that are important in directing
downstream pathways and ultimate product mix, and develop an integrated view of
lipid oxidation. In doing so, attempts are made to bridge basic chemistry to applied
lipid and food chemistry. Old literature is cited liberally, despite current trends to
ignore anything outside the previous two to five years, because the fundamental
chemistry is still relevant, the early researchers in the field deserve recognition
for their ground-breaking observations, and the information needs to be revisited
to remind us of what already has been done to prevent rediscovering the wheel.
Furthermore, consideration of fundamentals too often gets lost in the sophistication
of applications, particularly in biological systems. Lipid oxidation processes in
foods or biological tissues may be more complicated, but will still follow funda-
mental mechanisms identified in simpler chemical reactions. Greater consideration
of details learned from fundamental chemistry should help clarify and elucidate
mechanisms and kinetics in complex media.
In particular, this chapter will stress the need to look beyond the classic radical
chain reaction. Lipid oxidation mechanisms have been proposed based on kinetics,
usually of oxygen consumption or appearance of specific products (e.g., LOOH) or
carbonyls (e.g., malonaldehyde), assuming standard radical chain reaction
sequences. However, when side reactions are ignored or reactions proceed by a
pathway different from that being measured, erroneous conclusions can easily be
drawn. The same argument holds for catalytic mechanisms, as will be shown in
the discussion about metals. In the past, separation and analysis of products was
laborious, but contemporary methods allow much more sensitive detection and
identification of a broad mix of products. Thus, multiple pathways and reaction
tracks need to be evaluated simultaneously to develop an accurate picture of lipid
oxidation in model systems, foods, and biological tissues.
In vivo lipid oxidation will not be covered, although the fundamental chemistry
presented certainly applies wherever lipid oxidation occurs. Also, in light of the
product and reaction pathway complexities presented in this chapter, kinetics of
lipid oxidation will not be covered. That is not to say that kinetics are not important.
However, kinetic analyses are always based on assumptions, and kinetic equations
derived in different studies are often difficult to reconcile even in simple systems.
The broader consideration being urged in this chapter poses even greater chal-
lenges. A citation from the past remains cogently relevant today: in view of the
numerous possible routes that might be followed in the initiation, propagation, and
termination stages of the decomposition process, kinetic analysis of the results has
proved to be difficult [(12) citing (13)].
1.1. Classic Radical Chain Reaction Scheme

Lipid oxidation has long been recognized as a free radical chain reaction (1418),
and the classic chain reaction scheme with three phases has been repeated in many
forms. Figure 1 is one version. Sometimes secondary abstraction reactions of lipid
alkoxyl radicals (LO ) and peroxyl radicals (LOO ) are presented as initiation
reactions because they form L radicals. That is true when lipid oxyl radicals
are from outside sources, e.g., lipoxygenase reactions followed by Fe2 and Fe3
reactions with LOOH. However, in the following discussion, LO and LOO
deriving from the initial L or its subsequent reactions are considered to mediate
propagation or chain branching (initiation of secondary chains) rather than ab initio
initiation.
The driving force in the chain reaction is the repeated abstraction of hydrogens
by LOO to form hydroperoxides plus free radicals on a new fatty acid. The process
continues indefinitely until no hydrogen source is available or the chain is inter-
cepted. The radical chain reaction imparts several unique characteristics to lipid
oxidation:
1. Lipid oxidation is autocatalyticonce started, the reaction is self-propagat-

ing and self-accelerating.
2. G (product yield) 1, i.e., many more than one LOOH are formed and more
than one lipid molecule are oxidized per initiating event. Chain lengths as
long as 200 to 300 lipid molecules have been measured (19, 20) showing how
effective a single initiating event can be. However, this also points out one
reason why it has been so difficult to study initiation processesinitiators
become the proverbial needle in a haystack once oxidation chains become
established.
3. Very small amounts of pro- or antioxidants cause large rate changes.
4. The reaction produces multiple intermediates and products that change with
reaction conditions and time.
These features present distinct challenges in measuring and controlling lipid oxida-
tion, and are part of the reason why lipid oxidation is a major problem in vivo and
in storage stability of foods.
Citation of the classic chain reaction for lipid oxidation persists even though, as
product analysis and studies of mechanisms have become more sophisticated, there
is now considerable evidence that only Reactions 1, 2, and 5 (and perhaps also 6) of
Figure 1 are always present. Research has shown that, although hydrogen abstrac-
tion ultimately occurs, it is not always the major fate of the initial peroxyl or alkox-
yl radicals. Indeed, lipid alcohols from H abstraction are relatively minor products
of lipid oxidation. There are many competing alternative reactions for LOO and
LO that propagate the radical chain but lead to different kinetics and different pro-
ducts than expected from the classic reaction sequence (5, 6, 21). A more detailed
consideration of each stage shows how this basic radical chain sequence portrays
only a small part of the lipid oxidation process and products, and a new overall
reaction scheme for lipid oxidation is needed.
INITIATION 273
2. INITIATION (LH ! L )
Initiation of lipid oxidation produces the ab initio lipid free radicals, L . The initia-
tion process is not well understood, so it is usually represented in reaction schemes
merely as an X or ? over the reaction arrow. Lipid oxidation is a very facile
reaction that is nearly ubiquitous in foods and biological systems, so it is often trea-
ted as an instantaneous reaction that just happens, and has been referred to as
spontaneous (22). Nevertheless, lipid oxidation is not a spontaneous reaction!
Thermodynamically, oxygen cannot react directly with double bonds because the
spin states are different (Reaction 1). Ground state oxygen is in a triplet state (two
free electrons in separate orbitals have same spin direction, net positive angular
momentum), whereas the double bond is in a singlet state (no unpaired electrons,
paired electrons are in the same orbital and have opposite spin, no net angular mo-
mentum). Quantum mechanics requires that spin angular momentum be conserved
in reactions, so triplets cannot invert (flip spins) to singlet states. Reaction then
demands that the double bond be excited into a triplet state, which requires prohi-
bitive amounts of energy (Ea 3565 kcal/mole). Thus, no direct reaction occurs.
O O + C C ROOH 1
Triplet Singlet
To overcome this spin barrier, initiators or catalysts are required to start the lipid
oxidation process by removing an electron from either the lipid or oxygen or by
changing the electron spin of the oxygen. As only trace amounts of catalysts are
needed, many situations that appear to be spontaneous or uncatalyzed are actually
driven by contaminants or conditions that have gone undetected or unconsidered.
Indeed, in most foods, biological systems, and laboratory experiments, it is fair
to say that multiple catalysts and initiators are always operative.
The most common initiators are described below. Somewhat more detail than in
most reviews of lipid oxidation is presented because control of lipid oxidation ulti-
mately demands control of initiation. Antioxidants that scavenge lipid free radicals
after they are formed are always playing catch up, and may be totally or partially
ineffective if the total radical load from initiation (whether from known or unknown
sources) is excessive. To achieve full protection against lipid oxidation and attain
long-term stability of any material, control strategies must include elimination, or at
least inhibition, of initial alkyl radical production in lipids.
2.1. Catalysts
2.1.1. Metals
Redox-active metals are the initiators of perhaps greatest importance for lipid
oxidation in oils, foods, and biological systems because they are ubiquitous and
active in many forms, and trace quantities ( micromolar) are sufficient for effec-
tive catalysis (2326). Only metals undergoing one-electron transfers appear to be
active catalysts; these include cobalt, iron, copper, manganese, magnesium, and
vanadium. Metals that oxidize by two-electron transfers, e.g., Sn2 and Tl, are not
active (23).
The mechanisms and rates of metal-catalyzed initiation operative in individual
reaction systems are determined by a complex mixture of factors: the metal and
type of complexes it forms (inner sphere or outer sphere), the chelator or complex-
ing agent, redox potential of the metal and its complexes, solvents, phase localiza-
tion of the metal, and availability of oxygen or preformed hydroperoxides. The
reactions outlined below show the multiplicity of mechanisms possible.
Direct initiation through higher valence metals involves direct electron
transfer from the metal to a bond in the lipids and is the simplest mechanism
for metal catalysis. Electron transfer to methyl linoleate is exothermic
(H 62:8 kJ; 15 kCal), so is probably the dominant initiation mechanism
with lipids (23, 27). Ab initio lipid radicals are formed directly by removing an
electron from a double bond (Reaction 2) (28, 29) or, more generally, from the
CH bond of any labile H in lipid molecules (e.g., allylic hydrogens) (Reaction 3),
or via subsequent secondary hydrogen abstraction reactions, as designated in the
bracketed reactions.
LH
RCH CHR + M(n+1)+ RCH CHR + Mn+ L + RH 2
LH
RH + M(n+1)+ R + H+ + Mn+ L + RH 3
RCOOH + M(n+1)+ RCOO + H+ + Mn+ R + CO2 4

LH
RCHO + M(n+1)+ RCO + H+ + Mn+ L + scission products
5
Reactions 2 and 3 have been proposed as the primary mode of catalysis for Co
(30), Mn (31), and Cr (32). However, it must be pointed out that metal reactivity
can change tremendously with complexing agent, which shifts redox potentials, and
with solvent, which alters acid/base properties and electron transfer efficiency. Elec-
tron transfer oxidations to generate L are extremely rapid in nonpolar media (33, 34),
including neat oils, and are less efficient in aqueous or polar protic solvents.
Analogous electron transfers involving the carboxylic acid group of fatty acids
(Reaction 4) or lipid oxidation products such as aldehydes (Reaction 5) (35) can
also occur to form radicals that are potential initiators. Reaction 4 with free car-
boxylic acids has been demonstrated with cobalt and short-chain organic acids
(29, 36, 37), so the potential exists for its occurrence with fatty acids. The aldehyde
reaction (Reaction 5) is strongly catalyzed by Cu2, Co3, and Mn2 (3840) and,
being inhibited by water competition for ligand sites, occurs primarily in organic
solvents or neat lipids. However, the reaction is relatively slow and not competitive
with the first three reactions under most food conditions.
The rate and selectivity of the direct electron transfers of Reactions 25 are
influenced by the type of metal complex formed. In outer sphere complexes,
electrons flow directly between the valence shell of the metal and the target group;
INITIATION 275
electron transfer is fast and selective. Inner sphere complexes involve ligand bind-
ing to the metal and electron flow is through the ligands; electron flow is slow and
less discriminating (41). Iron forms mostly outer sphere complexes. Copper forms
mostly inner sphere complexes with organic substrates, especially in nonpolar sol-
vents, but most inorganic copper salts catalyze direct electron transfer through outer
sphere complexes. Cobalt forms inner sphere ligand complexes in nonpolar solvents
such as oils (42); but in polar solvents and with polar ligands, cobalt catalyzes elec-
tron transfer by an outer sphere mechanism (29, 43, 44). The difference may seem
academic, but it partially explains differences in reactivity, kinetics, and products
for different metals and in some cases for different complexing agents, and it points
out the need to understand mechanisms when determining which products to ana-
lyze to most accurately evaluate extent of oxidation.
Direct initiation by lower valence states (Mn] of metals proceeds through
formation of activated complexes with O2 (23, 45)mostly via inner sphere com-
plexes. As free reduced metals react rapidly with oxygen (Reaction 6a), this
mechanism is active primarily when chelators specifically stabilize the reduced
metals. These reactions also proceed mostly facilely in nonpolar solvent (46),
e.g., in hydrophobic lipid phases of membranes or in oils.
Mn+ + O2 M(n+1)+ ...O2
LH
M(n+1)+ + O2 HOO L + H2O2 6a
LH
L + M(n+1)+ ...O2H 6b
LH LH
L + Mn+ + HOO L + H2O2 6c
LH
LO + M(n+1)+ ...OH 6d
Mn+LH LH
M(n+1)+ L + M(n+1)+ + HOO L + H2O2 6e
Direct initiation by either mechanism is characterized by a lack of induction

period (47) and is most efficient by metals that are strongly oxidizing (Co and
Fe) or can form metal-oxygen complexes (Co and Cu).
Indirect initiation of lipid oxidation by reduced metals (Co2, Fe2, V2, Cr2,

Cu , Ce3, Mn2) occurs by two different mechanisms, depending on the pO2 of
the system and levels of preformed or nonlipid hydroperoxides:
a. autoxidation of reduced metals to generate oxygen radicals that then react

with lipids (27) occurs at moderate to high pO2, e.g., for iron:
H+
Fe2+ + O2 Fe3+ + O2 HOO L + H2O2 7
2 O2 or O2 / HOO H2O2 + O2 8
H2O2 + Fe2+ HO + OH + Fe3+ 9

HO + LH H2O + L 10
Evidence for this process has been obtained in systems of charged micelles
prepared from linolenic acid (48) and by chemiluminescence in very early stages
of lipid oxidation in oils and a variety of foods (49).
b. reduction or oxidation of hydroperoxides (either from other sources or from

preformed lipid hydroperoxides) to RO or ROO , respectively (Reactions 11
and 12), which then react with lipids; dominates under conditions of low
metal, substrate, and oxygen concentration (27, 35). Lipid hydroperoxide
reduction is an extremely facile reaction. The activation energy is consider-
ably lower than that of H2O2 (EaLOOH 12.5 kCal; EaHOOH
35 kCal) and
the rate of reduction is correspondingly several orders of magnitude faster:
kLOOH 5 109 L mol1sec1 (50, 51) and kHOOH
104 M1sec1 (52).
fast LH
Fe2+ + ROOH Fe3+ + RO + OH ROH + L 11
extremely slow LH
Fe3+ + ROOH Fe2+ + ROO + H+ ROOH + L 12
Metals that form complexes with oxygen also form intermediate complexes with
hydroperoxides during oxidation and reduction, particularly at low hydroperoxide
concentrations and in nonpolar solvents, as shown in Reactions 13 and 14 for cobalt
(5357). However, in polar solvents, cobalt acts by direct electron transfer, as in
Reactions 11 and 12 (58). Copper forms similar complexes with hydroperoxides (59).
Co2+ + LOOH [Co2+ HOOL] Co3+OH + LO 13
Co3+ + LOOH [Co3+ HOOL] Co2+ + LOO + H 14
Metal autoxidation and hydroperoxide decomposition are both very active pro-
cesses in foods, oils, and biological tissues where metals are always present. Con-
sidering the constant presence of peroxides from various sources in all natural
materials, it could reasonably be argued that peroxide decomposition is the major
practical source of initiators for lipid oxidation. However, these reactions are per-
haps even more important in accelerating chain branching in later stages of oxida-
tion when higher concentrations of LOOH accumulate.
Whatever the operative mechanism for a given system, the effect of metals is
tremendously amplified when redox cycling occurs. Coordination of redox pairs
of metals has the same effect in early stages of lipid oxidation that bimolecular
decomposition has in later stages (60):
Metal redox cycles Bimolecular LOOH decomposition.
Mn+ + LOOH Mn+1 + LO + OH
Mn+1 + LOOH Mn + LOO + H+

2 LOOH LO + LOO + H2O 2 LOOH LO + LOO + H2O
15
INITIATION 277
Initiation by hypervalent metal-oxygen complexes, e.g., Fe4 O. The ques-

tion of oxidation catalysis by hypervalent iron also needs to be raised because new
evidence is suggesting that some of the mechanisms of metal catalysis described
above may actually be driven by hyperoxidized iron. Ferryl iron complexes
[Fe(IV)O; FeO2] (61) and perferryl iron [Fe(V)] catalyze oxygen insertion
into C H to yield epoxides, ketones, and alcohols. However, the mechanisms for
both formation and reactions of Fe(IV) complexes are still unclear, and their invol-
vement in initiation of free radical autoxidations is hotly debated. Walling (62),
highly respected for his research on Fenton chemistry, disputes the Fe4 pathway
and argues that one-electron oxidation to Fe3 is the major pathway for most iron
compounds. Nevertheless, it is well-known that hypervalent iron complexes are
transient intermediates in many heme enzyme mechanisms, as will be discussed
later, and there is now unequivocable spectroscopic and EPR evidence for Fe(IV)
participation in nonheme iron enzymes as well (6365). Still, hypervalent iron com-
plexes were considered too difficult to form and too unstable to be relevant in solu-
tion chemistry without porphyrins or proteins as electron sinks until observations
that iron reacted with hydrogen peroxide in acid to give the same nonselective pro-
ducts as HO in pulse radiolysis, whereas in neutral and alkaline solutions, products
were more stereospecific and selective (66). This led to the proposal that hyperva-
lent iron does form transiently in some solution reactions and may be the catalytic
species involved rather than hydroxyl radicals. The two-electron oxidation of fer-
rous iron yields an equivalent ferryl peroxyl complex, 2 FeII O2 [FeOOFe]IV.
There are ten total unpaired spins on each side of the equation, the thermodynamics
are favorable (H 17 kCal, F 11), and the reaction can occur without a net
spin change (67). Pulse radiolysis studies show that FeIV and FeV have significant
lifetimes when complexed with simple ligands like hydroxide and pyrophosphate
and, as such, are plausible intermediates in iron-catalyzed oxidations of organic
compounds (68). Thus, participation of Fe(IV)O may explain aspects of kinetics
and product distributions that have not fit traditional Fe3/Fe2 mechanisms
(Reactions 614).
There is now substantial evidence that the metal-oxygen complexes described
above do indeed form hypervalent intermediates that catalyze both radical and non-
radical oxidations (63, 6978). Most is known about Fe(IV) and Fe(V) complexes,
providing support for the idea that hypervalent iron is at least one catalyst in Fenton
reactions (79); analogous complexes have been identified for Cu2 (73, 80) and
Co2 (73). Both Fe2 (74) and Fe3 (63, 75) complexes participate, although
through different routes: Fe2 HOOH yields Fe4O; Fe3 HOOH yields the
3
Fe OH complex, which is functionally equivalent to Fe4 (79).
Figure 2 presents overall reaction schemes for the Fe2 and Fe3 reactions. The
schemes include radical and nonradical pathways and represent reactions for both
H2O2 and ROOH. In the figure, ROOH is used to indicate lipid hydroperoxides
to avoid confusion with metal ligands, L, and for simplicity, only the lipid
species are carried completely through reaction sequences. These reactions have
been determined using H2O2, but have not yet been demonstrated specifically with
lipids. Nevertheless lipid hydroperoxides are expected to follow the same general
A. Ferryl iron (Fe 4+) complexes from Fe 2+-hydroperoxide reactions:

Site-specific
a oxidation
b
1
HLFe2+ + ROOH HLFe2+ (ROOH) HLFe3+ + RO + HO
Cage Reaction
3 2
ng
k- biti
Bac
OH Fe2+
HLFe4+ LFe3+ LFe2+ + H2O + OH
OR 8
Direct e transfer through 4 6 2 e transfer through

outer sphere complex RH RH (or ROOH) inner sphere complex
7 OH
HLFe3+ + OH + R + H2O HLFe4+ + ROH
RH
ROOH
6
LFe2+ + R + H2O
(ROO )
B. Ferryl (Fe 4+) and perferryl (Fe5+) iron complexes from Fe 3+-hydroperoxide reactions:
R R
O O O
O O R O H
heterolytic LFe III LFe III LFe IV
H
O H O O H
O OR H H stereospecific
LFeIII hydroxylation
O H
H O LOOH
homolytic RO + H + LFe IV LOO , L
LH
radical O H
generation
Figure 2. Formation of ferryl iron in initiation and catalysis of lipid oxidation: Reaction schemes
for formation of hypervalent iron states by Fe2 and Fe3 complexes and subsequent reactions
leading to radicals that can initiate lipid oxidation. L, metal ligand; R, alkyl or acyl group. Fe2
sequence (71, 73); Fe3 sequence (81), adapted.
pathways as H2O2, although perhaps even more facilely because the O O bond
energy is lower in lipids (HOOH 51 kCal mol1 vs. LOOH 2535 kCal mol1).
In Fe2 reactions (Figure 2A), the initial Fe-hydroperoxide complex formed
with H2O2 or LOOH can undergo a traditional one-electron oxidation
(Reaction 1), yielding Fe3 and hydroxyl or alkoxyl radicals, respectively, in a
INITIATION 279
cage reaction. In systems where the radicals can diffuse out readily, they escape to
react and initiate new lipid oxidation chains (A), or while still in the reaction cage,
the oxyl radical can backbite on the Fe3 (B) and oxidize it to Fe4 (Reaction 2).
Alternatively, the Fe-hydroperoxide complex can generate the ferryl iron complex
directly by two-electron oxidation to the Fe4 complex (Reaction 3). Fe4 reactions
are responsible for the catalytic power and greatly increased radical production.
Fe(IV)O abstracts hydrogens even more rapidly that HO (k > 109 L M1s1).
It can abstract allylic hydrogens from unsaturated fatty acids to form the ab initio
L radical or it can abstract H from lipid hydroperoxides to give LOO that will
propagate radical chains. Thus, through either a one-electron process involving out-
er sphere complexes (Reaction 4) or a two-electron process with inner sphere com-
plexes (Reaction 56), radicals are produced in any unsaturated fatty acid or lipid
hydroperoxide that comes in contact with the Fe4 complex.
It should be stressed that the radicals evolving from Reaction 3 are not from the
initial complexed hydroperoxide, but rather are in new lipid molecules. The initial
hydroperoxide serves only to activate the iron to Fe4 in contrast to Reaction 1 in
which the hydroperoxide was the direct reactant and source of propagating radicals.
In ferryl iron reactions, oxygen groups from the initial hydroperoxides are inserted
or transferred directly to a substrate without radical intermediates, yielding alco-
hols, ketones, epoxides, or water. This finally explains earlier observations of
crypto HO , hidden HO that hydroxylated target compounds but could not be
detected free in solution (82). In terms of kinetics, oxidation rates much greater
than would be predicted for trace levels of hydroperoxides and iron can thus be
achieved by Fe4 because Reactions 37 in Figure 2A are much faster than Reac-
tion 1, the selectivity of Fe4 in hydrogen abstractions is greater than either HO or
RO , and Fe4 both initiates and propagates radical chains. Reaction 8 depicts the
reduction of Fe4 complexes in the presence of excess Fe2 to yield two Fe3 com-
plexes with concurrent release of water and hydroxylated products. This is
one explanation for the loss of catalytic effectiveness at high concentrations of metals.
In the Fe3 reactions (Figure 2B), hydroperoxides bind to the iron atom and sub-
sequent formation of the Fe4 complex is accompanied by either heterolytic scis-
sion of the O O bond to form hydroxylated products or homolytic scission to
release hydroxyl or lipid alkoxyl radicals. Current evidence suggests that
Fe3 H2O2 and Fe3 LOOH form different Fe4 complexes, so H2O2 undergoes
preferential heterolytic scission, whereas homolytic scission is the almost exclusive
route for organic hydroperoxides (81). For LOOH, increased conversion to initiat-
ing LO and rapid H abstractions by Fe4 to produce L or LOO combine to tre-
mendously accelerate generation of new chains of radical reactions, and it accounts,
at least in part, for the great catalytic effectiveness of even traces of lipid hydroper-
oxide.
Both Fe2 and Fe3 complexes undergo two-electron oxidations to yield Fe4
and Fe5 states, respectively. The Fe5 state, in particular, is achievable with inor-
ganic and small organic ligands because both electrons needed for oxidation come
from the Fe. This doesnt happen with hemes, where one electron comes from the
iron and the other is taken from the porphyrin or apoprotein (81).
There is much still to be learned about conditions required for formation of fer-
ryl or other hypervalent iron complexes, the actual structure of the complexes under
different circumstances, the kinetics and mechanisms by which they react, and the
overall consequences to lipid oxidation. The factors that appear to be most impor-
tant include the following:
1. Ligand structure. Highly electrophilic ligands are most effective in producing

Fe4 (73). Changing the ligands alters the lifetime of FeIVO complexes.
Longer lifetime translates as lower reactivity; shorter lifetime results from
higher reactivity, but makes the state more difficult to detect and study (65).
2. Redox potential of the complex (73).
3. Spatial arrangement of ligand components relative to the iron atom (64, 71,
74, 78).
4. Acid-base properties of the ligands (64, 77). The presence of a Lewis base in
the ligand exerts a tremendous push effect on the OH group in the
hydroperoxide, enhancing both formation of FeIVO and homolysis of
OO in Fe3 OOH complex (increased release of HO ) (77).
5. Relative proportions of iron and hydroperoxide. High iron favors oxygen
insertion and formation of ketones, whereas 1 : 1 Fe : hydroperoxide shifts
products to epoxides (73); excess ROOH yields large amounts of free radicals
caused by a shift of the iron to high spin [FeL(Z1-OOH)2] states and rapid
reaction of iron with ROOH instead of substrates (83).
6. Solvent and presence of water. 15% water decreases the redox potential of
iron complexes and increases homolytic scission to HO radicals; in aprotic
solvents, heterolytic scission and oxygen insertion products predominate (69).
7. Chemical structure of hydroperoxide forming the initial complex. This alters
the structure and spin state of the Fe4 complex and, consequently, affects
dominant product pathways (73). H2O2 forms low spin complexes that
undergo heterolytic scission, whereas alkyl hydroperoxides form high spin
complexes that release alkoxyl radicals in homolytic scissions (81).
2.1.2. Light
2.1.2.1. Ultraviolet LightDirect Effects Direct initiation of lipid oxidation by

ultraviolet light,
h
R1CH(R2)R3 R1C(R2)R3 + H or R1CHR2 + R3 16
requires either direct deposition to sufficient energy to break covalent bonds or

transformation of light energy to chemical energy that can catalyze the reaction.
The Eas for L H and L L scission reactions are higher than the correspond-
ing bond energies (
98.4 kCal/mol and 83.1 kCal/mol, respectively), and this
photon energy is available only at wavelengths <
254 nm (Table 1). In fact, however,
most ultraviolet light damage to lipids occurs at wavelengths less than 200 nm.
INITIATION 281
TABLE 1. Energies of Light at Various Wavelengths vs. Typical Energies of Bonds

in Lipids.
eVa kJb kCalb Bond Dissociation Energy E

(Physicists) (Chemists) (Biologists) Bond kJ/molc kCal/mold
200 6.2 596 143 CC 612 146

230 5.4 518 124 H
O 463 111
260 4.8 458 110 H
C 412 99
290 4.3 411 98 O
C 360 86
320 3.9 372 89 C
C 348 83
350 3.5 341 82 N
C 305 73
380 3.3 314 75 O
O 157 35
410 3.0 291 70
440 2.8 271 65
470 2.6 254 61
510 2.4 234 56
540 2.3 221 53
570 2.2 209 50
600 2.1 199 48
630 2.0 189 45
660 1.9 181 43
700 1.8 170 41
a
See (84).
b
Calculated from E Nhc/l, where N Avogadros no.(6.02*1023 photons/mol), h Plancks constant
(1.58*1034 cal/s or 6.6*1034 J/s or 4.36*1015ev/Hz), c speed of light (3*1017nm/s), l wavelength (84).
c
See (85).
d
See (86).
Although ultraviolet light is thermodynamically capable of producing L radi-

cals directly in lipids, the process is not a competitive reaction. In solution, ioniza-
tion generally requires energies of about 56 eV (87), available only at wavelengths
< 230 nm, so direct L production is not easily achieved by ultraviolet irradiation.
When ionization does occur, there usually is not enough energy to push molecular
segments apart, except when the sample is heated, so radicals recombine in cage
reactions and do not initiate chains. Also, UV initiation is kinetically slow and
very selective because the absorbed energy must match E between energy states
of elements and bonds.
The principal light-absorbing groups of lipids are double bonds, peroxide O O
bonds, and carbonyls; the last two are most important. The primary mechanism by
which ultraviolet radiation initiates lipid oxidation is actually indirect, mediated
through homolytic scission of any preformed hydroperoxides to generate the true
initiators LO , HO , and RO that abstract hydrogens from lipid molecules
and form the ab initio L .
ROOH RO + OH ROH (17a)

h (UV) L2H
HOOH HO + OH H2O + L2 (17b)
LOOH LO + OH LOH (17c)
When the reaction involves LOOH, UV light is also a potent catalyzer of pro-
pagation and, from a practical standpoint, exerts its main effects in that stage. In
fact, it is often difficult to maintain LOOH on the lab bench for reaction or analysis,
especially under fluorescent lights, because the decomposition is quite rapid. Hand-
ling samples for analysis of LOOH and separation of hydroperoxides by column
chromatography are best done under red light or at least with the vessel or column
wrapped in aluminum foil or other light-impermeable material and also in the
cold, as will be shown later.
A second source of UV-induced radicals to initiate lipid oxidation is excitation
of carbonyl compounds (88). The carbonyl n ! p transition (340 kJ/mol) occurs
when light is absorbed at 350 nm and lower wavelengths (87).
h LH
C O C O* C O C OH + L 18
The production of H2O2 during UV-irradiation in aqueous solutions should not

be overlooked as another source of initiating radicals from ultraviolet light. H2O2
yields of 3.7 and 1.3 mmol per mole L and Ln, respectively, have been measured in
solutions exposed to UV light (89), more than enough for very active initiation of
lipid autoxidation.
Contrary to what might be expected from their reactivity, double bonds are
not effective targets for UV light. The energy of the p ! p excitation transition
in conjugated dienes is 560 kJ/mol and in isolated double bonds is 680 kJ/mol,
which is only achievable at the lower limit of the ultraviolet ranges (215 nm and
<180 nm, respectively) (87). Long periods of irradiation are required because
absorption of light must produce excited states and ionization before bond scission
can occur. Free fatty acids, even saturated ones, are more susceptible to UV
radiation than esters because the C C and C H bonds a to the COOH are
activated 58 kCal/mol by mesomerism and thus are more susceptible to rupture
by light energy. Chain reactions are not involved, and decarboxylation products
result (12).
Electron paramagnetic resonance (EPR) studies of lipid free radical production
during UV radiation have found it exceedingly difficult to detect L or subsequent
LOO in highly purified systems (90). Using nitrosodurene as a spin trap to detect
free radicals too short-lived for direct observation by EPR, a mixture of free
radical adducts were observed, consistent with H abstractions at allylic carbons
for unsaturated fatty acids and carbons a and b to COOH in saturated fatty
acids (91, 92). However, since UV irradiation produces radicals in both benzene
and nitrosodurene (both of which were trapped), the lipid radicals detected are
more likely to have been produced by secondary H abstractions than light-induced
bond scissions. Similarly, radicals were only detected in light-irradiated unsaturated
fatty acids at 77 K when photosensitizers were included. Thus, UV-induced
direct bond scission that could start radical chain reactions in lipids does not
seem likely.
INITIATION 283
2.1.2.2. Photosensitization of lipid oxidation by visible light Visible light (>400 nm)
lacks the energy to produce radicals directly. However, when the low level quantum
energy of visible light is collected by specifically absorbing molecules, it is trans-
formed to chemical energy that can drive reactions. This process, called photosensiti-
zation, involves excitation of the sensitizer, then transfer of the excitation energy to
bonds to form free radicals directly (Type 1) or to oxygen to form singlet oxygen,
which then adds to double bonds of unsaturated fatty acids without generating radicals
(Type 2):
Type 1 sensitization free radical ! L e transfer reaction

Type 2 sensitization 1 O2 ; singlet oxygen ! LOOH no free radicals produced
By these two reactions, photosensitization provides the spin state requirements

cited above for reaction of oxygen with double bonds, namely a change in oxygen
spin state from triplet to singlet or loss of a bonding electron from the target
molecule. Type 1 reactions are oxidations, whereas Type 2 reactions are oxygena-
tions (oxygen insertions) that are 1500 times faster than with normal triplet
oxygen (93). Photosensitizers in foods and biological materials are usually, but
not exclusively, pigments. Chlorophyll, in particular, acts as Natures light
gatherer, collecting low-energy visible light and converting it to chemical energy
in plants. Other photosensitizers include flavins (especially riboflavin), porphyr-
ins, aromatic amino acids, and any molecules with carbonyls or an extended
conjugated double bond system (94). Some photosensitizers, including chloro-
phyll, catalyze by both free radical and singlet oxygen mechanisms, with the
dominant reactions depending on substrate and reaction conditions (9597).
Other photosensitizers are very specific in their reactions (98). With nearly all
sensitizers, regardless of final mechanism or product, the initial steps
involve excitation of the sensitizer to its lowest triplet level, 3S* (requires the least
amount of energy). The triplet sensitizer then directs subsequent reaction, trans-
ferring the excitation either to the lipid substrate (Type I) or to oxygen (Type II)
(97, 99).
h
1S 3S*
LH O2 H+
SH + 0S + LOO
Type 1 Redox/ L LOOH
H+
(Gollnick)
O2
(S + L+ ) or (S+ + L )
Free Radical L
0S + LOOH
e H+
Type II Oxygenation
A. Direct LH
1S + 1O 0S + LOOH (Kautsky-Foote)
2
3O
LH
1S + 1O LOOH (Gollnick)
B. Indirect 2
S O2* LH H+
3O 0S (Schnberg-Schenck)
2 + LOO LOOH
1
O2 itself is not a free radical generator, but rather generates hydroperoxides
that are precursors for initiating radicals. In a concerted ene reaction, 1O2
attaches to either carbon of a double bond and abstracts an allylic proton to
form a hydroperoxide directly (Reaction 19); no free radical is involved. There
is little preference for carbon position, so approximately equal amounts of
LOOH are produced at both ends of the original double bond. At the same
time, a new double bond in trans-configuration is formed between the other dou-
ble bond carbon and the allylic position. When the hydroperoxide is decomposed
to radicals by metals, light, or heat, subsequent hydrogen abstractions initiate
autoxidation chain reactions.
R R
H H 19
O O
O R O
R
In polyunsaturated fatty acids with nonconjugated double bonds, 1O2 reacts with
each CC as if it were isolated, so it yields roughly equivalent amounts of hydro-
peroxides at both internal and external positions. However, if the double bonds are
conjugated (e.g., natural conjugated linoleic acid or oxidized linoleic acid), cyclic
endoperoxides are formed.
O O O O
20
Dioxetane formation by 1O2 does not occur in lipids because it requires an elec-
tron-donating atom such as N or S next to the double bond (100).
It has been proposed that 1O2 can be generated in chemical reactions (so-called
dark biochemistry) by unstable oxygen adducts, endoperoxides, metal complexes
(101, 102), and peroxyl radical recombinations (103105), and that the low levels
of internal hydroperoxides produced thereby initiate lipid autoxidation chains (102,
106, 107, respectively).
O2 + O2 1
O2 + H2O2 21
O2 + H2O2 OH + OH + 1O
2 22
2 (R2)CHOO (R)2CHOOOOCH(R)2 3(>C=O) + O2 + (R2)CHOH
>C=O + 1O
2
23
INITIATION 285
However, the production of 1O2 in the dark remains highly controversial. Bielski
and Allen, Matsoura et al., and Nilssa and Kearns have shown that this reaction
is highly unlikely both on thermodynamic grounds (108) and because 1O2 is con-
verted to HOO so rapidly by phenols (Reaction 24) (109) that it cannot be detected
(110).
ArOH + 1O
2 ArO + OOH 24
The general consensus remains that if 1O2 production in dark biochemistry occurs,
it is not competitive with other modes of catalysis and cannot be considered an
important initiator of lipid oxidation.
2.1.3. Heat High temperatures (e.g., frying temperatures) have sufficient energy
to break covalent C C or C H bonds in the acyl backbone to form a variety of
lipid alkyl radicals (111, 112), which then start the radical chains of oxidation.
Moderate temperatures have lower energy, so act primarily by breaking O-O bonds
in traces of ROOH or LOOH preformed by other reactions, particularly metals,
lipoxygenase, or photosensitizers. The RO , LO , and OH thus generated abstract
hydrogens from neighboring lipids to form L and initiate radical chains. As shown
by the activation energies for the individual stages of lipid oxidation, LOOH
decomposition and its subsequent contribution to propagation is the major catalytic
effect of heat (113, 114). Effects of increased LOOH decomposition are amplified
by increased rates of subsequent H abstractions by LO and LOO , which is
reflected in the doubling of oxidation rate for every 10 C rise in temperature (115).
Reaction Activation Energies (Ea) kCal/mole

(L O2) 0
kp (LOO LH)
515
kt (2 ROO )
4
kt (2 R ) 5
kt (R ROO ) 1
* ki (monomolecular) 31
* kii (bimolecular) 50 uncatalyzed system
2.1.4. Lipoxygenase Lipoxygenases catalyze the aerobic oxidation of fatty

acids with cis-nonconjugated pentadiene structures to generate optically active
conjugated LOOH without releasing a lipid free radical. Hydroperoxides are
synthesized in a cage reaction involving electron transfer to the lipid from the fer-
rous iron atom in the enzymes active site (116) and removal of the bisallylic hydro-
gen as the rate determining step (117119). Oxygen bound to a separate site on the
enzyme is activated to react with the free radical, then H donation from the
enzyme completes the LOOH before it is released. As the oxygen always adds
anti to the hydrogen removal, the resulting conjugated dienes are always trans-,
cis-relative to the hydroperoxide (117).
R1 R2
LH H+
O2
E (Fe3+) E Fe2+...L E (Fe2+)...LOO E (Fe3+)...LOO
(E Fe)...L...O2 (H+)
Ternary complex
HOO R2 OOH
R1 R1 R2
E (Fe3+) + LOOH
Feedback inhibition by high [LOOH] 25
Radical oxidation chains are initiated when LOOH is decomposed to initiating

LO and OH radicals by light and heat, to LO /LOO by metals, or to LO by the
enzyme itself (120). Very low levels produced in plant or animal tissues may pro-
vide the invisible initiators that make lipid oxidation sometimes appear sponta-
neous. Perhaps just as important, LOOH produced by lipoxygenase can accumulate
to relatively high levels under appropriate conditions (e.g., cold and dark, as in fro-
zen unblanched materials), then lead to a cascade of rapid oxidation when LOOH
decomposes.
It should be noted that although oxygen is not required for formation of the
bisallylic radical, it is necessary for formation of high yields of hydroperoxides.
Hence, when lipoxygenase is being used to synthesize lipid hydroperoxides, full
oxygenation must be ensured, and conversely, when lipoxygenase action needs to
be inhibited without thermal inactivation, reduced oxygen pressures offer an excel-
lent means of control.
2.1.5. Heme Proteins and Porphyrins Heme catalysis of lipid oxidation was
first reported in 1924 (121), but it was another 30 years before research to determine
mechanisms and effects began in earnest. In pioneering studies, Watts and Chang
(148) observed that ferric hematin forms were the most active catalysts and pro-
posed a fundamental electron transfer mechanism (122125). A few years later,
Tappels work in model systems suggested that hemes form complexes with pre-
formed hydroperoxides, and radicals are generated in subsequent decomposition
of the complex (126132). Love and Pearson (133) then proposed that free inorgan-
ic iron released from hemes, rather than the hemes themselves, catalyzed lipid oxi-
dation in meats. However, this theory was inconsistent with earlier observations that
hemes were more effective catalysts than free iron, and questions were further
raised when Fe3-hematin complexes were more active in model emulsions than
FeSO4 and FeCl3 (134). Although all these theories address some behaviors of
INITIATION 287
heme systems, none of them completely accounts for the kinetics, product mixes,
and solvent effects of heme catalysis (99).
That heme compounds catalyze lipid oxidation in food and biological systems
has been extensively documented (128, 135140), but how this occurs is still not
clear. The greatest obstacle for unraveling heme catalysis in foods is the compli-
cated composition and structure of the reaction system. The kind of compartmen-
talization that isolates heme proteins in living tissues may or may not be retained
after food processing, the cellular chemistry maintaining redox balance begins to
decline immediately after slaughter or harvest, and previously protected sites
become exposed. Under these conditions, overall measures of increased lipid
oxidation can be obtained, but it is exceedingly difficult to determine details of
reaction mechanisms.
Application of data obtained from simple clean reaction systems in biological
or chemical studies of heme catalysis also has its problems. Chemical model sys-
tems use chelators, model hemes, and substrate structures that are quite different
from those existing in foods. Reaction sequences change with heme, substrate, sol-
vent, and reaction conditions. Intermediates are often difficult to detect (141), and
derivations of mechanisms by measuring products and product distributions down-
stream can lead to erroneous or incomplete conclusions. It is no surprise, then, that
there remains considerable controversy over heme catalysis mechanisms. Further-
more, mechanisms determined in these defined model systems with reaction times
of seconds to minutes may or may not be relevant to lipid oxidation being measured
in the complex matrices of foods stored for days or weeks under conditions where
phospholipids, fatty acid composition, heme state, and postmortem chemistry com-
plicate the oxidation once it is started (142). Hence, the mechanisms outlined below
should be viewed as guides rather than absolutes. More research should be focused
on determining, by kinetic and product analyses, which reactions actually occur and
are of practical importance in specific food systems.
Current evidence indicates that hypervalent iron complexesferryl iron (FeIV,
FeO22, Fe(IV)=O) or perferryl iron (FeV)are involved in the catalytic mechan-
ism, but there is still controversy over the details of reaction mechanisms and what
proportion of heme catalysis it accounts for. Very recently, some very elegant chem-
istry has elucidated binding and O O bond scission mechanisms and identified
heme structural elements critical for oxidation catalysis (143, 144). Paradoxically,
although the early theories of heme catalysis have been largely dismissed, they
nevertheless are consistent with aspects of hypervalent iron behavior. Ferryl iron
chemistry encompasses and explains the most important features noted in early
studies (99):
1. The porphyrin-Fe structure is an absolute requirement for catalysis.

2. Catalytic activity varies tremendously among heme proteins, partly due to
exposure/or accessibility of the hematin structure and partly due to other
unidentified factors.
3. Fe3-hemes are most active even without oxygen; Fe2-hemes require

oxygen for catalysis.
4. No change of heme iron valence seems to be involved.
5. There is a solvent-related pH dependence that varies with the specific reaction
components and conditions (e.g., liposomes vs. membranes vs. emulsions,
fatty acids vs. phospholipids, buffer type, heme compound).
6. Catalysis reverses to inhibition at high heme levels.
The simplified reaction scheme given in Reactions 2630 is a synthesis from

several authors (139, 141, 143158) and together with Figure 3 provide a general
representation of current understanding of heme-mediated formation and reaction
of Fe4. All iron is complexed to a porphyrin, P, and has hydration and hydroxyl
ligands; F is a scission fragment; and radicals capable of initiating lipid oxidation
are noted in bold type. An extensive discussion of heme catalysis with major
emphasis on foods is available in a recent review by Baron and Anderson (159).
H
O R
O
FeIII Porph + ROOH FeIII Porph
(When protein has internal H source)

heterolysis homolysis
O e OH
ROH + FeIV Porph FeIV Porph + RO
ROOH ROO
CH CH CH2 CH CH CH CH CH2 CH CH
OOR
ROOOOR
epoxides
CH CH CH CH CH
RO + O2
LH
FREE RADICAL CHAIN REACTIONS
Figure 3. Heme-catalyzed formation of species that can initiate lipid oxidation: generation and
reaction of ferryl iron complexes [FeIV O, FeIV(OH)]. Adapted (143, 160); used with permission.
INITIATION 289
heterolysis (P)FeIII(OH) (H2O) + ROOH (P + )FeIV O + RO + 2 H2O 26
H abstraction ROH + (P)

Internal cyclization R epoxide
26a
(P + )FeIV O + H2O2 HPFeIV O + HOO 27
(P + )FeIV O + LOOH HPFeIV O + LOO 28
homolysis (P)FeIII(OH) (H2O) + ROOH (P)Fe IV(OH)
+ RO + 2 H2O 29
-scission F + aldehydes/ketones
H abstraction ROH + X O
O2
internal cyclization R epoxide 29a
O O
OO O
(ROO )
2 ROO [ROOOOR] 2 RO + O2 30
For hemes to be more effective initiators than Fe3 and Fe2, either removal of
an electron from the double bond, reduction of preformed hydroperoxides to gen-
erate L or LO , or both of these reactions must be activated, or another mechanism
entirely must be operative. Model system studies have now shown that the basic
activating reaction involves binding of preformed hydroperoxides, either H2O2 or
LOOH, to ferric hemes to generate hypervalent Fe in a very fast reaction
(k
109). In the concerted process, the negatively charged porphyrin ligand releases
H2O and weakens the O O bond, the hydroperoxide is decomposed heterolytically
(Reaction 26; left reaction series, Figure 3) to produce an alcohol, or homolytically
(Reaction 27; right reaction series, Figure 3) to produce alkoxyl radicals, respec-
tively, and an O is transferred to the iron to form the ferryl complex, Fe4O.
This reaction is very sensitive to environment, particularly solvent and proton avail-
ability; and the OO scission mode and products vary with the heme, hydroperox-
ide structure, solvent, and reaction environment. Heterolytic scission results in one
of the oxidizing equivalents being transferred to the porphyrin apoprotein, forming
a free radical that localizes on tyrosine (161163) or tryptophan (164). This radical
can be quenched by H abstraction from hydroperoxides, producing peroxyl radicals
(30, 31).
In protective heme enzymes such as catalase and peroxidases, the dominant
process is heterolytic, and amino acids such as histidine in the apoprotein are in
close proximity in the active site to transfer protons to the RO in situ. For
ROOH LOOH, a lipid alcohol is released and no initiation or branching can

occur. In aprotic solvents or acid environments, however, H abstraction is delayed
and the radicals remain active. When the heme is myoglobin, hemoglobin, or a
heme protein where an internal proton source is not available, the reaction mechan-
ism is more likely to be homolytic, yielding alkoxyl radicals with no radical on the
porphyrin.
There is disagreement about the fate of the RO , R0 , and F radicals, and it is
even less clear which species initiate new lipid oxidation chains. Most obviously,
any of these would appear to be potential initiating radicals for lipid oxidation, but
perhaps not directly. Unless a proton source is immediately available, there is a
strong driving force for the bound LO to cyclize internally to epoxides, at the
same time generating epoxyallylic radicals (Reactions 26a and 29a) that are
more stable than alkoxyl radicals (see Section 3.2.2). Indeed, most model system
studies of heme catalysis have found that cyclization dominates overwhelmingly
with fatty acids (147, 148, 156). The peroxyl radicals formed by oxygen addition
to epoxyallylic sites (Reaction 29a) are slow and specific in reaction, giving them a
much better chance of escaping the heme complex reaction cage to react elsewhere.
Even so, the slowness of peroxyl radical reactions also argues against their initiat-
ing lipid reactions much faster than normal autoxidation, so it is much more likely
that the peroxyl radicals recombine outside the reaction cage (but still inside the
heme crevice) and dismutate to LO radicals, which react much more rapidly.
Although the distinctions between heterolytic and homolytic pathways may be
important for enzymes in vivo and may also provide some support for Tappels
theory of lipid hydroperoxide decomposition, what happens to the activating
ROOH (HOOH or LOOH) is inconsequential to lipid oxidation in foods. HO
from H2O2 may be diffusible and highly reactive, but it does not initiate lipid oxi-
dation (144). Subsequent abstraction reactions of LO dominate with Hb (165) and
Myb (166), giving 9-LOOH as in autoxidation. Increased decomposition of H2O2 or
lipid hydroperoxides alone cannot account for the explosive oxidation that can
occur in the presence of hemes because (1) catalysis rate would then be directly
proportional to the hydroperoxide concentration, and (2) final rates would approx-
imate those controlled by LO in secondary stages of autoxidation. Neither condi-
tion seems to fit existing data. Something more is needed to connect these reactions
to active, accelerated lipid oxidation.
One missing link was provided in observations that myoglobin-H2O2 catalysis of
linoleic acid oxidation gives highly regio- and stereospecific hydroperoxides,
almost exclusively 9S-OOH (144, 167), which indicated some type of specific fatty
acid binding to myoglobin. Comparative studies with Myb mutants revealed that
fatty acids bind at the entrance to the heme pocket (Figure 4). The hydrocarbon
terminus of a fatty acid penetrates into the crevice in the geometry required to
form a trans-10,11-double bond. Abstraction of the pro-R hydrogen generates a pla-
nar 1,4-dienyl radical directed toward the heme ring and protected from oxygen.
Oxygen then adds from the opposite direction (from outside the crevice), and
the 9-OOH forms preferentially because it is exposed, while C-13 is inside the
crevice.
INITIATION 291
M M M M
V P V P O2
A D
HS
HR
B C CO2H CO2H
M P M P
V M V M
O O
HS
M M
OOH
V P
HR
H
O
HS
FeIV
CO2H
M P
V M
Figure 4. Model proposed for the binding and oxidation of polyunsaturated fatty acids in the
myoglobin heme crevice. From (144), used with permission.
A second missing link is that the critical driver responsible for the dramatically
increased lipid oxidation rate is the Fe4 itself, not radicals from the decomposition
of contaminating hydroperoxides. Ferryl iron is a strong oxidant, kinetically
equivalent to HO in reactivity (154) but more selective due to its lower redox
potential (168). Ferryl iron rapidly abstracts H from the doubly allylic C-11 of
linoleate (now conveniently oriented toward the heme iron core) (144) and it
abstracts hydrogens from hydroperoxides even more rapidly (154), in contrast to
the very slow oxidation with nonheme Fe3:
(P + )Fe4+(O) + LOOH LOO + Fe3+OH + H2O 31
This has two consequences: (1) most importantly, direct initiation of radicals
in lipids bound to the heme, and (2) assurance of lipid release as LOO rather
than LOOH. Chain propagation may proceed through LOO directly or through
epoxyallylic peroxyl radicals from LOO cyclization.
A third missing link important for rapid catalysis was recognition that once
formed, Fe4 states could be maintained by electron transfers to the apoprotein
without involving the iron center. This is shown as the reversible reaction
(P + )Fe4+(O) (P)Fe4+(OH) 32
in Figure 3. Thus, electrons can be shuttled facilely between two reactive states
without the loss of oxidizing power and reduction of Fe4 to less reactive Fe3
(160). Together, these three factors provide a powerful system for extremely effec-
tive catalysis of lipid oxidation.
Fe2-hemes also generate ferryl complexes, albeit more slowly, and this oxidant
source may be important over longer reaction times or during storage. With H2O2 as
the oxygen source, Fe2-myoglobin catalysis of lipid oxidation is initially slower
but eventually reaches the same rate as Fe3-myoglobin (169). However, peroxides
are not always absolute requirements. Direct (slow) reaction of heme-Fe2 with
oxygen,
(P)Fe2+(H2O) + O2 Fe4+(O) + H2O 33
may not be detected as activity in short-term assays designed to mimic enzyme

active sites, but nevertheless may provide a nonhydroperoxide source replenishing
Fe4(O) in longer reactions. Fe2 hemes may also contribute to delayed catalysis
by regenerating Fe3 hemes after some peroxyl radicals have been formed (2, 149,
152),
ROO + Mb(FeII)O2 ROOH + Mb(FeIII) + O2 34
Variable catalytic activity between different heme proteins (137, 170, 171) and
between the same hemes from different species (172) has long been recognized.
The recent elucidation of the fatty acid binding (144) and clarification of O O
bond cleavage mechanisms by ferryl complexes (143, 173, 174) provide insights
into why this happens. The composition and arrangements of amino acids in the
heme crevice, as well as heme pocket size and orientation, affect lipid binding
and proton transfer, while the heme structure and ligands influence electron transfer
processes and stabilization of the ferryl complex. Attainment and stabilization of
Fe4(O) long enough for reaction requires both appropriate adjustment of the
heme redox potential and steric shielding of the bound oxygen at a fixed coordina-
tion position on the iron. Small perturbations in the active site deactivate oxygen
and lead to its release as O
2 /HO2 (149), which are not very reactive with lipids.
All of these factors and the reaction environment influence whether O O bond
cleavage is homolytic or heterolytic, pro-oxidant or antioxidant, under given con-
ditions (143). Considering this new information on lipid binding and mechanisms
of ferryl iron formation, it should now be straightforward to interpret, model,
and even predict catalytic activity based on individual heme protein and ligand
structures.
Similarly, this new information provides explanations for the shift from pro-oxi-
dant to antioxidant at high heme concentrations that has long been recognized (123,
175177). High heme concentrations increase heme association and limit fatty acid
access to the heme pocket (177). Under low oxygen conditions or when oxygen has
been depleted by reaction, excess ferrous hemes oxidize instead by combination
with reactive ferryl complexes, reducing them to ferric complexes (Reaction 35)
with lower reactivity. High heme concentrations oxidize the radicals generated dur-
ing formation of Fe4(O) (Reaction 36), or reduce them if the hemes are ferrous, so
INITIATION 293
no subsequent reactions can occur. Any alkoxyl radicals produced in ferryl forma-
tion, although kinetically inconsequential at low concentrations, become competi-
tive at high heme concentrations and can convert the hydroperoxides being
generated to alcohols and peroxyl radicals (Reaction 40)a net reduction in pro-
pagation capacity.
(P)Fe2+(H2O) + (P + )Fe4+(O) 2 (P)Fe3+(OH) (H2O) 35
(P)Fe3+(OH) (H2O) + R (or F ) (P)Fe2+ (H2O) + ROH (FOH) 36
RO + ROOH ROH + ROO

37
(fast reactions) (slow reactions)
Whether the porphyrin apoprotein radical shown in reactions above has a role in
catalyzing lipid in oxidation is still being debated. Current evidence suggests that
the heme protein radical is required for electron transfer in the ferryl iron-heme
complexes (157) and that it may co-oxidize proteins or other molecules (163,
178), but is probably not involved in direct catalysis of lipid oxidation (144).
For food applications, another mechanism must also be considered as a possible
minor contributor. Considering the photosensitization capabilities of porphyrin
rings in chlorophyll, Schaich (99) questioned whether analogous reactions could
be catalyzed by hemes in foods in which normal molecular and cell environments
are disrupted and porphyrin rings can become exposed. This possibility has
now been verified by EPR spin trapping evidence that hematin, but not intact
heme proteins, produce 1O2 (179), and in observations that protoporphyrin IX cat-
alyzed oxidation of rat liver microsomes only in the light, whereas in the dark it
inhibits lipid oxidation (180). Photosensitization, which can only occur at the sur-
face, would not be expected to compete with ferryl iron produced by intact hemes
in the interior of muscle foods before cooking, but it may indeed contribute to oxi-
dation in processed foods in which some disintegration of the heme complexes
occurs.
2.1.6. Ozone The reactivity of ozone with unsaturated fatty acids has long been
recognized, and indeed, the reaction has practical applications in localization of
double bonds (181). As a damage reaction, atmospheric ozone (O3) [e.g., from pol-
lution or sterilization processes (182)] rapidly adds across double bonds in nearly
all organic molecules to form ozonides (trioxides), which then undergo a number of
different subsequent reactions, not all of which produce free radicals. However,
there remains some controversy over whether direct or indirect mechanisms
dominate.
Ozone adds directly to double bonds in fatty acids to form ozonides (183185).
These decompose to lipid alkoxyl and peroxyl radicals that abstract hydrogens to
initiate radical chains (186). In the process, internal rearrangements within the
original lipid molecule(s) yield hydroxy epoxides and hydroxy epidioxides with
1,3- and 1,4-cyclic hydroperoxides:
R R
Initial products
R R R R R R
O3
O
HO HO O O HO O O
R R R R R R O2
38
O O O OO O OOH
O
M+, uv
LOO
LH
O + OH LH
LH
O2
Radical chain LOO L L L
Indirect initiation of lipid oxidation by ozone is similar except that it occurs via
decomposition of ozonides in non-lipid molecules to form alkoxyl and peroxyl radi-
cals that subsequently abstract hydrogens from fatty acids. Two mechanisms have
been proposed, both of which yield the same final lipid products (186):
R CH CH R + O3
nonlipid molecule
O OO O OOH O O + OH
RCH CHR RC CHR RCH CHR
39
PUFA
O O O
O O
C C RC O +
H CHR
O2
PUFA OO
INITIATION 295
Ozone preferentially reacts with the most unsaturated fatty acids present (187);
arachidonic acid and higher PUFAs are particularly sensitive. Trans-double bonds
and fatty acids have been reported to react with ozone much more slowly than cis-
double bonds (21), but this observation may be an artifact of measuring only initial
ozonides. In fact, trans-fatty acids do react with ozone, but the initial ozonides
decompose and rearrange more rapidly to generate peroxy-epoxide or peroxy-ozo-
nide complexes and free acids (188). This is another example of how, as in lipid
oxidation itself, downstream as well as initial products must be measured to obtain
a full and accurate picture of reaction.
Ozone reactions are not very fast (k
105) and do not change the rate or product
mix of lipid autoxidation once established (189). Nevertheless, ozone markedly
shortens induction periods by contributing to early accumulation of the critical
concentration of lipid radicals and hydroperoxides necessary to trigger the onset
of rapid oxidation. Ozone also reacts with LOOH to produce radicals that propagate
the oxidation chain:
LOOH + O3 LOO + HO + O2 40
Whichever initial reaction occurs with ozone, once active oxidation equilibrium is
established, LOO and LO propagation reactions dominate and effects of ozone on
oxidation rates and product mixes becomes insignificant (190).
2.1.7. Free Radicals In the discussion above, all the initiating processes gener-
ate some form of radical that ultimately reacts with lipids to produce the ab initio
lipid radical that starts the autoxidation chain. The kinetics of the initiation, how-
ever, are governed by the speed of individual radical reactions with lipids, which
can vary tremendously. Table 2 lists rate constants for a number of reactions impor-
tant in initiation of lipid oxidation. For the most part, the rate constants speak for
themselves. Nevertheless, a few comments need to be added.
Not surprisingly, hydroxyl radicals have the fastest reaction rates with lipids.
However, HO are so strongly oxidizing that their reactions are also very nonspe-
cific, and they attack lipids indiscriminantly at all sites along acyl chains (195, 207).
These radicals then migrate (by intramolecular abstraction) to the doubly allylic
Hs in dilute monomer solutions, or abstract Hs from doubly allylic sites of neigh-
boring lipids in concentrated solutions, yielding the dienyl radicals that, when
oxygenated to LOO , become the main chain carriers.
It is important to note that saturated fatty acids are not immune to effects of
oxidation. The strongly oxidizing radicals HO and RO abstract hydrogens at rea-
sonable rates even from saturated fatty acids (106 for RO and 109 for HO ). The
subsequent LsatOO radicals then abstract hydrogens from neighboring unsaturated
fatty acids and thus can be sources of external radicals initiating radical chains in
PUFAs (9, 208).
Values for ROO are average rates for all organic peroxyl radicals; peroxyl radi-
cal rate constants vary little with R structure unless there is a halogen atom a to the
radical peroxyl group (9). Although O
2 has been invoked as an initiator of lipid
TABLE 2. Lifetimes and Hydrogen Abstraction Rates of Various Radicals that Initiate
Lipid Oxidation.
Half-life with Typical Substrate, Ave. rx Rate,

Radical 103 M, 37 C k (L mol1sec1) Reference
9 9 10
HO 10 sec 10 10 191
RO 106 sec 106108 191
ROO 10 sec 101103 191
L 108 sec 104108 191
AnOO 105 sec 192a
O2
1 193b
HOO 100103 194b
18:1 18:2 18:3 20:4

9 9 9
HO
10 9.0 10 7.3 10
1010 9,195
Monomer 8.0 109 8.0 109 196
Micellar 1.3 109 2.5 109 195
Non-allylic H 4 102 3.4 103 7.0 103 1.0 104 196
RO 3.3 106 8.8 106 1.3 107 2.0 107 9
t-BuO 3.8 106 9.1 106 1.3 107 2.1 107 197
(trans) 3.3 106 (trans) 8.8 106 197
aqueous 6.8 107 1.3 108 1.6 108 1.8 108 198
ROO 1.1 6 101 1.2 102 1.8 102 199201
O2 no rx no rx <1 <1 196, 200
(MLOOH) 7.4 103 202
HOO no rx. 1.1 103 1.7 103 3.1 103 193
<3 102 200
O3
CCl4 6.4 105 6.9 105 203

aq SDS 9.5 105 1.1 106 203
SO3 1.8 106 2.8 106 3.9 106 194
GS <2 10 6
8 106 1.9 107 3.1 107 204
1
O2 0.74 105 1.3 105 1.9 105 2.4 105 205
O 7.5 102 9.7 103 1.2 104 1.9 104 196
NO2 1.2 106 6.2 106 6.6 106 206
a
Aqueous solution.
b
H abstraction from unsaturated alkenes.
oxidation, the rate constants in Table 2 show clearly that O

2 does not react with

unsaturated fatty acids or their hydroperoxides. O2 is a weak reactant, both as a
reducing and oxidizing agent (E 0:33 V for O2/O
2 and 0.94 V for O2 /

H2O2) (209). In lipid oxidation, O2 is probably most active in recycling traces
of contaminating metals, particularly iron, or as a source of highly reactive HO ,
which very rapidly take over reactions, obscuring initial effects of O
2 . This has
been demonstrated in reactions of Fe-EDTA complexes with linolenic acid (210).
However, the conjugate acid, HO2 , abstracts doubly allylic H atoms of linoleic,
linolenic, and arachidonic acids (211). At pH 7.0, only about 1% of O
2 solutions

is present as HO2 , but the latter drives any radical abstractions. In acid solution,
only HO2 is active.
2.1.7.1. Radicals from Secondary Reactions One area of initiation that has gone
totally unnoticed is reaction of radicals produced in solvents or other system
INITIATION 297
components, which then react with lipids. Whether the primary initiator is heat,
radiation, or metals, many of the initial oxygen radicals produced react more
rapidly with solvent components than with lipids. For example, HO react with
alcohols (e.g., used as solvents in model systems) at rates as high as 1012 L
mol1s1 (212, 213), and the alcohol radicals then react with lipids. Decomposition
of MLOOH in 80% ethanol, for example, yields >7% ethoxylated products (214),
and more than 60% of products from photolysis of MLOOH in methanol were
methoxylated (215). Radicals induced in cyclohexane by photolysis also react
with MLOOH (216). Similar co-oxidation occurs with Triton-X as an emulsifier
(217). Tris, phosphate, and other buffer components form radicals that can be
detected by EPR, and EDTA forms several radicals that are strongly reducing in
nature (218220). The role these system radicals may play in overall lipid
oxidation is not yet known, but their possible involvement should be considered
in designing test systems and calculating and interpreting oxidation kinetics.
2.2. Sites of Radical Initiation by Hydrogen Abstraction

and Formation of Peroxyl Radicals
Hydrogen abstraction by free radicals is generally quite specific, occurring prefer-
entially at allylic hydrogens where the C H bond energies are lowest (Table 3).
The order of reactivity is doubly allylic Hs between two double bonds > singly
allylic Hs next to double bonds Hs a to the COOH group > Hs on methylene
groups farther down the acyl chains. The one exception to this rule is the hydro-
xyl radical, HO , which is so electrophilic and reactive that it abstracts Hs indis-
criminantly from all positions along the acyl chain (195). The radicals formed
either migrate to the acyl carbon with the weakest bonding, i.e., the allylic Hs,
or abstract allylic hydrogens from a neighboring lipid molecule.
Older literature always presents the initial radicals in equivalent resonant posi-
tions with equal probability of forming hydroperoxides. The three resonant posi-
tions for linoleic acid or its ester and the three hydroperoxides resulting
from these are shown in Reaction 41 (224). Comparable resonant structures have
been published for oleate, linolenate, and arachidonate (222, 225).
TABLE 3. Bond Energies of Hydrogens at Various Positions in Acyl Chains (bold font):
Sites of Preferential Hydrogen Abstraction.
E (kJ/mol) E (kcal/mol)a Relative Ease of H Abstractionb
CHCH2
H 431 105
CH2
H CH2
CH3 419 99
CH2
H CH
CH2 356 85
HCH
R CHCHCH2
CH3 322 77
R(CH2CH)
HCHCH2
310 74 1
CHCH
R HCHCHCH 272 65 62
ROOH 377 90
a
See (221, 222).
b
See (223).
CH3 (CH2)4 CH CH CH2 CH CH CH2 (CH2)6COOCH3
R1 R2
R1 CH CH CH CH CH R2
41
OOH
OOH
OOH
Product distributions show the inaccuracy of this notion. It is now recognized

that after hydrogen abstraction from the allylic hydrogens, the free electron becomes
distributed across a resonance stabilized double bond system (Reaction 42). The
highest electron density is in the center, and the outside positions are relatively
electron deficient. Thus, oxygen preferentially adds at the outermost points. When
oxygen pressures are greater than 100 mm Hg, the addition occurs at diffusion-
controlled rates (k > 109 L M1sec1) (223), so is essentially instantaneous as
long as oxygen is availableone reason L radicals are so difficult to detect,
even by electron paramagnetic resonance.
O2
H+
42
H H
e deficient points
Translating this into observed behavior, isolated double bonds behave as if there
were two separate resonant systems of equal probability, so oleic acid yields
(C9 C11) and (C8 C10) hydroperoxides from the two resonance systems,
respectively, in approximately equivalent amounts (18:1, Figure 5). In 1,4-diene
systems, H abstraction occurs preferentially at the doubly allylic hydrogen between
the two double bonds, and the resonance system with the unpaired electron extends
across both double bonds with electron density focused at the central carbon (11)
and electron deficient positions at external carbons 9 and 13 (18:2, Figure 5).
In higher polyunsaturated fatty acids with multiple 1,4-diene structures (18:3
and 20:4, Figure 5), the resonant systems from multiple doubly allylic radicals
INITIATION 299
Electron delocalization Hydroperoxide positions

+
11 10 9 8 H
+
18:1
13 12 10 9 H+
18:2
16 15 13 12 10 9 H+
18:3
15 14 12 11 9 8 6 5 H+
18:4
Figure 5. Doubly allylic H abstraction sites, electron resonance distributions, and corresponding
locations of hydroperoxide formation in unsaturated fatty acids. Heavy arrows denote dominant
positions for hydroperoxide formation.
overlap. In theory, then, hydroperoxides should form at internal positions in equal

proportion to the external positions. Nevertheless, only minor amounts of internal
hydroperoxides are observed, and then only with three or more double bonds,
because they lack the C OO bond stabilization by conjugation and undergo rapid
b-elimination of the oxygen to regenerate the original 1,4-diene radical (226). In
addition, internal peroxides have a very strong tendency toward cyclization (227
230). Consequently, the dominant hydroperoxides of autoxidizing fatty acids are
always found at the external positions, regardless of the number of double bonds,
except under two circumstances; (1) in autoxidation, equal distribution of LOOH at
all positions without any cyclic products is found only in media of high H donating
powere.g., when 35% tocopherol is added (6), and (2) internal hydroperoxides
are characteristic of singlet oxygen photosensitized oxidation, as was discussed in
Section 2.1.2.
Hydroperoxide positional distributions in unsaturated fatty acids undergoing
autoxidation and photosensitized oxidation are presented in Table 4.
Hydroperoxides have geometric as well as positional isomers on lipid chains.
When the hydrogen is abstracted at an allylic carbon, the double bond shifts one
carbon to a position b to the abstraction site, and it reforms in the trans rather
than cis configuration. The trans,cis-conjugated diene structure is retained whether
oxygen adds or not, and provides the first detectable intermediate in lipids during
autoxidation (238).
For a long time, it was thought that the trans,cis-conjugated double bonds iso-
merize to trans,trans as oxidation progresses, so both trans,cis and trans,trans
forms are typically isolated for each hydroperoxide position. Linoleic acid, for
example, forms 9tc, 9tt, 13tc, and 13tt hydroperoxides. It is now known that this
TABLE 4. Hydroperoxide Positional Distributions in Oxidizing Fatty Acids.
5-OOH 6-OOH 8-OOH 9-OOH 10-OOH 11-OOH 12-OOH 13-OOH 14-OOH 15-OOH 16-OOH Ref.
18:19
Autoxidation 26.4 24.2 22.8 26.6 231
Photo-ox 1O2 47.5 52.3 232
Photo-ox Chl* 49.1 50.8 233
Thermal oxidation 25.1 25.1 24.9 24.9 231
18:29,12
Autoxidation 1 51 tr tr 49 1 234
Photo-ox 1O2 31.9 16.7 17 34.5 232
Photo-ox Chl* 30.2 19.8 19.8 30.1 233
18:39,12,15
Autoxidation 33.4 10.1 12.5 43.9 235
Photo-ox 1O2 22.7 12.7 12.0 14.0 13.4 25.3 232
Photo-ox Chl* 21.6 14.3 15.3 15.7 12.0 21.1 233
20:45,8,11,14
Autoxidation 27 7 9 11 6 40 9
Photo-ox 1O2 14.4 4.8 12.9 13.2 14.4 13.3 6.9 20.3 236
22:64,7,10,13,16,19
Autoxidation % of
OOH at C20 (27.1), C17 (7.9), C16 (9.2), C14 (10.8), C13 (8.9), C11 (7.3), C10 (7.3), C8(7.9), C7(7.0), C4(6.5) 237
INITIATION 301
explanation is inaccurate. The fundamental mechanisms underlying cis-trans

isomerization and distributions in lipid hydroperoxides were recently elucidated
by Porter and his colleagues (7, 226, 239). Two critical factors control the process:
reversible b-elimination of oxygen from peroxyl radicals, and availability of strong
hydrogen donors. Reversible addition of oxygen to the pentadienyl system was first
proposed as the major action during the induction period about thirty years ago,
based on kinetic (240), EPR (241), and 17O evidence (242, 243). Since then, Porter
has contributed much new documentation of the phenomenon, but the concept still
does not seem to be recognized widely and incorporated into general schemes of
lipid oxidation.
Data of Porter and colleagues (5, 7, 11, 226, 244257) shows quite conclusively
that both positional and geometric isomerism proceed through the delocalized allyl
radical for oleate (Figure 6) or dienyl radical for linoleate and higher PUFAs
(Figure 7) via alternating removal of the peroxyl oxygen by b-scission, migration
of the free radical, and readdition of the oxygen at a new carbon position or orienta-
tion (5, 7, 11, 226257). There can be interconversion of peroxyl position and
orientation indefinitely as long as the radical is in the manifold. Once the peroxyl
radical is protonated, it becomes fixed as the hydroperoxide, but can return to the
manifold if the LOOH hydrogen is abstracted.
To explain different proportions of trans,cis and trans,trans isomers, Porter
distinguishes thermodynamic and kinetic processes (5, 226, 253). Kinetically,
hydroperoxides will form whenever an abstractable hydrogen atom is available,
but thermodynamically, the system equilibrium moves toward trans,trans isomers
in the absence of good H donors, as in organic solvents (248). The observed iso-
mer mix reflects the balance and competition between these two processes in a
given system. When good H donors are present, the trans,cis isomers kinetically
form first. The H atoms can come from a protic solvent, an antioxidant, a cosub-
strate, or the allylic hydrogens of the fatty acid chains themselves. For oleic and
linoleic acids with only slightly bent chains, trans,cis formation is favored in
oriented systems or at high concentrations that increase interchain contact. Trans,
trans isomers are favored in dilute solutions, aprotic solvents, and at elevated tem-
peratures in which there is less interchain contact and decreased H availability.
With linolenic, arachidonic, and higher polyunsaturated fatty acids, the fatty
acid chains bend back on each other, bringing double bonds and allylic hydrogens
from opposite ends of the chain into proximity with the peroxyl radicals.
When oxidized neat, higher PUFAs thus have an immediate internal H source
and characteristically yield high proportions of trans,cis peroxides (kinetic
products). However, when an H donor is lacking (e.g., low concentrations, aprotic
solvent, elevated temperature), trans,trans cyclic hydroperoxides become domi-
nant (251).
The tendency for higher cis isomers at external hydroperoxides and positions
closer to the COOH terminus reflects the greater H abstracting power of those
positions. Conversely, the increase in trans isomers with internal hydroperoxides
and as the hydroperoxide position moves toward the distal end of the fatty acid
chain reflects depressed activity at those sites. The cis/trans ratio changes with
OO OOH
RH
11 R2 11 R2
R1 R1
11-trans
O2 RH
9 9 R2 9 R2
R1 R2 R1 R1
OO OOH
9-trans
OO OOH
O2 RH
11 11 11
R1 R2 R1 R2 R1 R2
11-cis
11 8
R1 R2
OO HOO
O2 RH
R1 8 R2 R1 8 R2 R1 8 R2
8-cis
O2 RH
10 10 10
R1 R2 R1 R2 R1 R2
OO OOH
10-trans
R1 = C7H15, R2 = (CH2)6COOCH3 RH
8 R2 8 R2
RH = methyl oleate R1 R1
OO HOO
Figure 6. Radical sites and b-elimination manifold leading to isomerization of hydroperoxides
in oleic acid. Adapted from (11).
reaction system and with temperature. Cis isomers are enhanced by the presence of
antioxidants such as tocopherol and by high concentrations of lipids, whereas trans
isomers are enhanced by even mild heating which reduces contact between lipid
and potential H donors. Contrary to earlier reports, the cis/trans ratio does not
vary with extent of oxidation unless reaction conditions are changing or H abstrac-
tion from LOOH is occurring, allowing LOO to undergo b-scission.
OOH R1 OO R1 R2 OO
LH O2 O2 R2 LH HOO
R1 H H R2
R2 R2 (O2) (O2)
R1 R1
trans, trans -9-OOH trans, trans -13-OOH
O2 (O2) (O2) O2
R1 R2
R2 R1
(O2) O2 (O2) O2
R1 R2 R2
LH OO OO LH
R1 H
HOO R2 OOH
R1 R1 R2
trans, cis -13-OOH LH OO R2 OO LH trans, cis -9-OOH

O2 O2
H H
R1 (O2) R1 R2 (O2) R1 R2
R1 R2
Figure 7. Reaction scheme for positional isomerization of double bonds and formation of trans-, trans-hydroperoxides during oxidation of
linoleic acid via reversible b-scission of oxygen. Adapted from (246, 248, 250).
303
3. PROPAGATION
The classic free radical chain depicts propagation as proceeding directly and
entirely by hydrogen abstraction. In reality, however, H abstraction by LOO is
very slow (k 3662 L mol1 sec1) (200, 258) and selective, abstracting only
hydrogens with low bond energy (e.g., doubly allylic CH2 , thiols, phenols)
(259). Consequently, there is plenty of time for alternative reaction pathways to
compete and change the direction of oxidation (260) yielding distinctly different
products at different rates and having significant consequences to the ultimate mix-
ture of products. Addition, cyclization, and scission reactions compete with H
abstraction to reroute LO and generate products and additional radical species.
Ultimately, radicals are always transferred between molecules by hydrogen abstrac-
tions, but the original LOO may not be the propagating radical, and the product
mix is much more complicated than implied by the simple free radical chain. At
least one of the reactions (e transfer) stops rather than propagates the radical
chains.
Multiple mechanisms are well established in radical chemistry and have been
applied to peroxyl and alkoxyl radical reactions in lipid oxidation (6, 7, 261),
although not all rate constants and reaction details are available. Consideration of
the multiple competing pathways discussed below can explain complicated oxida-
tion kinetics, account for complex product mixes, enable more accurate evaluation
of the extent of oxidation, and facilitate design of more effective antioxidant
strategies.
3.1. Chain Propagation by LOO

LOO are the chain carriers in early stages of lipid oxidation. Competing reactions
of LOO include:
a. atom or group transfer (H-abstraction)

b. rearrangement/cyclization
c. addition to double bonds (! crosslinks)
d. disproportionation
e. b-scission
f. recombination
g. e transfer (LOO e ! LOO )
The first four reactions all contribute to chain propagation, although under
different conditions. Disproportionation leads to branching and a shift in kinetics,
and b-scission mediates isomerization, as was described in the previous section.
Recombination (f) and electron transfer (g) terminate radical chains. Electron
transfer is an active antioxidant mechanism that occurs particularly in the presence
PROPAGATION 305
of active redox agents such as metals. The mechanism(s) occurring in any

given system are determined by ease of H abstraction and double bond
structure in the target molecule, solvent, and reaction conditions, particularly
temperature.
3.1.1. Atom Transfer (hydrogen abstraction) by LOO ! Free Radical

Chain Reactions Hydrogen abstraction is the heart of the classic free radical
chain reaction schemes (Figure 1). Peroxyl radicals initially formed at any site
on a fatty acid pass the unpaired electron to adjacent lipid molecules by abstracting
hydrogens from an allylic position or a hydroperoxide, and the process repeats itself
indefinitely until the chain is intercepted.
LOO + LH LOOH + L 43
LOO + LOOH LOOH + LOO 44
H abstraction from dienes by peroxyl radicals (Reaction 43) is slow (k

62 L M1s1) (258) and highly selective for doubly allylic hydrogens (261, 262). H
abstraction from hydroperoxides (Reaction 44) is ten times faster (223). Two fac-
tors govern H abstraction by LOO : (1) relative availability of H sources in solvent
and lipids, and (2) viscosity of medium (263). H abstraction from other lipids (i.e.,
chain propagation) is facilitated in neat lipids and aprotic solvents in which the lipid
allyls are the only source of hydrogens, at high lipid concentrations where fatty
acid chains come in closer contact, and in higher polyunsaturated fatty acids
with multiple bisallylic hydrogens. Hydrogen abstraction is also facilitated in low
viscosity media, whereas chain lengths are greatly shortened in viscous solvents
(11, 264).
On the other hand, when the solvent or other components in the system have H
sources, competitive abstraction from nonlipid sites occurs and the net result is to
quench the radical and interrupt the chain rather than propagate it. Abstraction from
multiple H sources in a system is common, and subsequent oxidation at nonlipid
sites may account for oxygen consumption that exceeds LOOH formation in
many systems.
Hydrogen abstraction also increases at elevated temperature as thermal energy
decreases bond dissociation energy. Typical H abstraction rates for ROO at
room temperature are 1 M1s1, but this increases to 103104 L M1s1 at
65 C (223). For example, in linolenic acid autoxidized neat at room temperature
to PV 1113, products were not quantified, but estimates from intensities of
HPLC peaks gave about 40% LnOOH, 12% dihydroperoxides, 12% hydroperoxy
epidioxides, and 4% epoxides (228). At 40 C, H abstraction occurred more as a
secondary process. Hydroperoxides per se were still the main products, but fewer
were present as mono- and dihydroperoxides (36% total) and more had formed after
cyclization or addition (31%). Data are not available to distinguish whether this
occurred because rates of LOO cyclization increase more than H abstraction

with temperature or, alternatively, the increased rates of H abstraction by
Reaction 44 [12.75 109 M1s1 for Reaction 43 vs. 5.6 109 M1s1 for Reac-
tion 44 (261)] forces a shift in the equilibrium balance and cyclization products
accumulate at the expense of hydroperoxides in secondary processes. However,
at 80 C, H abstraction clearly dominated, yielding 84% LOOH and 16% cyclic
products.
Hydroperoxides were the first lipid oxidation products discovered (115), and
eventually a hydroperoxide should result from each alternative pathway, as will
be described below. Thus, there is logic to the common practice of measuring
hydroperoxides to detect early stages of lipid oxidation. Nevertheless, hydroper-
oxides alone do not give an accurate quantitative or qualitative picture of the
extent of lipid oxidation because there is no way to account for either LOOH
decomposition or alternative reactions. Oxidation can be seriously underestimated
and system effects can be misinterpreted when monohydroperoxides are consid-
ered to be the only product and determinations of oxidation extent and kinetics are
based on LOOH concentrations alone. Although many studies have focused on
identifying structures of products, few have actually calculated total product
yields and distributions. The limited data available show clearly that LOOH
is not the only product, even in early stages of lipid oxidation. In some cases,
hydroperoxides may ultimately form only after addition, cyclization, or other
rearrangement; in some systems, conventional monohydroperoxides may not
form at all.
3.1.2. Rearrangement/Cyclization of LOO When abstractable hydrogens are

not immediately available, peroxyl radicals find pairing electrons by adding to
down-chain double bonds, forming cyclic products. The most important internal
rearrangement or cyclization of LOO proceeds by 1,3-addition of the peroxyl radi-
cal to the neighboring cis-double bond, attaching to the b carbon to form a 5-exo
ring and leaving a radical on the g carbon of the double bond (Reaction 45). Addi-
tion of oxygen to the radical generates a second peroxyl radical (new position),
which abstracts a hydrogen from a neighboring lipid molecule to propagate the
chain and form a hydroperoxy epidioxide product (Reaction 45a) (265). 5-exo
cyclization by LOO (k
103 s1) (7) is faster than b-scission of oxygen (27-430
s1) (11) and H abstraction (<1-400 M1s1) (88, 223, 247), so it should be
able to compete as an initial process, especially in fatty acids with three or more
double bonds.
Reaction sequence 45 shows this process at C-13 of linolenic acid for simplicity,
but comparable cyclization also occurs at C-10 in linolenic acid and at C-8, C-9, C-
11, and C-12 in arachidonic acid (252). The cyclic product mixes of oxidized Ln
and An typically show multiple positional and geometric isomers (227, 266). In the
interest of space, the isomerization and racemization that accompanies cyclization
will not be discussed here. The reader is referred to papers by Gardner (6, 267) and
Porter (7, 11, 252, 268) for more details.
PROPAGATION 307
OO O O
R R
OO O O
R
45
L2H
L2
HOO O O
R
45a
Cyclization requires the presence of a cis-double bond homoallylic to a hydro-

peroxide (230, 269), as shown in Reaction 45. In addition, cyclization of peroxyl
radicals at internal positions is considerably faster than secondary oxidations of
hydroperoxides at either external position. About 25% of peroxyl radicals in lino-
lenic acid and 33% of peroxyl radicals in arachidonic acid are internal (Table 4).
Thus, linolenic and arachidonic acids are particularly prone to formation of cyclic
peroxides. These factors together make intramolecular cyclization 46 fold faster
than b-scission in higher polyunsaturated fatty acids (247).
Initial cyclization of LOO via 1,4-addition to the g-carbon of the neighboring
double bonds forces endo cyclization to a 6-oxo ring and is kinetically unfavorable
(k
10 s1) (11, 230). However, both 6-oxo exocyclic peroxides (Reaction 46) and
endoperoxides (Reaction 46a) have been observed as secondary oxidation or rear-
rangement products in arachidonic acid oxidation (252, 270). The acyl chains of
fatty acids with four or more double bonds (An, EPA1, and DHA1) have hairpin-
like configurations, bringing double bonds from opposite ends into close proximity.
Although the 1,3cyclic peroxides are all exo, with these fatty acids there is
increased tendency towards cross-chain addition to form endo peroxides (270)
and toward multiple internal LOO additions to form bicycloperoxides (Reaction
46a) and polyperoxides (Reaction 46b) in prostaglandin-like structures (11, 230,
261, 271, 272). Note that each cyclization produces another radical capable of initi-
ating new oxidation chains!
Without enzymatic catalysis, the endo and bicyclic peroxides (Reaction 46a)
usually account for less than a few percent of arachidonic acid oxidation products,
and the dominant pathways are formation of exo peroxides (Reaction 45) and
polyperoxides (Reaction 46b).
L2 L2H
OO OOH
O2 46
O O O
OO O O O
1
EPA: eicosapentaenoic acid, 20:5o3; DHA: docosahexaenoic acid, 22:6o3.
(CH2)3COOR
C5H11
OO
(CH2)3COOR
O a
a O (CH2)3COOR 46a
O O C5H11
C5H11
OO O O
O2 (CH2)3COOR (CH2)3COOR
O O
O O 46b
C5H11 C5H11
As mentioned earlier, in linolenic acid and higher PUFAs, even with cyclization,
some peroxyl radical may eventually abstract a hydrogen externally to propagate
the oxidation chain. The net result of internal cyclization, however, is a reduction
of the number of molecules oxidized: Two or more moles of oxygen are absorbed
per fatty acid, but only one radical transfer occurs and the chain length is extended
by only one. Hence, although the most highly unsaturated fatty acids are innately
the most oxidizable, paradoxically their oxidation chains may be shorter and their
propagation rates may be lower than for linoleic acid.
Peroxyl radicals of linoleic acid do not undergo cyclization to epidioxides during
autoxidation because the requisite cis-double bondhydroperoxide structure is not
present. However, 1O2 photosensitized oxidations produce hydroperoxides at the
internal 10- and 12- positions in quantities almost as high as at the external 9-
and 13- positions, and internal hydroperoxides do have the required b-cis-double
bond. Hydrogen abstraction from the internal hydroperoxides yields LOO that
undergo cyclization and produce propagating hydroperoxy epidioxide radicals
(Reaction 47) and the corresponding epidioxy- hydroperoxide products (Reaction
47a) in high proportions (232, 273, 274).
OO OO
R2 O O O O
R1 R2 R1 R2 R1 47
L2H
L2
HOO O O
R2 R1 47a
PROPAGATION 309
TABLE 5. Product Distributions in Methyl Linolenate (MLn) Autoxidized Neat at 40 C.

Data from (228).
PV 904 1286
9
9-OOH c/t 27.8%
1:1 ratio >
> MLn (unreacted) 87.9 74.8%
>
>
t/t 24.5 >
> Epoxides 0.2 0.3
=
13-OOH c/t 27.8 Mono-OOHa 3.5 8.4
t/t 20.7 >
> HOO-epidioxideb 3.8 7.7
>
>
>
> Epoxy-HO dienes tr tr
;
Di-OOH 0.9 2.9
Polars 3.7 5.9
a
Mixture of 9-OOH and 13-OOH.
b
All at internal positions.
Availability of hydrogens drives abstraction reaction. Thus, solvent, lipid con-

centration, extent of oxidation, and temperature all play critical roles in shifting
the balance between external hydrogen abstraction and internal addition, i.e., direct
chain propagation vs. internal cyclization. Low oxygen pressures particularly favor
cyclization (275). In aprotic solvents and dilute solutions at room temperature, in
which external abstractable hydrogens are absent or limited, LOO cyclization at
various positions accounts for all the products (266). However, in neat lipids, a
situation that provides both availability and proximity of abstractable hydrogens,
abstraction competes with cyclization to generate mixed products. The apparent
proportions vary with temperature and extent of oxidation. At least two studies
have found about 30% cyclic products at 40 C (228, 265).
Epidioxide-OO radicals are very reactive and are particularly prone to dimeri-
zation with even moderate heat (276). This makes detection of their involvement in
oxidations sometimes difficult. Using the example cited above (228), with contin-
ued oxidation to PV 1286 meq oxygen/kg oil, epidioxides remained a major pro-
duct, but mono- and dihydroperoxides, as well as the polar products, increased
more (Table 5). Dimerization of the hydroperoxy epidioxides accounted for at least
part of the decrease in proportional percentage of epidioxides as well as the
increase in polar dimers. Increased H abstraction rates at elevated temperatures con-
tributed to the higher hydroperoxides, but increased decomposition of hydroperox-
ides to alkoxyl radicals at the elevated temperature also introduces more competing
reactions. Faster H abstractions by LO at external positions lead to increased
propagation rates to form new hydroperoxides, pointing out how even moderate
heat can introduce considerable complication in propagation mechanisms.
3.1.3. Addition of LOO to Double Bonds Peroxyl radicals are quite specific
in their addition preferences, and competition with hydrogen abstraction is gener-
ally unfavorable except under select conditions. The ROO addition becomes com-
petitive when abstractable hydrogens are limited (aprotic solvents, low temperature)
and when there is a double bond that is conjugated, terminal, or 1,1-disubstituted
(277). Hence, simple short-chain alkenes with allylic hydrogens react 80100% by
hydrogen abstraction, whereas alkenes that are conjugated or have radicals at a
terminal double bond with no allylic hydrogens (as in some scission products) react
80100% by addition (278, 279). Longer chain alkenes such as fatty acids give
mixed products.
Factors controlling addition are strength of the bond formed, steric hindrance,
polar effects, and stability of the resulting radical adduct (280). Although addition
reactions of small molecular ROO and RO can be very fast, steric factors and the
large number of reactive CH2 groups on unsaturated fatty acids decrease the ease of
addition reactions in lipids. Conjugation and trans-double bonds in oxidized lipids
counterbalance the steric impediments and enhance peroxyl radical additions (281).
Copper salts catalyze addition reactions of peroxyl radicals (282), which has some
interesting implications in food systems.
Propagation via addition of LOO to double bonds forms both monomer pro-
ducts (epoxides and epidioxides) and dimers or polymers; the propagating species
are peroxyl radicals formed at new positions and alkoxyl radicals released by
b-elimination. In early stages of oxidation, LOO adds to double bonds to form
an initial dimer complex (Reaction 48), which then reacts further to generate
new radicals. The ultimate product depends on the nature of the target double bond.
OOL
48
LOO + R1 CH2 CH CH R2 R1 CH2 CH CH R2
LOO adds to isolated or nonconjugated double bonds, then undergoes 1,3-cycli-

zation to form an epoxide and an allylic radical, eliminating LO in the process.
The allylic radical then adds oxygen to form a new peroxyl radical (Reaction
49). This is a true branching reaction as two new propagating radicals (LO and
epoxyOO ) with increased reactivities are generated from the initial LOO .
O2
R1 CH2 CH CH R2 LO + R1 HC CH CH R2 L2(epoxy)OO
OOL O
49
This presents an interesting analytical quandary. Epoxides are major products of

lipid oxidation and derive from LO cyclization as well as LOO additions (see
Section 3.2.2). Consequently, it may be difficult to determine the mechanism that
is operative in a given reaction system, and indeed, both may contribute. For exam-
ple, Hendry (283) reacted a series of ROO with their parent compounds at 60 C
and found 40% of the products were epoxides. Rate constants of k 20 to
1130 M1sec1 were calculated assuming the reactions were all additions, but at
the elevated temperature of the study, hydrogen abstraction to form the hydroper-
oxides, followed by homolytic scission to alkoxyl radicals, could also have con-
tributed to the yields.
PROPAGATION 311
In later stages of oxidation, the likelihood of LOO addition increases because

doubly allylic hydrogens have been removed during initial reactions and the double
bond system has been shifted to conjugated. This limits competition by hydrogen
abstraction and greatly facilitates LOO addition. Thus, addition reactions of LOO
to conjugated oxidation products (Reaction 50) increase during later stages of oxi-
dation and produce the characteristic polymers and increased viscosity of very oxi-
dized oils (284286). Note that even in advanced oxidation, the polymer product
still retains a propagating free radical, distinguishing this from peroxyl radical
recombinations that yield peroxo dimers without radicals (see Termination reac-
tions). Gardner (6) hypothesized that LOO could add to cis-bonds of unoxidized
linoleic acid as well as conjugated double bonds of products to form the same type
of polymer as in Reaction 48. Whether LH or LOOH is the LOO target for addi-
tion, chain propagation could then continue by three pathways: (1) eliminate an
alkoxyl radical and form epoxides via Reaction 49, or add oxygen to regenerate
peroxyl radicals and then (2) add to another LOOH to continue the polymerization
process, or (3) abstract a hydrogen to propagate the radical chain and form a stable
dihydroperoxide dimer. Elevated temperatures favor polymerization via pathway 2
(287). Pathway 3 could account for the low levels of dimers that have been detected
in early stages of oxidation (288).
LOO additions increase with heat (289), extent of oxidation (290), and solvent
polarity (266). Dimer levels of methyl linolenate autoxidized neat at room tempera-
ture varied from 0.1% to 10.1%, proportional to peroxide values (290). MLn auto-
xidized at 40 C to PV 1062 gave 6.8% dimers; 80% of these were from LOO and
20% were from epidioxide-OO additions. The dimer linkages were mostly
CO O C at lower temperatures, but shifted to C C and C O C as the tem-
perature increased (276). At PV 4002, LOO additions increased to 55% of the
products. Epidioxide peroxyl radicals, in particular, showed a very strong tendency
to add to double bonds, with greater than 90% dimerization at 40 C.
OO OOH O O OOH
+
R1 R
R2
(1) O2
HOO OO OOR2
HOO O R1
R 2O + (2)
R2OOH R R1
R
LH
R1 R (3) L
O HOO HOO HOO OOR2

HOO O OOR2
R R1
R R1
50
3.1.4. Disproportionation (radical self-recombination) of LOO Peroxyl

radical recombination is usually written as a termination reaction generating two
nonradical products (Reaction 51a, see also Section 4). However, some mechanisms
are more appropriately considered as propagations because new alkoxyl radicals
rather than stable products are formed (Reaction 51). The following reaction
has been found to occur generally with peroxyl radicals (291, 292). The rate of dis-
proportionation varies with the nature of the alkyl groups, R, but the mechanism is
not altered (293). With t-butyl peroxyl radicals, dismutation to alkoxyl radicals
(Reaction 51) is twenty five times faster than peroxide formation via oxygen
elimination in a cage reaction (Reaction 51a) (291). Similar dominance of RO pro-
duction has been observed with other peroxyl radical species, and subsequent reac-
tions of RO lead to greatly accelerated oxidation (292, 294). Thus, as sufficient
concentrations of LOO accumulate in later stages of lipid oxidation for dispropor-
tionation to occur with lipids, Reaction 51 may contribute to the very rapid increase
in oxidation rates in the bimolecular rate period, which will be discussed further
below.
R1OO + R2OO [R1OOOOR2] R1O + OOOR2
R1O + O2 + OR2 51
R1OOR2 + O2 51a
(51) 2k = 9.5 103 L M1sec1. (51a) 2k = 3.9 102 L M1sec1.
It must be noted that the propagation Reaction 51 only occurs in lipids oxidized
neat or in aprotic solvents. In polar solvents or aqueous solutions, the preferred
reactions of ROO shifts to b-scission, the rate of ROO decomposition increases
dramatically, and dismutation becomes a termination rather than propagation pro-
cess (207, 295). The rate constant for LOO recombination is (2k 2 107 L
mol1 sec1) in aqueous solution at pH 10.5 (196). In organic solvents, yields of
polar scission products increase with solvent polarity, whereas scavengeable radi-
cals and radical cage products decrease. Increasing solvent viscosity also favors ter-
mination over propagation by increasing radical cage products and decreasing
radical release in the dismutation (296). This reaction will be discussed further
under Termination (Section 4).
3.1.5. b-Scission of LOO Beta-scission in LOO cleaves the C-O bond and
releases O2, leaving an alkyl radical behind. In linoleic acid, the rate of b-scission
is competitive with H abstraction from allylic positions, accounting for its critical
role in isomerization (247), as was discussed in Section 2.2. Perhaps the most
important practical implication of b-scission is the shift in isomer distribution at
elevated temperatures, and this in turn alters the ultimate products. During heating,
13-OOH isomerizes to 9-OOH (Table 6) and the scission product mix correspond-
ingly approaches that of 9-OOH (297).
PROPAGATION 313
TABLE 6. Isomerization of 13-OOH to 9-OOH and Corresponding Shift in Products

During Heating of Linoleic Acid. Data from (297).
% yield
Product Scission Point 9-OOH 13-OOH Pure 9-OOH Pure 13-OOH
Hexanal 13-OOH a 1.7 8.3 1 28

Me octanoate 9-OOH a 5.0 4.3 37 24
2,4decadienal 9-OOH a 20.0 12.5 51 33
Me oxononanoate 9-OOH b 10.5 15.2 12 16
3.2. Propagation by Alkoxyl Radicals, LO Alkoxyl (LO ) radicals are

responsible for propagation of the radical chain during the very rapid oxidation
that ensues after the induction period ends. In the earliest stages of oxidation,
LOO cyclization and addition reactions can proceed before LOOH formation
via H abstraction, but LO can only be generated via LOOH decomposition, so their
reactions become important as secondary events in oxidation. Nevertheless,
because LO react faster than LOO by several orders of magnitude, LO becomes
dominant almost as soon as LOOH breaks down.
There are four major mechanisms for radical chain propagation by alkoxyl radi-
cals. The mechanism dominating in a given system is determined largely by double
bond structure, solvent conditions, and steric factors (21):
a. hydrogen abstraction
b. rearrangements/cyclization
c. addition
d. a-and b-scission (fragmentation)
3.2.1. Hydrogen Abstraction by LO LO abstractions are very fast (k

107-
8 1 1
10 L M s ), but less selective than LOO (198); they abstract both allylic and
bis-allylic hydrogens, whereas LOO abstracts only the latter (261). Allylic hydro-
gens are particularly susceptible to abstraction by sec alkoxyl radicals (21), so the H
abstractions by lipid alkoxyl radicals, as written in the classic free radical chain
(Reaction 52), should be a preferred reaction in lipid oxidation:
R1 CH R2 + LH R1 CH R2 + L
52
O OH
However, cyclization and scission reactions of LO compete with H abstraction and
can often limit the effectiveness of this reaction in chain propagation.
Factors influencing the rates of H abstraction by alkoxyl radicals are H abstract-
ability on target molecules > structure of the alkoxyl radical > solvent system
(298). Hydrogen availability and solvent have critical effects in lipid oxidation;
TABLE 7. Rate Constants for H Abstraction from PUFA

by t-BuO in Various Solvents (197, 198). Reactivity of LO
is Comparable (299).
k 106 L mol1 s1

Fatty Acid Nonpolar Solvent Aqueous Solution
Oleic 3.8 68
Linoleic 8.8 130
Linolenic 13.0 160
Arachidonic 20.5 180
the structure is essentially constant as sec alkoxyl radicals. Hydrogen abstraction

from other lipid chains by LO is most effective in neat lipids in which the lipid
allylic groups are the only source of hydrogens. The relative rates of abstraction
from different fatty acids are approximately proportional to the number of allylic
or doubly allylic hydrogens, as is shown in the reaction rate hierarchy of
O < L < Ln < An in Tables 2 and 7, but beyond that there seems to be little pre-
ference for one bis-allylic position over another (197, 300). Interestingly, the high
susceptibility of allylic hydrogens to abstraction, along with the bent chain config-
urations of polyunsaturated fatty acids, also enhances preferential internal H
abstraction by lipid alkoxyl radicals, leading to competing cyclization and epoxide
formation (301303). This will be discussed in more detail in Section 3.2.3.
Hydrogen abstraction by LO to propagate free radical chains is facile also in
nonpolar aprotic solvents when lipids are at high concentrations. However, at mod-
erate lipid concentrations, H abstraction must compete with internal rearrangements
and scission (304), and at low concentrations it may become insignificant (305).
Surprisingly, the most important effect of LO on propagation may well be via
abstraction of hydrogens from the lipid hydroperoxides as they form (Reaction 53),
thus regenerating LOO and eliminating the need for other catalysts to decompose
the hydroperoxides and begin chain branching.
LO + LOOH LOH + LOO 53
The bond dissociation energy of the hydroperoxide hydrogen is higher than the
allylic hydrogens (90 vs. 6585 kCal mol1, respectively), but hydrogen bonding
between the LO and LOOH greatly decreases the Ea for the abstraction (306).
In mixtures of fatty acids and their hydroperoxides, t-butoxyl radicals abstract
hydrogens almost exclusively from the hydroperoxides (230). The rate constant
for (t-BuO ROOH) is 2.5 108 M1s1, nearly diffusion controlled (307). Simi-
larly, cumylalkoxyl radicals abstract H from hydroperoxides faster than reported for
alkyl substrates (306).
In aqueous and protic solvents where H sources are plentiful, hydrogen abstrac-
tions by LO are faster kinetically, but less effective in chain propagation (Table 7).
Production of LOH can be detected in protic solvents (308), but the yields of hydro-
xylated products remain low because selectivity of H abstraction decreases and H
PROPAGATION 315
abstraction must compete with increasing rates of b-scission (309) (see Section
3.2.4). The availability of hydrogens from water and other dissolved solutes
increases the likelihood of H abstraction from molecules other than fatty acids
(310), in which case the chain reaction is not propagated (2, 9). H abstraction as
a termination reaction will be discussed further in Section 4.
The rate of H abstraction by RO increases with temperature in all solvents. This
leads to marked acceleration of oxidation in neat lipids and in nonpolar solvents
where the only H sources are fatty acids, and it also favors LOOH formation
over cyclization. This is evident in the marked increase of mono-, di-, and trihydro-
peroxides over epoxides as oxidation temperature increases from room temperature
to about 80 C (228, 276, 311, 312). However, heat has less effect in polar and aqu-
eous solvents (310). The activation energy for H abstraction is lower than for
b-scission, so there is less thermal enhancement of abstraction rate and also less
selectivity of abstraction sites in polar solvents. More importantly, higher tempera-
tures enhance scission more than abstractions so, particularly at T > 100 C, the
relative importance of H abstraction by LO and LOO in propagation is dimin-
ished (278, 313) and secondary processes begin to dominate.
One additional H abstraction reaction must be mentioned. Internal 1,5 (Reaction
54) or 1,6 (Reaction 55) hydrogen abstraction generates an alcohol and a radical
(21) in a position that may or may not be normal for autoxidation. Intramolecular
H abstraction involving a six-membered transition state (Reaction 55) has been
identified in saturated alkyls with long side chains (304). Occurrence of the corre-
sponding reaction in unsaturated fatty acids would produce oxidation at sites pre-
viously attributed to HO attack (314).
Oleic acid:
O
RHC H O CHR1 RHC HO CHR1 O2
RCH2CH2CH CH CHR1
OO OH
RCHCH2CH CH CHR1 54
Linoleic acid:
O
RHC H O CHR1 RHC HOCHR1
RCH2CH CH CH CH CHR1
OH OH
RCHCH CH CH CH CHR1 55
3.2.2. Rearrangement/Cyclization of LO Cyclization of LO involves 1,2

addition to an adjacent double bond to form epoxides and epoxyallylic radicals
(Reaction 56).
O O
56
R1 HCH CH CH CH CH R2 R1 HCH CH CH CH CH R2
This is a very fast reaction that, under some conditions, can even exceed rates of
H abstraction (309). Cyclization of LO to epoxides is the dominant reaction in
aprotic solvents (including neat lipids), when lipids are at low concentration
(275) or highly dispersed on a surface (315, 316), at room temperature (147,
308, 317), and at low oxygen pressures (275, 278); and the reaction accelerates
with increasing polarity of the aprotic solvent (308310). However, the stability
of LO is reduced considerably in polar solvents (309, 310). Although epoxyallylic
radicals from cyclization have been observed in pulse radiolysis studies of LO in
aqueous solutions (308), H abstraction and scission reactions are much faster. This
pattern can be seen in the change of cyclic products yields when oxidation was con-
ducted in different solvents (Table 8). The change in competition over time is also
apparent.
Cyclization of LO is stereospecific. The configuration of epoxides is fixed by
the conformation of the fatty acid alkoxyl radical at the point of cyclization rather
than postcyclization isomerization (319, 320). As with LOO , there is a stronger
tendency for LO to cyclize from internal positions, probably due to the orientation
of the -O relative to the bis-allylic hydrogens, and consequently, photosensitized
oxidations yield high concentrations of cyclic products (321). The levels and posi-
tional distribution of these products are characteristic markers distinguishing auto-
xidation from photosensitized oxidation.
Temperature has relatively little effect on cyclization because the activation
energy for the rearrangement is low. Cyclization thus dominates in neat lipids at
TABLE 8. Variation in Dominant Propagation Mechanism and Product Distribution

for Linoleic Acid Oxidized in Different Solvents.
Product Distribution (%)

a b
Solvent and System LOH/LOOH Cyclic Scission Other Unknown Reference
CHCl2, FeCl3, early 100 266

Anhydrous MeOH 38 7580 1315 318
Cyclohexane, 7.5 mM 15 68 18c 216
80% ethanol 30 11 7d 7e 214
FeCl3/cysteine
a
Total of all H abstraction products, all isomers.
b
Total of all products that had any cyclic component.
c
Oxo dienes.
d
Hydroxyl ethoxylated products from rx with solvent radicals.
e
Unidentified soluble products and volatile scission products.
PROPAGATION 317
TABLE 9. Effect of Alkene Structure on Preference for Addition

vs. Abstraction by t-BuO Radicals at 40 C. Data from (21).
Alkene Abstraction (%) Addition (%)
CHCH
R R (trans) 95 34
CHCH
R R (cis) 83 17
CHCH2
R 97 3
R2
CCH2 83 17
room temperature, but as the temperature increases, H abstraction and scission

become more important directors of propagation. At high temperatures
(>100 C), rearrangement is a relatively minor process (278, 317). Metals, particu-
larly Fe and Cu, activate cyclization and direct internal rearrangements to dihy-
droxy and hydroxyene multiple positional isomers (317). Iron catalyzes
isomerization and conversion of HO-epoxides to ketols (322).
3.2.3. Addition of LO to Double Bonds Addition of LO to double bonds

does not occur with the ease of LOO additions. Alkoxyl radicals have unusually
strong preference for allylic attack, so intermolecular H abstraction or internal
cyclization will dominate as long as allylic hydrogens are present. Addition is
favored by absence of allylic hydrogens and by conjugation (Table 9), conditions
that only hold after oxidation has started. Hence, propagation by LO addition is
most active in catalyzing chain branching in secondary stages of oxidation. In con-
tradistinction to LOO additions, LO addition increases with cis configuration and
asymmetrical substitution on double bonds (323), so when LO does add to lipid
hydroperoxides, it adds to the cis-rather than trans-double bonds (324).
LO
57
LO +
Propagation by LO addition is most important in neat lipids and organic sol-
vents (308). Although LO additions do occur in aqueous solvents, they are gener-
ally not competitive with scission and rearrangement reactions. Heat catalyzes the
addition. Addition of LOOH to methyl linoleate at 210 C results in complete con-
version of the LOOH to dimers containing both reactants (289). Although the exact
structure was not determined, the dimers were presumably LO-ML adducts forming
after thermal decomposition of LOOH to LO .
3.2.4. b-Scission of LO Beta-scission of alkoxyl radicals leads to scission of

C bond on either side of the LO group to yield a mixture of carbonyl pro-
the C
ducts and free radicals, typically aldehydes, alkanes, and oxo-esters, from the initial
alkoxyl radicals (297). Scission produces the volatile products so characteristically
associated with rancidity, and the mix can become quite complex in secondary
stages of oxidation.
A simplified scission is shown in Reaction 58. The a and b fragmentation in this
case refer to the position of chain scission relative to the
COOH on the fatty acid.
More complete scission maps for oleic, linoleic, and linolenic acids are presented in
Figures 810. Some of the radicals deriving from the scissions rearrange to nonra-
dical products internally, but most of them abstract hydrogens to propagate the radi-
cal chain. Unsaturated fragments, particularly those containing conjugated dienes,
are still susceptible to oxidation and their subsequent reactions also contribute to
chain branching.

.............................. ..............................
R1 CH R2 R1 + CH R2 OR R1CH + R2 58
O O O
Scission of alkoxyl radicals is a solvent-dependent, solvent-driven process (325).

Fragmentation of the carbon chain proceeds through formation of a transition state,
which mediates the transformation from nonpolar alkoxyl radical to polar cleavage
products. Water and polar protic solvents stabilize both the increasingly polar tran-
sition state and the carbonyl products by providing solvation and hydrogen bonding
to support the transition state and reduce the activation energy for bond rupture
(263, 305, 323, 326, 327); H from the solvent then adds immediately to the scis-
sion radicals to provide the driving force for the reaction (224). This process is
shown in Reaction 59 for a-scission (21). Scission is also favored when lipids
are in dilute solution in nonpolar organic solvents where there is reduced competi-
tion from hydrogen abstraction.
O O O 2+
R CH R1 +... RCHO + R1
R CH R1 R CH R1
Increasing polarity
59
Scission is rapid in polar solvents. The ks for alkoxyl radicals in aqueous solu-
tion is 106107 s1 (328330), 10100 times faster than rates in nonpolar organic
solvents (331333) that have dielectric constants comparable with fatty acid methyl
esters. Even though this is somewhat slower than H- abstraction (Table 10), scission
usually competes effectively, and under appropriate conditions, scission can dom-
inate. In polar media, scission accounts for at least half of the LO reactions even in
early oxidation. For example, Bors (308) found
48% fragmentation,
48% H
abstraction, and 4% unreacted t-BuO in aqueous solution on a pulse radiolysis
time scale (ms to s). n-6 Fatty acids oxidized in Tris-KCl FeSO4/ascorbic acid
for up to 24 hrs gave the scission fragment 2-hydroxyheptanal as the sole product
(334). Scission accounted for 710% of the oxidation products in neat triolein, but
PROPAGATION 319
OLEIC ACID
O
CH3(CH2)6CH=CHCH(CH2)7COOH
OCH(CH2)7COOH CH3(CH2)6CH=CHCHO
9-oxo-nonanoic acid 2-decenal
+ +
CH3(CH2)6CH=CH (CH2)7COOH
(1-nonenol) CH3(CH2)6CH=CHOH
8-HO-octanoic acid
HO HO
nonanal CH3(CH2)6CH2CHO HOCH2(CH2)6COOH
RH R RH R
1-nonene CH3(CH2)6CH=CH2 CH3(CH2)6COOH octanoic acid
O2 / H O2 / H
(CH3(CH2)6CH=CHOOH) (HOOCH2(CH2)6COOH)
CH3(CH2)6CH=CHO + OH HO + OCH2(CH2)6COOH
nonanal CH3(CH2)6CH2CHO CH3(CH2)6CH2 (CH2)6COOH OCH2(CH2)6COOH

+ +
HCHO HCHO 8-oxo-octanoic acid
Formaldehyde
Following the same fragmentation pattern -

-scission -scission
8-O 8-oxo-octanoic acid + decanal 2-undecenal + 7-HO-heptanoic acid

1-decene heptanoic acid
nonanol 7-oxo-heptanoic acid
nonane 6-HO-hexanoic acid
nonanal hexanoic acid
formaldehyde 6-oxo-hexanoic acid
formaldehyde
10-O 10-oxo-8-decenoic acid + octanol nonanal + 9-oxo-nonanoic acid

octane 8-nonenoic acid
octanal octanol
heptanol octane
heptane octanal
heptanal formaldehyde
formaldehyde
11-O 11-oxo-9-undecenoic acid + heptanol octanal + 10-oxo-decanoic acid

heptane 9-decenoic acid
heptanal nonanol
hexanol nonane
hexane nonanal
hexanal formaldehyde
formaldehyde
Figure 8. Typical initial scission patterns of oxidizing oleic acid. Data from (340, 341). Paren-
theses indicate unstable intermediates; brackets denote products from secondary scissions.
LINOLEIC ACID
O
CH3(CH2)4CH=CHCH=CHCH(CH2)7COOH
OCH(CH2)7COOH CH3(CH2)4CH=CHCH=CHCHO
9-oxo-nonanoic acid 2,4-decadienal
+ +
CH3(CH2)4CH=CHCH=CH (CH2)7COOH
CH3(CH2)4CH=CHCH=CHOH
(1-HO-2,4-nonadienol) 8-HO-octanoic acid
HO HO
CH3(CH2)4CH=CHCH2CHO HOCH2(CH2)6COOH
3-nonenal RH R RH R
CH3(CH2)4CH=CHCH=CH2 CH3(CH2)6COOH octanoic acid
1,3-nonadiene O2 / H O2 / H
(CH3(CH2)4CH=CHCH=CHOOH) (HOOCH2(CH2)6COOH)
CH3(CH2)4CH=CHCH=CHO + OH HO + OCH2(CH2)6COOH
CH3(CH2)4CH=CHCH2CHO OCH2(CH2)6COOH
3-nonenal CH3(CH2)4CH=CHCH2 (CH2)6COOH 8-oxo-octanoic acid
+ +
HCHO HCHO
Formaldehyde
-scission -scission
13-O 13-oxo-9,11-tridecadienoic acid hexanal + 12-oxo-9-dodecenoic acid

+ pentanol 9,11-dodecadinoic acid
pentane 11-HO-9-undecenoic acid
pentanal 9-undecenoic acid
butanol 11-oxo-9-undecenoic acid
butane formaldehyde
butanal
formaldehyde
Figure 9. Typical initial scission patterns of oxidizing linoleic acid. Data from (340, 341).
Parentheses indicate unstable intermediates; brackets denote products from secondary
scissions.
this shifted to 6675% in the presence of acid (335). In a series of solvents, an

alkoxyl derivative of benzene gave 2481% scission, with the proportion increasing
with solvent polarity; H abstraction was able to compete only when scission was
less than 35% (336).
In contrast, scission is a minor process in neat lipids or aprotic solvents at room
temperature where background levels of scission products range from about 1% in
monoacid triacylglycerols (337) to 1020% in free esters of O, L, and Ln (14, 335)
PROPAGATION 321
LINOLENIC ACID
O
CH3(CH2)CH=CHCH2CH=CHCH=CHCH(CH2)7COOH
OCH(CH2)7COOH CH3CH2CH=CHCH2CH=CHCH=CHCHO
9-oxo-nonanoic acid 2,4,7-decatrienal
+ +
CH3CH2CH=CHCH2CH=CHCH=CH (CH2)7COOH
CH3CH2CH=CHCH2CH=CHCH=CHOH
(1,3,6-nonatrienol) 8-HO-octanoic acid
HO HO
CH3CH2CH=CHCH2CH=CHCH2CHO HOCH2(CH2)6COOH
3,6-nonadienal RH R octanoic acid
RH R
CH3CH2CH=CHCH2CH=CHCH=CH2 CH3(CH2)6COOH
1,3,6-nonatriene
O2 / H O2 / H
(CH3CH2CH=CHCH2CH=CHCH=CHOOH) (HOOCH2(CH2)6COOH)
CH3CH2CH=CHCH2CH=CHCH=CHO + OH HO + OCH2(CH2)6COOH
CH3CH2CH=CHCH2CH=CHCH2CHO OCH2(CH2)6COOH
CH3CH2CH=CHCH2CH=CHCH (CH2)6COOH
3,6-nonadienal 8-oxo-octanoic acid
+ +
HCHO HCHO
Formaldehyde

-scission -scission
16-O 16-oxo-9,12,14-hexadecatrienoic acid propanal + 15-oxo-9,12-pentadecadienoic acid

+ ethane 9,12,14-pentadecatrienoic acid
14-HO-9,12-tetradecadienoic acid
9,12-tetradecadienoic acid
14-oxo-9,12-tetradecadienoic acid
formaldehyde
13-O 13-oxo-9,11-tridecadienoic acid 3-hexenal + 12-oxo-9-dodecenoic acid
+ 2-pentenol 9,11-dodecadienoic acid
2-pentene 11-HO-9-undecenoic acid
2-pentenal 9-undecedoic acid
butanal 11-oxo-9-undecenoic acid
butene formaldehyde
formaldehyde
12-O 12-oxo-9-dodecenoic acid 2,4-heptadienal + 11-HO-9-undecenoic acid

+ 3-hexenal 9-undecenoic acid
3-hexene 11-oxo-9-undecenoic acid
2-pentenol 10-oxo-decanoic acid
2-pentene 9-decenoic acid
2-pentenal formaldehyde
formaldehyde
Figure 10. Typical initial scission patterns of oxidizing linolenic acid. Data from (340, 341).
Parentheses indicate unstable intermediates; brackets denote products from secondary
scissions.
TABLE 10. Solvent Effects on Rates of H Abstraction (ka) and b-Scission

(kb) of Cumyloxyl Radicals (CumO). Data from (327).
H Abstraction b-Scission
6 1 1
ka 10 M s kb 105 s1 ka/kb M1
CCl4 1.1 2.6 4.5
C6H6 1.2 3.7 3.2
C6H5Cl 1.1 5.5 2.0
(CH3)3COH 1.3 5.8 2.3
CH3CN 1.2 6.3 1.9
CH3COOH 1.3 19 0.7
on a total weight basis. The fact that it is observed at all under these conditions is
probably due to acceleration of scission in the presence of double bonds through the
increase in polarity. Dipole-dipole interaction between alkoxyl radicals on one fatty
acid with double bonds on an adjacent unsaturated fatty acid forms a charge transfer
transition state that induces electron and charge redistributions, thus facilitating
scission (263).
+
CR2 CR2 CR2
R1(R2)HC O + R1(R2)HC O R1HC O + R2 +
CR2 CR2 CR2
60
Alkoxyl radical (LO ) scission makes its greatest contribution to propagation at

elevated temperatures (323) that overcome the large Ea and log A (Arrhenius factor)
for scission (338). Heat accelerates alkoxyl radical scissions in all solvents,
although the pattern of cleavage may change as temperature increases because pri-
mary scission of the alkoxyl radical (Reaction 49) gradually increases at the
expense of alternative reactions (339), and at high enough temperatures, secondary
scissions also occur. Cleavage of polyperoxides, for example, is unimportant at
60 C, but becomes a major contributor to propagation at T > 100 C (278). Simi-
larly, fragmentation of sulfenyl alkoxyl radicals varies from 21% at 10 C to 40%
at 50 C (338). Shifts in scission products at different temperatures have been
reviewed in detail by Grosch (340).
What determines whether a scission will be a or b relative to the -COOH is an
age-old question that still has not been completely answered. There are suggestions
in the chemical literature that scissions should occur between the alkoxyl radical
and the double bond (314, 341),
O
61
R CH CH CH CH2 R R CH CH > CH2R,
PROPAGATION 323
TABLE 11. Distribution of a and b Scissions as a Function

of Alkoxyl Radical Distance from COOH in Oleic Acid. Data
from (312).
LO Position % a-Scission % b-Scission
8 3.2 7.4
9 10.4 8.0
10 22.0 6.5
11 23.0 10.6
but this has been questioned on energetic grounds, i.e., the dissociation energy for
vinyl bonds is 109 kcal but for allylic bonds is 60 (111). An alternative explanation
is that scission will occur preferentially at the site fulfilling the thermodynamic
requirement to form the most stable product, e.g., saturated aldehydes are more
stable than unsaturated aldehydes. However, there are disagreements over whether
the stability of the radical (333, 342344) or the carbonyl product (327, 341) is the
determining force. Inductive effects of the COOH group increase the tendency
toward a scission, but this has less overall influence than the aforementioned fac-
tors. The presence of an acid group favors selective cleavage between OOH and
double bond (a or b depending on position of OOH). Heat and metals induce one-
electron redox reactions, which generate a-monocarbonyls (scission on COOH
side of alkoxyl radicals) and an ejected radical that can initiate new chains (335).
Thus, the scission pattern for oxidizing fatty acids is mixed and varies with the pro-
duct structure and reaction conditions.
Evaluation of products from oleate oxidation provides a simple example of how
these factors interact in directing scission (312). The tendency for a-cleavage
increases as the alkoxyl radical position moves away from the carboxylic acid;
there is relatively little positional preference for b-cleavage (Table 11). This pattern
is consistent with preferential scission between the alkoxyl radical and the double
bond as well as formation of saturated aldehydes.
The dominant products do indeed derive from scission between the alkoxyl radi-
cal and the double bond, but a variety of scissions that are less favorable thermo-
dynamically occur at the same time, generating the complex mixture of products
shown in Figures 810 and Table 12. For monohydroperoxides, scission varies
with the position of the alkoxyl, with the longest saturated product receiving pre-
ference. For alkoxyl radicals from dihydroperoxides, dominant cleavages are still
between the CO and double bond, but
40% occur at the alkoxyl nearest
the
COOH, and half that occur on the CH3 terminal alkoxyl radical (345).
It is important to recognize that scission does not necessarily stop after reaction
of initial alkoxyl radicals. Scissions of secondary products generated during lipid
oxidation also contribute to propagation and to the ultimate product mix (346).
Malonaldehyde is perhaps the best known example of this, as will be discussed
further in Section 4.2.
TABLE 12. Scission Products from Unsaturated Fatty Acids Oxidized

at Room Temperature. (Data from 3, 8, 273, 290, 314, 340, 341, 345, 347349.)
OLEIC ACID
Major Products Product Classes and Carbon Chain Length
Nonanal Hydrocarbons 68
Octanal Alkanals 24, 5, 6, 7, 8, 9, 10,11
Undec-2-enal 2-Alkenals 69, 10, 11
Undecanal Acids 1, 69
Alkanols 58
2-decenal Alkylformates 28
LINOLEIC ACID
Hexanal Hydrocarbons 35
2,4-decadienal Alkanals 3, 4, 5, 6, 7, 8
2-octenal Alkenals 7, 8, 9, 10
2-heptenal Dienals 9, 10
Oxo-alkanals 7, 8, 9
Ketones 7, 8
Alcohols 3, 4, 5, 6, 7, 8
Acids 1, 5, 6, 7, 9
Esters 1, 6, 7, 8
LINOLENIC ACID
2,4-heptadienal Hydrocarbons 13
3-hexenal Alkanals 13, 6
Propanal Alkenals 4, 5, 6, 7
2,4,7-decatrienal Dienals 7, 8, 9
2-pentenal Trienals 10
Octadienal Ketones 5, 18
Pentene-3-one Alcohols 3, 4, 5, 6, 7, 8
Octadiene-2-one Acids 1, 5, 6, 7, 9
Esters 1, 6, 7, 8
Oxo-alkanals 1, 6, 7, 8
ARACHIDONIC ACID
Hexanal Alkanal 2, 6, 7
2,4-decadienal Alkenal 7, 8, 9, 11
2,4,7-tridecatrienal Dienal 9, 10, 11, 12
2-heptenal Ketones
2-octenal Alkanes 5, 6
Pentanal Aldehyde esters 4, 5
1-octen-3-one
4-decenal
3,5-undecadien-2-one
2,6-dodecadienal
5-oxo-pentaoate
PROPAGATION 325
3.3. Propagation Reactions of LOOH; Mono- vs. Bimolecular LOOH

Decomposition and Chain Branching
It should be obvious from the discussion above that hydroperoxides are the key
intermediates controlling the progress of lipid autoxidation. As long as peroxyl
radicals remain in the b-scission manifold, with O2 continuously being added to
or eliminated from lipid alkyl radicals, perceptible oxidation does not progress.
This is the well-known induction period (Figure 11) during which oxygen absorp-
tion and traditional chemical changes in the lipids are difficult to detect for several
reasons:
1. Until H abstraction occurs, no net change occurs.

2. Standard analytical techniques are not sensitive enough or fast enough to
detect the low levels of initial products.
3. Only traditional products such as conjugated dienes and hydroperoxides have
been analyzed in most cases. If cyclization occurs before H abstraction to
form hydroperoxides, the oxidation may be missed by standard peroxide
value analyses.
4. Hydroperoxides are often breaking down as fast as (or faster than) they are
formed.
Ultimately, production of lipid hydroperoxides, even by circuitous routes,

becomes the major process driving the oxidation reaction forward. LOOH are the
first stable products of lipid oxidation, accumulating in the absence of pro-oxidant
heat, metals, hemes, ultraviolet light, peroxyl radicals, or antioxidant acids or
nucleophiles. However, from a practical standpoint, one or more of these or other
decomposing factors are nearly always present, so the low energy O O and O H
bonds undergo a variety of scission reactions. Indeed, a large proportion of the
M B
I
Extent of reaction
LO products
2 LOOH LO + LOO
Secondary oxidation processes
LOOH LO + OH/OH

LOO LOOH ,epidioxide
L LOO
O2 O2
LH
Time of reaction
Figure 11. Diagrammatic representation of changes in dominant reactions and products over
the course of lipid oxidation. Three separate rate periods are usually designated: Induction
period (I), monomolecular rate period (M), and bimolecular rate period (B).
LOO and all of the LO involved in propagation are not ab initio radicals, but
derive from some form of LOOH decomposition (350).
To briefly recap what has already been covered in Section 2, redox-active metals
break the O-O bond by electron transfer, hence LOOH decomposes heterolytically
to generate radicals and ions. Reducing metals such as Fe2 and Cu generate
alkoxyl radicals (LO ) and hydroxide ions (OH ), whereas oxidizing metals such
as Fe3 and Cu2 give peroxyl radicals (LOO ) and hydrogen ions (H):
LOOH + Mn+ (fast) LO + OH K ~ 109 M1sec1 62a
LOOH + M(n+1)+ (slow) LOO + H+ 62b
Hydroperoxides are also recycled by reaction with peroxyl radicals:
LOOH + LOO LOOH + LOO 63
These reactions must be distinguished from homolytic decomposition by heat

and UV light that break the O O bond by energy deposition, yielding alkoxyl
and hydroxyl radicals ( OH). The O O in organic hydroperoxides (BDE 25
38 kCal/mole) begins decomposing at about 50 C and is completely decomposed
at 160 C (297).

LOOH LO + OH 64
Homolytic scission is much more catastrophic in terms of lipid oxidation because

two propagating radicals are released per hydroperoxide, LO is more reactive and
more selective than LOO , and HO is extremely reactive. HO is rather unselec-
tive, abstracting hydrogen atoms all along the acyl chain, and it can also readily add
to double bonds (still generating a radical). Hence, the net effect of LOOH decom-
position is a transition in mechanism and kinetics. Lipid oxidation essentially gath-
ers steam, increasing in rate and extent as LO becomes the dominant chain carrier
and secondary chains are initiated. This process, often referred to as chain branch-
ing, greatly amplifies and broadcasts the effects of initiation: if not intercepted by
nonlipid molecules, a single initiating event can result in sequential oxidation of
literally hundreds of molecules in the primary chain and in secondary branching
chains, as shown in Figure 12. The net effect is a noticeable increase in measurable
oxidation, as seen in the monomolecular rate period (Figure 11).
Oxygen uptake remains slow and LOOH decomposes monomolecularly (Reac-
tions 6264) during early stages of lipid oxidation as main chains are extending,
branches are developing, and hydroperoxide concentrations are low. However,
the process becomes more complicated as LOOH accumulates, e.g., in the presence
of lipoxygenase or the absence of decomposers at low temperature, in the dark, or
with metal chelation. At high [LOOH], i.e., greater than 1% oxidation (114), it has
been proposed that LOOH transition dimers form via hydrogen bonding and bimo-
lecular decomposition ensues, leading to greatly accelerated oxidation (277, 351):
2 LOOH LOOH...HOOL LO + H2O + OOL 65

PROPAGATION 327
Main radical chain

O2 L2H O2 L3H O2 L4H O2 L5H
L1 L1OO L2 L2OO L3 L3OO L4 L4OO L5 etc
.
L1OOH L2OOH L3OOH + L4OOH
Branching chains
L1O + OH L2OO + H L3O + H2O + OOL4
L 6H L7H L8H L9H L10H
L6 L7 L8 L9 L10
Figure 12. Chain propagation and branching in lipid autoxidation. The main chain starts at the
the ab initio radical, L1 , and is driven by cyclical addition of oxygen to form LOO , then
abstraction of hydrogens to generate new propagating L and product LOOH. Branching
reactions are secondary chains originating from radicals produced via a variety of LOOH
decompositions.
Now both alkoxyl and peroxyl radicals are present equally, and although LO dom-
inates kinetically, LOO still produces secondary chains. This shift can be seen gra-
phically in the dramatic increase in oxygen consumption rates and production of
LOOH in the bimolecular rate period (B) in Figure 11.
Does bimolecular decomposition actually occur? This mechanism has been
widely included in discussions of lipid oxidation, but it is also somewhat controver-
sial. Supporting the theory is the rapid O2 uptake and shift in product mix (352), as
well as the tendency of hydrophilic hydroperoxides to dimerize in nonpolar solu-
tions (353) (e.g., neat lipids) and at high concentrations where bimolecular decom-
position is thermodynamically favorable (less endothermic than unimolecular
homolysis) (353). Contradicting this are poor fits of oxidation kinetics for some
compounds (354). Possible sources of inconsistencies between studies include
the measures used to determine kinetics (appearance of product vs. loss of starting
material vs. oxygen consumption), specific assumptions made in deriving kinetic
equations, and particularly the nature of the oxidizing compound. The rate of reac-
tion increases and the fit of kinetic data improves with a decrease in C H bond
dissociation energies and with increased chain length. For example, methyl oleate,
for which the oxidation kinetics are faster, has a better fit with bimolecular break-
down theory than either n-decene or ethyl benzene (355).
The latter observations with methyl oleate, together with thermodynamic consid-
erations and EPR evidence for free radical intermediates, suggest an alternative
explanation for the dramatic increase in oxidation rates once hydroperoxides
accumulate, namely that bimolecular decomposition may be specific to allylic
hydroperoxides and proceed via LOO radical-induced decomposition rather than
by dissociation of hydrogen-bonded dimers (280). Reaction sequence 63 is analo-
gous to Reactions 49 and 50a, where one slowly reacting radical reacts with a
nonpropagating hydroperoxide to generate three very reactive radicalstwo LO

and one OH. The heavy arrows indicate the favored pathway.
OOH LOO OOH

LOO + R1 CH2 CH CH CH CH2- R1 CH2 CH CH CH CH2-
O OOH LOO O
LO + R1 CH2 CH CH CH CH2- R1 CH2 CH CH CH CH2- + OH
O O O O
R1 CH2 CH CH CH CH2- + OH LO + R1 CH2 CH CH CH CH2-
66
Radical-induced decomposition is thermodynamically favorable (Ea 37.5 kCal),

and is also more consistent with the characteristics of bimolecular initiation by
hydroperoxides originally proposed by Russell (356), the kinetics measured in lipid
oxidation systems, and significant epoxide products reported in many studies. Most
importantly, the radical-induced decomposition described in Reaction 63 provides a
powerful cascade of reactive radicals to fuel the very rapid increase in oxidation
during the bimolecular rate period.
3.4. Factors Influencing Propagation Pathways (abstraction vs.

scission reactions vs. rearrangement) of LOO and LO
The net lipid oxidation observed is a net sum of all the competing reactions occur-
ring in a given system:
LOO Reactions LO Reactions LOOH Reactions
b-scission of O2 ! isomers H abstraction Decomposition !

H abstraction Addition chain branching
Cyclization to epidioxides Cyclization to epoxides
Addition b-scission ! fragment products
Dismutation
Hopefully, this chapter has made it clear that there is no fixed sequence of reac-
tion pathways for lipid oxidation. Rather, the pathways most active probably
change with reaction system, determined by the type and concentration of lipid,
PROPAGATION 329
the solvent, phase distributions of catalysts, surface and interfaces, and numerous
other factors. As a consequence, no one standard assay will give a complete or
accurate picture of the progress of lipid oxidation. Indeed, one of the difficulties
in sorting out controlling factors is that so few lipid oxidation studies have analyzed
products quantitatively as well as qualitatively, and even fewer have measured mul-
tiple classes of products simultaneously. Several decades of detailed, painstaking
product analyses, as discussed above, have now provided a reasonably clear picture
of what kinds of compounds are generated during lipid oxidation, but we still need
coordinated quantitative analyses of all the classes of products to determine relative
contributions of the various pathways under specific reaction conditions. Such
information would tremendously improve our ability to tailor oxidation analyses
to individual systems as well as to design more effective antioxidant strategies.
Arguments have been presented in the literature that the structure and configura-
tion of the target molecule at the time of radical attack sterically and thermodyna-
mically establish the reaction mechanisms, whereas system conditions, particularly
temperature, have relatively little effect. Based on short-term oxidation of simple
alkenes, Van Sickle and coworkers (275, 357) proposed that the ratio of H abstrac-
tion to addition is determined by the alkene structure and is constant over a very
wide temperature range. There is some support for this position in the thermody-
namics of H abstraction vs. addition with different double bond structures
(Table 13). Clearly, doubly allylic hydrogens are the most susceptible to abstrac-
tion, and with this structure, H abstraction has a slight edge over addition most
of the time. Allylic hydrogens of isolated double bonds are less susceptible to
both H abstraction and addition, reflecting relatively low reactivity (as with oleic
acid). In contrast, the conjugated double bond is activated chemically as a result
of its extended resonance system: only in this structure is addition competitive
with H abstraction, and both reactions are strong. Thus, conditions that favor addi-
tion actually develop during lipid oxidation. Although addition is of little impor-
tance in early stages, it becomes quite important in secondary stages of oxidation
for linoleate and higher PUFAs. However, Van Sickles theory is not totally applic-
able to lipids because decades of research has shown quite clearly that system con-
ditions play a major role in determining which propagation mechanisms dominate
in lipid oxidation.
Reaction preferences in lipid oxidation have mostly been deduced from product
analyses; the few rate constants available for lipid reactions have been determined
TABLE 13. Ease of Oxyl Radical H Abstraction vs. Addition

for Different Double Bond Configurations. Data from (279).
H abstraction H addition
kcal/mol kcal/mol
RCH2CHCHCH2R0 15 8
CHCHR0
RCHCH 19 20
RCHCH CHCHR0
CH2 26 8
in pulse radiolysis studies in the laboratories of Patterson (195197, 358) and Bors
(194, 198, 261, 299, 308). Nevertheless, we can gain some insights from attempts to
determine relative contributions of H abstraction, rearrangements, radical additions,
and scissions in oxidation of small alkenes that lack the steric complications of fatty
acid chain length and polyunsaturation (206, 275, 312, 327, 336, 357). Relevant rate
constants are compiled in Table 14. The table includes all fatty acid reaction rates
available, and these are supplemented with rates from related compounds, primarily
tert-butyl and cumyloxyl radicals. This approach is justified because model systems
have shown that H abstraction rates are determined primarily by the bond strength
of the H being abstracted and are relatively independent of the R-group of the
abstracting oxy radicals (278). Also, tert-butyl peroxyl and alkoxyl radicals, as
well as the corresponding oxyl radicals of cumene, have been shown to be
reasonable models for unsaturated fatty acids (261, 299, 308, 332, 333). Therefore,
consideration of the comparative rate data that has accumulated in defined chemical
systems can help elucidate the logic of oxidation processes in lipids. Most critically,
it shows how we are usually looking at a totally different process when systems are
oxidized under different conditions, and our interpretations of product data and
designs of antioxidant strategies must recognize and account for alternative oxida-
tion pathways.
When searching for rate constants to support the product distributions identified
under different conditions of lipid oxidation begain, numbers were expected that
would establish a distinct kinetic hierarchy. Surprisingly, what is most apparent
from the rate constants in Table 13 is the lack of clear priority of any of the reac-
tions so that it becomes difficult to establish any rules for expected reactivity.
Rather, the dominant products in any given reaction must be specifically system
dependent. Some of the distinction between reactions may be blurred in ranges
of values encompassing multiple sources of oxyl radicals that only approximate
reactions of lipid radicals, and this argues for more research focused specifically
on lipid reactions. Nevertheless, several important patterns do emerge.
1. The literature has long noted that alkoxyl radical reactions were faster than
peroxyl radicals and that reaction rates increased with the solvent polarity.
The values in Table 13 reveal the magnitude of those differencesseveral
orders of magnitude in most cases.
2. Both peroxyl and alkoxyl radicals abstract hydrogens much faster from
hydroperoxides than from lipid allylic positions, a fact that has been little
appreciated previously and can have great consequences to oxidation kinetics
and product distributions.
3. There is a surprising lack of clear preference for one reaction over another,
except that H abstraction has a slight priority in general. Thus, most systems
should be expected to produce mixtures of products rather than a single class,
and only small modifications in reaction conditions (including extent of
oxidation) are sufficient to shift the balance between abstraction, cyclization,
and scission reactions, altering the product distribution.
TABLE 14. Rate Constants for Competing Reactions of Lipid or Related Peroxyl and Alkoxyl Radicals.a
ROO Reference LO Reference

1 1 4 7 1 1
H abstraction, LH nonpolar organic <1400 M s 88, 223, 247, 258, 359 10 10 M s 197, 240, 327, 331
polar, aqueous 106108 L M1s1 198
H abstraction, LOOH nonpolar organic 600 M1s1 223, 360 2.5 108 M1 s1 307
polar, aqueous NA NA
Cyclization nonpolar organic 101103 s1 11, 247 104105 s1 307, 361
Addition nonpolar organic NA 104108 M1 s1 258, 332
b-scission oleate 18 s1 11, 250, 253 103105 s1 org 305, 327, 364
linoleate 27430 s1 11, 226 104105 s1 polar org 305, 327
106107 s1 aq 198, 328, 333
Dismutation nonpolar organic 106109 L M1s1 192, 258, 354, 362, 363 1091010 M1s1 361, 362, 364, 365
polar, aqueous 107108 L M1s1 261 NA

oleate-OO 106 M1s1 195, 196, 223
a
Data included authentic fatty acids whenever possible, plus primarily cumyl, tetralinyl, and t-butyl peroxyl and alkoxyl radicals.
NA: data not available.
At the risk of being redundant, let me summarize conditions that shift chain
propagation mechanisms in lipid oxidation:
a. Hydrogen abstraction from other fatty acid chains by LOO and LO is
favored under conditions providing close contact between lipid chains with-
out competition from other H sourcesi.e., in aprotic environments such as
neat lipids and the lipid interior of membranes, where lipid chains are closely
associated. In solvents, H abstraction is favored at moderate lipid concentra-
tions where enough substrate is present to supply hydrogens. However, at low
lipid concentrations, cyclization or scission dominate, whereas at high concen-
trations, radical additions and recombinations become more important (279).
b. Hydrogen abstraction rates increase with solvent polarity and temperature
but under these conditions, accelerated propagation of lipid oxidation as in (a)
must compete with H abstraction from solvent or other nonlipid sources and
also with increased rates of scission.
c. Cyclization is favored when oxygen is limited and abstractable hydrogens are
not available, i.e., in neat lipids, aprotic solvents, and low lipid concentra-
tions. Cyclization is facilitated by polyunsaturation, radical formation at
internal positions, and iron chlorides. As temperature increases, cyclization
diminishes in importance as a propagation mechanism because it is less
affected by temperature than other propagation processes and because
epidioxide peroxyl radicals have an increasing tendency to dimerize rather
than abstract hydrogens.
d. Scission is favored over H abstraction in polar protic solvents that provide the
protons necessary to stabilize the scission products, but an excess of water
shifts propagation to termination as protons for stabilization of secondary
products are drawn from nonlipid sources and increased hydrolysis yields
tertiary lipid oxidation products. Scission also increases markedly with
temperature as thermal energy facilitates bond rupture.
e. Propagation by addition is generally a minor reaction whenever hydrogen
sources are readily available, but increases when abstractable hydrogens
are limited in aprotic solvents, particularly when there is a conjugated double
bond. Thus, addition becomes more important once oxidation chains are
established. Addition also increases with lipid concentration, but under these
conditions it also must compete with increased rates of H abstraction.
All the pathways outlined above eventually lead to H abstraction to form inter-
mediate products that then breakdown to secondary products. Why, then, is the dis-
tinction between propagation mechanisms important, other than as an academic
exercise? The answer is that shifting among propagation pathways critically affects
the kinetics of oxidation, whether determined by oxygen consumption or appear-
ance of specific products, and can induce large differences in the ultimate mix of
products, particularly volatiles. This has several important implications and conse-
quences. The first is analytical. If the dominant pathway is not being monitored, an
TERMINATION 333
inaccurate picture of the rate, extent, and character of lipid oxidation is generated
and reactivity is misinterpreted. For example, when peroxide values alone are used
to follow oxidation under conditions favoring cyclization or scission, much of the
lipid change may be missed altogether. Second, changes in the product distributions
critically alter flavors and odors from lipid oxidation, and also the potential for
secondary effects such as nonenzymatic browning and reactions with proteins.
Finally, without information about dominant and active propagation pathways,
the most effective strategies for inhibition of the oxidation may not be applied.
For example, using only phenolic antioxidants in systems where scission is domi-
nant will probably not be sufficient to stop production of off-flavors and odors. To
achieve long-term stability, antioxidant approaches must be tailored specifically to
control all active propagation pathways.
4. TERMINATION
Termination is one of those nebulous handwaving terms used to imply that a pro-
cess is coming to a close. In lipid oxidation, termination is an even fuzzier con-
cept in that, from a practical standpoint, the lipid oxidation chains probably never
fully stop. In addition, a specific radical may be terminated and form some product,
but if this occurs by H abstraction or rearrangement, another radical is left behind
so the chain reaction continues. Net oxidation slows down when H abstractions or
other radical quenching processes exceed the rate of new chain production, but it
would be difficult indeed to totally stop the entire radical chain reaction. Thus, in
the discussion below, termination refers to an individual radical, not the overall
reaction.
Free radicals terminate to form nonradical products by four major mechanisms:
a. Radical recombinations
b. A variety of cleavage reactions when proton sources are present to stabilize
products
c. Co-oxidations of other molecules (radical transfer)
d. Eliminations
LOOH decompositions and rearrangements, sometimes listed as termination reac-

tions, are major sources of propagation LOO , LO , and OH radicals, so were dis-
cussed previously in Section 3.3. The mechanisms dominating in a given system are
influenced by the nature and concentration of the radicals, the oxygen pressure, and
the solvent.
4.1. Radical Recombinations

The number of variations possible for radical recombination is nearly limitless, and
this accounts, in part, for the broad range of oxidation products detected in lipid
1.0
R
ROO 25C
35C
% Termination

ROO / R
RO rxs 45C
1 10 100 20 40 60 80 100
pO2 (mm Hg) pO2 (mm Hg)
Figure 13. Effects of oxygen and temperature on termination processes in lipid oxidation.
Adapted (114).
oxidation. Nonetheless, recombinations are not random, and distinct patterns of

favored recombinations have been identified.
Temperature and oxygen pressure are key determinants of radical recombina-
tion pathways. As shown in the well-known curves of Figure 13 (114, 115), L
reactions dominate under low oxygen (pO2 1 to about 80100 mm Hg)
and high temperature (reduced O2 solubility) conditions, high pO2 favors
LOO reactions (more likely additions than recombinations), and LO contribu-
tions to the product mix dominate when LOOH or LOO decompositions are fas-
ter than their formation, i.e., in secondary stages of oxidation and at moderate
temperatures and oxygen pressures (15). These oxygen effects on product distri-
butions are indeed striking, but they should not be misconstrued as the only role
of oxygen. Oxygen plays critical and complex roles in all three stages of lipid
oxidationinitiation, propagation, and terminationalthough the effects are
different in each stage (363), as was implied in discussion of Sections 2 and 3
above.
4.1.1. Peroxyl Radicals Secondary peroxyl radicals, as are found in most lipid
acyl chains, recombine rapidly (2k 108-109 M1s1) (192, 362) to form a variety
of products, including alcohols and ketones (Reaction 67) (361, 362, 366), ketones
and alkanes (Reaction 68) (60, 292), or acyl peroxides and peroxyl radicals (Reac-
tion 69) (264, 367, 369). The alcohols thus produced are indistinguishable from H
abstraction products of an original LO , but the ketones and dialkyl peroxides are
unique to recombination reactions. As any R3OO and RO released from Reaction
68 or Reaction 69a react further, peroxyl radical recombinations also have the
potential for propagating lipid oxidation (Section 3.1.4).
TERMINATION 335
Concerted addition:
OO R1 R1 OH O
2 R1CHR2 HC OOOO CH R1CHR2 + R1CR2 + O2
Noniradical
R2 R2 termination
67
OO O O R3H
2 R1CR3 2 R1CR3 + O2 R1C R2 + R3
68
R2 R2 R3OO
Radical
propagation by diffusion
propagation 69a
Russell tetroxide
(8090%)
2 RO + O2
2 ROO [RO O 2 OR]
(1020%) Nonradical
Cage reaction
ROOR + O2
termination 69b
Stepwise addition:
O2
ROO R ROO 70
ROO
Whether radical or nonradical products dominate depends on the nature of the

peroxyl radical, the solvent, and the temperature (292). The self-reaction is facili-
tated in neat oils or aprotic solvents where high LOO concentrations can accumu-
late and H abstraction from external molecules is limited; such LOO
recombinations have been extensively cited as the dominant termination product
under high pO2 conditions. Thus, it is surprising that in reality, LOO recombina-
tion is a major reaction only for oleic acid where the reaction is relatively slow
(2k 110 106 M1 sec1) (223) even though there are fewer competing reac-
tions (192). Some evidence for Russell mechanism in oxidizing linoleic acid has
been presented (367, 368). However, in the higher PUFAs, there is a much stronger
tendency toward internal rearrangements to epoxides, etc. (369), as has been
discussed above, and the LOO disappears very rapidly by other reactions
[2k 107 L mol1 sec1 for L and Ln (196); 2k 4.8 108 mol1 sec1 for An
(192)]. LOO still forms crosslinks, but via addition reactions rather than peroxyl
recombinations.
The a-hydrogen is particularly important for stabilizing products, so secondary
ROO or LOO (Reaction 67) terminate 100500 times faster than tertiary ROO
(Reaction 68) (240, 360362, 364). In oleic acid, most peroxyl radicals are sec, but
tert peroxyl radicals may derive in secondary oxidations of scission products. This
may explain why oleic acid produces Russell products, although in lesser amounts
than would be expected (253). Tertiary ROO produce ketones and release new pri-
mary peroxyl radicals that can initiate radical chains, rearrange, or be quenched by
solvent (294).
There is considerable controversy over whether and how the Russell Mechanism
involving tetroxide intermediates (107) actually occurs in lipids, and whether
the oxygen is released as 1O2. In early work, Ingold proposed that the Russell
mechanism (Reaction 69) was the most important termination process for sec
peroxyl radicals (223), and the mechanism has been widely invoked. Nevertheless,
although the ketone and alcohol products are found in reactions of small primary
peroxyl radicals (366), the prescribed O2 elimination and alcohol-ketone nonradical
products have not always been observed with more complicated lipid peroxyl radi-
cals (367, 370372). Indeed, Reactions 67 and 68 are probably unimportant at room
temperature where 8090% of the ROO ends up as RO and < 20% of total reac-
tion leads to nonradical products (292). Thus, as long as pO2 is not limiting, LOO
recombination is more active as a propagating reaction than in termination to non-
radical species. However, as the temperature increases, this proportion reverses
as b-scission of oxygen from LOO predominates and LOO concentrations are
decreased below the level required for effective self-reaction (292, 366). Under
these conditions, the reaction more likely proceeds via the stepwise radical addition
process (Reaction 70) proposed as a general alternative to the Russell mechanism (277).
It should be noted that the tetroxide intermediate proposed as the mechanism for
peroxyl radical disproportionation remains somewhat controversial. If it exists, it
has been argued that the oxygen should be released as 1O2 to avoid spin restrictions
(291). Some studies claim to have detected 1O2 from lipid hydroperoxides (366),
but the evidence has not been conclusive. One of the difficulties in determining
when 1O2 is produced is that O2 reduces singlet oxygen when water sufficient
to provide a hydration shell of five water molecules is present (373).
4.1.2. Alkoxyl and Alkyl Radical Recombinations A wide variety of alkoxyl

and alkyl radical recombinations have been proposed to explain lipid oxidation pro-
ducts observed in model reaction systems and in food or biological materials. Many
are hypothetical, based on detailed studies with simple compound, but not necessa-
rily verified in lipid oxidation. Nevertheless, the radical recombinations outlined
below do provide a pathway to products not generated in the reactions already dis-
cussed. Obviously, recombinations lead to polymers. Perhaps just as importantly,
however, recombinations of the fragment radicals formed in a and b scissions of
alkoxyl radicals generate low levels of volatile compounds and flavor components
that augment those produced in scission reactions and provide the undertones and
secondary notes that round out flavors (340).
R1O + R2O R1OOR2 dimer peroxides 71
R1O + R2 R1OR2 eathers 72
R1 CH R2 + R R1 C R2 + RH ketones, alkanes
73
O O
R1 CH R2 + RO R1 C R2 + ROH ketones, alcohols

74
O O
R1 + R2 R1 R2 alkane polymers 75
TERMINATION 337
There is little data available to provide a quantitative sense of the contribution of

these radical recombinations to the overall mix of lipid oxidation products. The
rates of recombinations generally follow the energy of the dimer bond formed
(198, 305, 323):
Bond E Rate Constant (M1 sec1)
R R 8090 10101011
R RO 8090 10101011
RO RO 3540 107109
That lipid alkoxyl radicals recombine (Reaction 71) at diffusion controlled rates
(k 109 M1s1) (198, 305) probably accounts for the presence of low levels of
peroxides even under mild conditions and low levels of oxidation. In one study, oxi-
dation of linoleic acid at 30 C gave
C
OO C dimers. Reactions 7173 were
found in linolenic acid oxidized under mild conditions to PV 585; this increased
to > 50% at PV 4000 and to > 75% after heating to 40 C (276). Alkoxyl radicals
from hydroperoxyepidioxides heated at 40 C generated > 90% dimers (276).
The reactivity described in the reactions above was determined in neat oils.
When oils are in polar solvents or dilute solution in nonpolar solvents, b-scission
dominates and radical recombinations are probably unimportant.
4.2. Scission Reactions

b-scissions of alkoxyl radicals are the major source of aldehyde products in lipid
oxidation. As discussed in Section 3.2.4, a major aldehyde product and a propagat-
ing radical are formed via scission of the initial alkoxyl radical in a fatty acid. How-
ever, products continue to form as unsaturated radical fragments oxidize and
undergo secondary scissions to produce carbonyls and alkanes of shorter chain
length (224), secondary products that contribute dramatically to the characteristic
odors and flavors associated with lipid oxidation. In fact, most evidence suggests
that initial oxidation primes fatty acids for additional attacki.e., oxidation con-
tinues on the same molecules rather than randomly attacking new virgin acyl
chains. The increased susceptibility to oxidation derives both from conjugation
and secondary products that oxidize more easily than the parent fatty acids. It is,
therefore, not surprising that identifying products of scission reactions of hydroper-
oxides and various radicals has been the subject of so much research (3, 8, 232, 273,
290, 314, 341, 345, 347349, 374, 375).
Review of all the scission reactions responsible for the hundreds of volatile pro-
ducts in lipid oxidation is beyond the scope of this chapter. The reader is referred to
the available reviews (3, 314, 340, 341, 347) for further details. The scission pattern
of hydroperoxide epidioxides from linoleic acid is included here to show how the
decompositions can become quite complex (Figure 14), and lists of typical products
resulting from scission reactions of oleic, linoleic, and linolenic acids are presented
in Table 12.
Linoleic acid
CH3(CH2)4CH=CHCH=CHCH(CH2)7COOH
O
OCH(CH2)7COOH
CH3(CH2)4CH=CHCH=CH 9-oxo-nonanoic acid
OH OO + H
CH3(CH2)4CH=CHCH=CHOH CH3(CH2)4CH=CHCH=CHOOH
OH
CH3(CH2)4CH=CHCH2CHO CH3(CH2)4CH=CHCH=CHO
CH3(CH2)4CHCHCH + HCHO
OO
2-nonenal CH3 (CH2)4
O
CH3(CH2)4CH=CHCHO + CH3(CH2)4CHO
2-Octenal Hexanal
+ OHC-CHO Glyoxal CH3 (CH2)4
O
2-Pentyl Furan
HO O O O
CH3 (CH2)3 (CH2)6 COOMe
13 12 10
H OH
Pentane 9-Oxo-Nonanoic acid
(2.4%) CH3(CH2)3CH3 OHC(CH2)7COOMe (3.8%)
Hexanal (45%) CH3(CH2)4CHO OHCCH=CH(CH2)6COOMe

(29%)
H2 Me 10-Oxo-8-Decenoate
O2
CH3(CH2)3CH=CHCHO
(1.1%) 2-Heptenal
Figure 14. Secondary scissions of intermediate products make important contributions to the
total mix of compounds generated during lipid oxidation, shown here for linoleic acid and esters.
Top: Oxidation and subsequent scission of radicals released in scissions of initial alkoxyl radicals
augment some of the original scission aldehydes, although by different routes, and produce
some different compounds as well, including the pentyl furan responsible for reversion flavor in
oils. Similarly, decomposition of epidioxides formed during photosensitized oxidation of linoleate
increase yields of major aldehydes and also produce longer chain aldehydes. Adapted from
(273, 314).
TERMINATION 339
This discussion would be incomplete without some mention of the most notor-
ious scission product of lipid oxidation, namely malondialdehyde (MDA). MDA is
a downstream scission product from five-membered cyclic hydroperoxides, which
can only be formed in linolenic and higher fatty acids (376, 377). Reaction 76
shows only one positional isomer of malonaldehyde, although at least four perox-
ides give comparable structures (376). Thus, formation of MDA first requires
appropriate conditions to generate cyclic peroxide precursors (251), i.e., internal
hydroperoxides, aprotic solvents, low lipid concentrations, and limited oxygen
pressures. Then conditions for cleavage of the endoperoxide must be supplied,
usually mild heat and acid (374). Yields of authentic MDA determined by GC-
MS in autoxidized fatty acids are usually less than 0.1% (374, 378), although up
to 5% MDA was found in photosensitized fatty acids (374) in which internal hydro-
peroxides are formed in high concentrations.
O H+ or O
(CH2)3COOR (CH2)3COOR
O C5H11 O C5H11
OOH OOH
(CH2)3COOR
+ C5H11
O O
OOH
76
4.3. Co-Oxidations with Non-Lipid Molecules

Lipid alkoxyl and peroxyl radicals abstract Hs from any available sources, includ-
ing nonlipid molecules such as amino acids (346, 379382), proteins (99, 383, 384),
nucleic acids (385, 386), antioxidants (318), carotenoids and other pigments, and
even carbohydrates (387). As was noted in the discussion above (Sections 3.1.1
and 3.2.1), this quenches the lipid radical and stops propagation of the immediate
radical chain. However, there is increasing evidence that the radicals transferred to
proteins and carbohydrates, in particular, may follow processes similar to lipids,
i.e., add oxygen to form peroxyl radicals that abstract Hs and initiate new radical
chains. In this way, lipids serve to broadcast oxidation damage to other mole-
cules in foods and biological systems (186, 390).
What is important in the context of termination reactions is that radicals formed
in nonlipid molecules combine with lipid radicals to generate co-oxidation products
(Reactions 77 and 78) that provide footprints of LOOH reactions (389) and
should not be ignored in consideration of lipid oxidation kinetics, mechanisms,
and overall effects in foods and biological systems. Co-oxidation products

limit extractability of lipids for analysis and, in addition, often remove lipids
from product streams normally analyzed. Consequently, these products have
probably been severely underestimated in studies of lipid oxidation in complex
systems.
Cysteine (379, 380)
LOO LOOH
77
+ RSH RS +
LO LOH
RS-OOL, RS-OL, RS-epoxy-L adducts

77a
COOH
OH HS NH2 O
eg.
OH
O
LOO LOOH
NH2 NH
+ + 78
LO LOH
>NH >N
N OOL, N OL adducts 78a
Lipid radical transfer has been demonstrated for trp, arg, his, and lys (99, 383,
384), all of which have reactive N groups on their side chains, and radical decom-
position products from these amino acids have been identified (381, 382, 390). Tyr-
osine and methionine degradation by oxidizing lipids has also been demonstrated
(390), but the intermediate radicals in the reaction may be too unstable for detec-
tion. Lipid radical adducts to amino acids are important flavor precursors (340) and
also may play critical roles in pathological processes in vivo (186, 388).
4.4. Elimination Reactions

HO and HOO can be eliminated from LOOH, respectively yielding an internal
carbonyl (ketone) (Reaction 79a) and a desaturated product with an additional
double bond (Reaction 79b) (391, 392). These are not major reactions, but never-
theless account for some of the lipid oxidation products identified under various
conditions.
EXPANDED INTEGRATED REACTION SCHEME 341
OOH
R1CH CH CH CH CH CH2 R2
OH OOH
O
R1CH CH CH CH C CH2 R2 R1CH CH CH CH CH CH R2 79
(a) (b)
5. EXPANDED INTEGRATED REACTION SCHEME
The classic free radical chain reaction mechanism used for more than five decades
to understand and track oxidation reactions was developed from product analyses
that were somewhat crude compared with the sophisticated chromatography and
spectroscopy available today. The reaction scheme is not wrong, but it may be
incomplete, at least for complex molecules such as polyunsaturated fatty acids.
Current information raises questions about the literal application of the classic
free radical chain sequence to lipid oxidation. Observed products do not match
those predicted: Many studies have now shown that hydroperoxides are not exclu-
sive products in early stages and lipid alcohols are not even major products after
hydroperoxide decomposition. Product distributions are consistent with multiple
pathways that compete with each other and change dominance with reaction con-
ditions and system composition. Rate constants show no strong preference for H
abstraction, cyclization, addition, or scission, which partially explains the mixture
of products usually observed with oxidizing lipids. It could be argued that the reac-
tions in Figure 1 accurately describe early processes of lipid oxidation, but LOO
rate constants considerably higher for cyclization than for abstraction contradict
this.
The picture emerging from integration of all these observations is that lipid oxi-
dation has multiple pathways available and that the balance of pathways taken in a
given system depends on solvent, fatty acid composition and concentration, initia-
tion mechanisms and catalysts present, temperature, oxygen pressure, and espe-
cially on availability of abstractable hydrogens from lipids and other sources.
These multiple pathways must be considered in determining appropriate analyses
for lipid oxidation, designing more effective strategies for stabilization of foods
where lipid oxidation is a major mode of deterioration, and understanding how lipid
oxidation may mediate pathological processes in vivo.
Therefore, a new integrated paradigm for lipid oxidation is proposed in which
the major alternative pathways are added to the classic free radical chain
(Figure 15). The traditional reaction sequence involving hydrogen abstractions is
presented vertically down the center of the scheme because most radicals formed
in alternative reactions ultimately abstract hydrogens to propagate the chain. This is
the core of the oxidation process. Pathways that compete with H abstraction are
AN INTEGRATED SCHEME FOR LIPID OXIDATION
LH
E, 1 e oxidation
Polymers Dimers
L
Addition Isomerization
O2 -Scission of O2
Cis trans Epidioxides-,
Dimers-
Endoperoxides-
-CH=CH- addition O2 LOO Cyclization
r
fe O2
Epoxides- ns
tra
+ n
t ro O2
LO ec
El
- Hydrogen abstraction
+/ from LH or RH
LOO+ or LOO
H+ Epoxides-
LOOH
Scission
n
tio
h Mn+ za
c li
Cy
HO + OL LO + OH + M(n+1)+
H
ab
ion str
C=C Addition
Re
a
iss cti
tion
on
co
Sc
mb
liza
LOH + L
ina
Aldehydes
Cyc
t
ion
Alkanes,
Oxo cmpds., Peroxides,
Polymers ketones
Epoxides
Scission
(Hydroxy-, hydroperoxy- )
radicals
Secondary oxidations
Figure 15. Integrated scheme for lipid oxidation accounting for multiple reactions pathways
competing with the classic hydrogen abstraction. Dotted lines indicate paths for oxygen addition
to secondary radicals formed in cyclic and addition products, with formation of new peroxyl
radicals.
shown for both peroxyl and alkoxyl radicals, and the H abstractions that are asso-
ciated with these alternative reactions and propagate the oxidation chain are either
designated specifically (dotted lines) or implied in the production of reactive radi-
cals. Cyclization and addition yield intermediate products with radicals at new sites.
These radicals can add oxygen and form peroxides that either enter the traditional
H abstraction flow, designated by the dotted lines, or undergo further addition
REFERENCES 343
or scission reactions outside the traditional scheme. Some of the products resulting
from these alternative pathways are the expected aldehydes, etc., but some are not.
Thus, alternative reaction paths increase the complexity of both the kinetics and the
product mix of lipid oxidation. In addition, an attempt has been made to distinguish
termination of individual radicals from termination of the oxidation chain by
including side radicals produced in each reaction. Products are generated by oxida-
tion and have impacts on the system, whether food or biological, but the process
nevertheless continues. Any radical deriving during lipid oxidation has the potential
to start a separate chain of its own, equivalent to the entire reaction scheme. This
approach more accurately portrays the perpetuity of lipid oxidation reactions in the
absence of antioxidants or interceptors.
This integrated scheme is a first step to broader recognition of the complexities
of lipid oxidation and should be considered a work in progress. The lack of rate
constants for lipid reactions, in itself, shows there is still much that we do not
know, and factors shifting the balance between pathways are only beginning to
be understood. The past twenty years have brought great progress in our under-
standing of the details of lipid oxidation reactions, and increasing sophistication
and sensitivity of analytical techniques promise to advance our knowledge even
faster in the next few years. Demands for increased stability in foods and control
of lipid oxidation in vivo will force us to look beyond the traditional hydroperox-
ides and consider the multiple pathways and products that may contribute critically
to system deterioration and toxic side reactions.
Hopefully, this chapter will stimulate and encourage broader consideration
of the multiple pathways of lipid oxidation, as well as more collaborative
research between food chemists, biochemists, and organic chemists to obtain the
reaction details that will ultimately be needed to control lipid oxidation in any
system.
REFERENCES
1. M. G. Simic and M. Karel, ed., Autoxidation in Food and Biological Systems, Plenum
Press, New York, 1980.
2. M. G. Simic, J. Chem. Ed., 58(2), 125131 (1981).
3. E. N. Frankel, Prog. Lipid Res., 23, 197221 (1985).
4. H. W.-S. Chan. ed., Autoxidation of Unsaturated Lipids, Academic Press, London, 1987.
5. N. A. Porter and D. G. Wujek, in A. Quintanilla, ed., Reactive Oxygen Species in
Chemistry, Biology, and Medicine, Plenum Publishing Corp., New York, 1987, pp. 55
79.
6. H. W. Gardner, Free Radic. Biol. Med., 7, 6586 (1989).
7. N. A. Porter, in C. Vigo-Pelfrey, ed., Membrane Lipid Oxidation, vol. 1, CRC Press,
Boca Raton, Florida, 1990, pp. 3362.
8. E. N. Frankel, J. Sci. Food Agric., 54, 495511 (1991).
9. M. G. Simic, S. V. Jovanovic, and E. Niki, in A. J. St. Angelo, ed., Lipid Oxidation in
Food, American Chemical Society, Washington, D.C., 1992, pp. 1432.
10. A. J. St. Angelo ed., Lipid Oxidation in Food, American Chemical Society, Washington,
D.C., 1992.
11. N. A. Porter, S. E. Caldwell, and K. A. Mills, Lipids, 30(4), 277290 (1995).
12. N. Pilpel and B. F. J. Hunter, J. Colloid Interfac. Sci., 33(4), 615622 (1970).
13. N. M. Emanuel, ed., The Oxidation Of Hydrocarbons In The Liquid Phase, Pergamon
Press, New York, 1965.
14. E. H. Farmer, Trans. Faraday Soc., 42, 228236 (1946).
15. J. L. Bolland, Quart. Rev., 3, 121 (1949).
16. J. L. Bolland, J. Chem. Soc., 445447 (1945).
17. D. Swern, in W. O. Lundberg, ed., Autoxidation and Antioxidants, vol. 1, Interscience,
New York, 1961, pp. 154.
18. E. H. Farmer, H. P. Koch, and D. A. Sutton, J. Chem. Soc., 541547 (1943).
19. S. M. Hyde and D. Verdin, Trans. Faraday Soc., 64(1), 144154 (1968).
20. J. P. Cosgrove, D. F. Church, and W. A. Pryor, Lipids, 22, 299304 (1987).
21. J. K. Kochi, in J. K. Kochi, ed., Free Radicals, vol. 2, John Wiley, New York, 1973,
pp. 665710.
22. Anonymous. Chem. & Eng. News, 1820 (1981).
23. N. Uri, in W. O. Lundberg, ed., Autoxidation and Antioxidants, Interscience, New York,
1961, pp. 55106.
24. G. Haase, and W. L. Dunkley, J. Lipid Res., 10, 561567 (1969).
25. K. M. Schaich, Lipids, 23(6), 570578 (1988).
26. K. M. Schaich, Lipids, 27(3), 209218 (1992).
27. F. W. Heaton and N. Uri, J. Lipid Res., 2(2), 152160 (1961).
28. W. A. Waters, Discussions Faraday Soc., 46, 158163 (1968).
29. P. Smith, and W. A. Waters, J. Chem. Soc. B., Phys. Org. Chem., 462467 (1969).
30. M. Hirano, K. Nakamura, and T. Morimoto, J. Chem. Soc., Perkin II, 817820 (1981).
31. W. G. Dauben, M. Lober, and D. S. Fullerton, J. Org. Chem., 34, 35873592 (1969).
32. N. Ikeda and K. Fukuzumi, J. Amer. Oil Chem. Soc., 54(3), 105109 (1977).
33. A. J. Chalk and J. F. Smith, Trans. Faraday Soc., 53, 12141234 (1957).
34. A. J. Chalk and J. F. Smith, Trans. Faraday Soc., 53, 12351245 (1957).
35. C. E. H. Bawn, Discuss. Faraday Soc., 14, 181190 (1953).
36. A. A. Clifford and W. A. Waters, J. Chem. Soc., 27962804 (1965).
37. A. A. Clifford and W. A. Waters, J. Chem. Soc. (B), 462473 (1969).
38. T. A. Cooper and W. A. Waters, J. Chem. Soc., 15381543 (1964).
39. C. E. H. Bawn, T. P. Hobin, and L. Raphael, Proc. Roy. Soc. A, 237, 313324 (1956).
40. W. A. Waters, J. Amer. Oil Chem. Soc., 48(9), 427433 (1971).
41. C. Walling, Acc. Chem. Res., 8, 125131 (1975).
42. J. F. Black, J. Amer. Chem. Soc., 100(2), 527536 (1978).
43. C. E. H. Bawn and J. E. Jolley, Proc. Royal Soc. (London), A237, 297312 (1956).
44. E. J. Y. Scott and A. W. Chester, J. Phys. Chem., 76(11), 15201524 (1972).
45. K. L. Rollick and J. K. Kochi, J. Amer. Chem. Soc., 104, 13191330 (1982).
46. C. Copping and N. Uri, Discuss. Faraday Soc., 46, 202212 (1968).
REFERENCES 345
47. T. J. Wallace and R. M. Skomorski, Chem. Ind., 348349 (1965).

48. K. Fukuzawa, K. Kishakawa, T. Tadokoro, A. Tokumura, H. Tsukatani, and J. M.
Gebicki, Arch. Biochem. Biophys., 260(1), 153160 (1988).
49. K. M. Schaich and Z. Rajfur, J. Amer. Oil Chem. Soc., in preparation.
50. W. G. Barb, J. H. Baxendale, P. George, and P. R. Hargrave, Trans. Faraday Soc., 47,
591616 (1951).
51. W. G. Barb, J. H. Baxendale, P. George, and P. R. Hargrave, Trans. Faraday Soc., 47,
462500 (1951).
52. O. K. Borggard, O. Farver, and V. S. Andersen, Acta Chem. Scand., 25, 35413543
(1971).
53. A. Banks, E. Eddie, and J. G. M. Smith, Nature (London), 190, 908909 (1961).
54. E. Ochiai, Tetrahedron, 20, 18191829 (1964).
55. A. Tkac, K. Vesely, and L. Omelka, Coll. Czech. Chem. Commun., 39, 35043514
(1974).
56. A. Tkac, K. Vesely, L. Omelka, and R. Prikryl, Coll. Czech. Chem. Commun., 40, 117
128 (1975).
57. R. Prikryl, A. Tkac, L. Omelka, and K. Vesely, Coll. Czech. Chem. Commun., 40, 104
116 (1975).
58. Y. Kamiya, S. Beaton, A. Lafortune, and K. U. Ingold, Can. J. Chem., 41, 20202032
(1963).
59. J. K. Kochi, Tetrahedron, 18, 483497 (1962).
60. R. Hiatt, T. Mill, and F. R. Mayo, J. Org. Chem., 33, 14161420 (1968).
61. W. C. Bray and M. H. Gorin, J. Amer. Chem. Soc., 54(5), 21242125 (1932).
62. C. Walling, Acc. Chem. Res., 31, 155157 (1998).
63. R. A. Leising, B. A. Brennan, L. Que Jr., B. G. Fox, and E. Munck, J. Amer. Chem. Soc.,
113, 39883990 (1991).
64. A. L. Feig and S. L. Lippard, Chem. Rev., 94, 759805 (1994).
65. J-U. Rohde, Science, 299, 10371039 (2003).
66. B. H. J. Bielski and M. J. Thomas, J. Amer. Chem. Soc., 109, 7761 (1987).
67. H. Taube, J. Gen. Physiol., 49(1) (Pt 2), 2950 (1965).
68. J. D. Rush and B. H. J. Bielski, J. Amer. Chem. Soc., 108, 523525 (1986).
69. H. Sugimoto and D. T. Sawyer, J. Amer. Chem. Soc., 106, 42834285 (1984).
70. S. Rahhal and H. W. Richter, J. Amer. Chem. Soc., 110, 31263133 (1988).
71. J. D. Rush and W. H. Koppenol, J. Amer. Chem. Soc., 110, 49574963 (1988).
72. V. K. Sharma and B. H. J. Bielski, Inorg. Chem., 30, 43064310 (1991).
73. H-C. Tung, C. Kang, and D. T. Sawyer, J. Amer. Chem. Soc., 114, 34453455 (1992).
74. B. Ensing and E. J. Baerends, J. Phys. Chem. A, 106, 79027910 (2002).
75. B. Ensing, F. Buda, and E. J. Baerends, J. Phys. Chem., 107, 57225731 (2003).
76. F. Buda, B. Ensing, M. C. M. Gribnau, and E. J. Baerends, Chem. J. Eur., 9, 34363444
(2003).
77. J. Kaizer, M. Costas, and L. Que, Jr., Agnew. Chem. Int. Ed., 42, 36713673 (2003).
78. J. Kaizer, E. J. Klinker, N. Y. Oh, J-U. Rohde, W. J. Song, A. Stubna, J. Kim, E. Munck,
W. Nam, and L. Que, Jr., J. Amer. Chem. Soc., 126, 472473 (2004).
79. S. H. Bossman, E. Oliveros, S. Gob, S. Siegwart, E. P. Dahlen, L. J. Payawan, M. Straub,

M. Worner, and A. M. Braun, J. Phys. Chem., 102, 55425550 (1998).
80. C. M. Jones and M. J. Burkitt, J. Amer. Chem. Soc., 125, 69466954 (2003).
81. A. Bassan, M. R. A. Blomberg, P. E. M. Siegbahn, and L. Que, Jr., J. Amer. Chem. Soc.,
124, 1105611063 (2002).
82. D. C. Borg, K. M. Schaich, and A. Forman, in W. Bors, M. Saran, and D. Tait, ed.,
Oxygen Radicals in Chemistry and Biology, Walter de Gruyter, Berlin, Germany, 1984,
pp. 123129.
83. A. Wada, S. Ogo, S. Nagatomo, T. Kitagawa, and Y. Watanabe, Inorg. Chem., 41(4),
616618 (2002).
84. E. Hecht, Physics: Calculus, 2nd ed., Brooks/Cole, Pacific Grove, California, 2000.
85. P. W. Atkins and J. A. Beran, General Chemistry, 2nd ed., Scientific American Books
(W.H.Freeman & Co.), New York, 1990.
86. L. W. Fine, Chemistry, Appleton-Century-Crofts, New York, 1972.
87. R. V. Bensasson, E. J. Land, and T. G. Truscott, Excited States and Free Radicals
in Biology and Medicine, Oxford Science Publications, Oxford, United Kingdom,
1993.
88. J. A. Howard, in J. K. Kochi, ed., Free Radicals, vol. 2, Wiley-Interscience, New York,
1973, pp. 362.
89. N. Baker and L. Wilson, J. Lipid Res., 7, 341348 (1966).
90. M. Geoffroy, P. Lambelet, and P. Richert, J. Agric. Food Chem, 48, 974978 (2000).
91. T. Chiba, K. Fujimoto, and T. Kaneda, J. Amer. Oil Chem. Soc., 58(5), 587590 (1981).
92. T. Chiba, K. Fujimoto, and T. Kaneda, Agric. Biol. Chem., 48(11), 27012705 (1984).
93. H. R. Rawls and P. J. Van Santen, J. Amer. Oil Chem. Soc., 47(4), 121125 (1970).
94. K. M. Schaich, CRC Crit. Rev. Food Sci. Nutr., 13, 131159 (1980).
95. Y. S. R. Sastry and G. Lakshminarayana, J. Amer. Oil Chem. Soc., 48(9), 452454
(1971).
96. J. Terao and S. Matsushita, Agric. Biol. Chem., 41(12), 24672468 (1977).
97. C. S. Foote, in W. A. Pryor, ed., Free Radicals in Biology, Vol. 2, Academic Press,
New York, 1976, pp. 85134.
98. R. W. Murray, in H. Wasserman and R. W. Murray, eds., Singlet Oxygen, Academic, New
York, 1979, pp. 59114.
99. K. M. Schaich, CRC Crit. Rev. Food Sci. Nutr., 13, 189244 (1980).
100. T. Schulte-Herbruggen and E. Cadenas, Photochem. Photobiol., 10, 3551 (1985).
101. E. A. Mayeda and A. J. Bard, J. Amer. Chem. Soc., 96, 40234024 (1974).
102. B. A. Svingen, F. O. ONeal, and S. D. Aust, Photochem. Photobiol., 28, 803809
(1978).
103. A. U. Khan, Science, 168, 476477 (1970).
104. A. U. Khan, Photochem. Photobiol., 28, 615627 (1978).
105. A. U. Khan, J. Photochem., 25, 327334 (1985).
106. E. W. Kellogg and I. Fridovich, J. Biol. Chem., 250, 88128817 (1975).
107. G. A. Russell, J. Amer. Chem. Soc., 79, 38713877 (1957).
108. B. H. J. Bielski and A. O. Allen, J. Phys. Chem., 81(11), 10481050 (1977).
REFERENCES 347
109. T. Matsuura, N. Yoshimura, A. Hishinaga, and I. Saito, Tetrahedron, 28, 49334938

(1972).
110. R. Nilsson and D. R. Kearns, J. Phys. Chem., 78, 16811683 (1974).
111. W. W. Nawar, J. Agric. Food Chem., 17(1), 1821 (1969).
112. W. W. Nawar, in O. R. Fennema, ed., Food Chemistry, 2nd ed., Academic, New York,
1966, pp. 225320.
113. R. Marcuse and P-O. Fredriksson, J. Amer. Oil Chem. Soc., 45(5), 400407 (1968).
114. T. P. Labuza, CRC Crit. Rev. Food Sci. Nutr., 2(3), 355405 (1971).
115. W. O. Lundberg and J. R. Chipault, J. Amer. Chem. Soc., 69, 833836 (1947).
116. H. Aoshima, T. Kajiwara, A. Hatanaka, and H. Hatano, J. Biochem., 82(6), 15591565
(1977).
117. M. R. Egmond, J. F. G. Vliegenthart, and J. Boldingh, Biochim. Biophys. Res. Commun.,
48(5), 10551060 (1972).
118. J. J. M. C. deGroot, G. J. Garssen, J. F. G. Vliegenthart, and J. Boldingh, Biochim.
Biophys. Acta, 377(1), 7179 (1975).
119. J. J. M. C. deGroot, G. J. Garssen, J. F. G. Vliegenthart, and J. Boldingh, FEBS Lett.,
56(1), 5054 (1975).
120. H. Aoshima, T. Kajiwara, and A. Hatanaka, Agric. Biol. Chem., 45(10), 22452251
(1981).
121. M. E. Robinson, Biochem. J., 18, 255264 (1924).
122. B. M. Watts and I. Chang, Food Technol., 3, 332336 (1949).
123. J. Kendrick and B. M. Watts, Lipids, 4, 454458 (1969).
124. B. M. Watts and B. T. Lehmann, Food Res., 17, 100108 (1952).
125. M. T. Younathon and B. M. Watts, Food Res., 24, 728734 (1959).
126. A. L. Tappel, Arch. Biochem. Biophys., 44, 378394 (1953).
127. A. L. Tappel, Food Research, 18, 560573 (1953).
128. A. L. Tappel, J. Biol. Chem., 217, 721733 (1955).
129. A. L. Tappel, in W. O. Lundberg, ed., Autoxidation and Antioxidants, vol. 1, Interscience
Publishers, New York, 1961, pp. 326366.
130. A. L. Tappel, in H. Schultz, E. D. Day, and W. O. Sinnhuber, eds., Symposium on Foods:
Lipids and Their Oxidation, AVI Publishing, Westport, Connecticut, 1962, pp. 122138.
131. V. P. Maier and A. L. Tappel, J. Amer. Oil Chem. Soc., 36, 1215 (1959).
132. V. P. Maier and A. L. Tappel, J. Amer. Oil Chem. Soc., 36, 812 (1959).
133. J. D. Love and A. M. Pearson, J. Agric. Food Chem., 22, 10311034 (1974).
134. M. M. Verma, V. Paranjape, and D. A. Ledward, Meat Sci., 14, 91104 (1985).
135. W. D. Brown, L. S. Harris, and H. S. Olcott, Arch. Biochem. Biophys., 101(1), 1420
(1963).
136. C. E. Eriksson, P. A. Olsson, and S. G. Svensson, J. Amer. Oil Chem. Soc., 48(9),
442447 (1971).
137. R. M. Kaschnitz and Y. Hatefi, Arch. Biochem. Biophys., 171, 292304 (1975).
138. J. D. Love, Food Technol., 37(7), 117120 (1983).
139. T. A. Dix, R. Fontana, A. Panthani, and L. J. Marnett, J. Biol. Chem., 260(9), 53585364
(1985).
140. K. S. Rhee, Y. A. Ziprin, and G. Ordonez, J. Agric. Food Chem., 35, 10131017 (1987).
141. R. Labeque and L. J. Marnett, J. Amer. Chem. Soc., 111, 66216627 (1989).
142. P. Gatellier, M. Anton, and M. Renerre, J. Agric. Food Chem., 43, 651656 (1995).
143. W. Nam, H. J. Han, S-Y. Oh, Y. J. Lee, M-H. Choi, S-Y. Han, C. Kim, S. K. Woo, and
W. Shin, J. Amer. Chem. Soc., 122(36), 86778684 (2000).
144. S. I. Rao, A. Wilks, M. Hamberg, and P. R. Ortiz de Montellano, J. Biol. Chem., 269,
72107216 (1994).
145. M. Hamberg, Lipids, 10(2), 8792 (1975).
146. E. Ochiai, J. Inorg. Nucl. Chem., 37, 15031509 (1975).
147. T. A. Dix and L. J. Marnett, J. Amer. Chem. Soc., 103, 67446746 (1981).
148. T. A. Dix and L. J. Marnett, J. Amer. Chem. Soc., 105, 70017002 (1983).
149. V. Ullrich, in W. Bors, M. Saran, D. Tait, eds., Oxygen Radicals in Chemistry and
Biology, Walter de Gruyter, Berlin, 1984, pp. 391404.
150. T. G. Traylor and F. Xu, J. Amer. Chem. Soc., 109, 62016202 (1987).
151. R. Labeque and L. J. Marnett, J. Amer. Chem. Soc., 109, 28282829 (1987).
152. T. C. Bruice, P. N. Balasubramanian, R. W. Lee, and J. R. Lindsay Smith, J. Amer. Chem.
Soc., 110(23), 78907892 (1988).
153. R. Labeque and L. J. Marnett, Biochemistry, 27, 70607070 (1988).
154. T. G. Traylor and F. Xu, J. Amer. Chem. Soc., 112, 178186 (1990).
155. I. Yamazaki and L. Piette, J. Amer. Chem. Soc., 113, 75887593 (1991).
156. A. L. Wilcox and L. J. Marnett, Chem. Res. Toxicol., 6, 413416 (1993).
157. Y. S. Choe, S. I. Rao, and P. R. Ortiz de Montellano, Arch. Biochem. Biophys., 314(1),
126131 (1994).
158. A. C. Onuoha, X. Zu, and J. F. Rusling, J. Amer. Chem. Soc., 119, 39793986 (1997).
159. C. P. Baron and H. J. Andersen, J. Agric. Food Chem., 50, 38873897 (2002).
160. W. Nam, M. H. Lim, H. J. Lee, and C. Kim, J. Amer. Chem. Soc, 122, 66416647 (2000).
161. M. J. Davies, Biochim. Biophys. Acta, 1077(1), 8690 (1991).
162. M. J. Davies, Free Radical Res. Commun., 10(6), 361370 (1990).
163. D. Tew and P. R. Ortiz de Montellano, J. Biol. Chem., 263, 1788017886 (1988).
164. M. R. Gunther, D. J. Kelman, J. T. Corbett, and R. P. Mason, J. Biol. Chem., 270(27),
1607516081 (1995).
165. H. Kuhn, R. Gotze, T. Schewe, and S. M. Rappaport, Eur. J. Biochem., 120, 161168
(1981).
166. D. Galaris, A. Sevanian, E. Cadenas, and R. Hochstein, Arch. Biochem. Biophys., 281,
163169 (1990).
167. W. Nam, S.-E. Park, I. K. Lim, M. H. Lim, J. Hong, and J. Kim, J. Amer. Chem. Soc, 125,
1467414675 (2003).
168. W. H. Koppenol and J. F. Liebman, J. Phys. Chem., 88, 99101 (1984).
169. N. Hogg, C. A. Rice-Evans, U. V. Darley, and M. T. Wilson, Arch. Biochem. Biophys.,
314(1), 3944 (1994).
170. P. J. OBrien and C. Little, Can. J. Biochem., 47, 493499 (1969).
171. H. B. Dunford, Free Radic. Biol. Med., 3, 405421 (1987).
172. M. P. Richards and M. A. Dettmann, J. Agric. Food Chem, 51, 38863891 (2003).
REFERENCES 349
173. S. J. Yang and W. Nam, Inorg. Chem., 37, 606607 (1998).

174. W. Nam, H. J. Choi, H. J. Han, S. H. Cho, H. J. Lee, and S-Y. Han, Chem. Commun.,
387388 (1999).
175. S. E. Lewis and E. D. Wills, Biochim. Biophys. Acta, 70, 336338 (1963).
176. K. Nakamura and T. Nishida, J. Lipid Res., 12, 149154 (1971).
177. C. P. Baron, L. H. Skibsted, and H. J. Andersen, J. Agric. Food Chem., 50, 883888
(2002).
178. C. E. Catalano, Y. S. Choe, and P. R. Ortiz de Montellano, J. Biol. Chem., 264(18),
1053410541 (1989).
179. K. Whang and I. C. Peng, J. Food Sci., 53(6), 18631865 (1988).
180. M. Williams, B. B. H. Krootjes, J. van Steveninck, and J. van der Zee, Biochim. Biophys.
Acta, 1211, 310316 (1994).
181. F. D. Gunstone and F. A. Norris, Lipids in Foods: Chemistry, Biochemistry, and
Technology, Pergamon Press, Oxford, United Kingdom, 1983.
182. B. D. Goldstein, O. J. Balchum, H. B. Demopolous, and P. S. Duke, Arch. Environ.
Health, 17, 4649 (1968).
183. O. S. Privett and M. L. Blank, J. Amer. Oil Chem. Soc., 39(11), 465469 (1962).
184. O. S. Privett and E. C. Nickell, J. Lipid Res., 58, 208211 (1963).
185. O. S. Privett and E. C. Nickell, J. Amer. Oil Chem. Soc., 43(6), 393400 (1966).
186. W. A. Pryor, Photochem. Photobiol., 28, 787801 (1978).
187. O. J. Balchum and J. S. OBrien, Arch. Environ. Health, 22, 3234 (1971).
188. E. C. Nickell, M. Albi, and O. S. Privett, Chem. Phys. Lipids, 17, 378388 (1976).
189. W. A. Pryor, J. P. Stanley, E. Blair, and G. B. Cullen, Arch. Environ. Health, 31(4),
201210 (1976).
190. M. E. Kurz and W. A. Pryor, J. Amer. Chem. Soc., 100(25), 79537959 (1978).
191. W. A. Pryor, Ann. Rev. Physiol., 48, 657667 (1986).
192. P. S. Rao, S. M. Ayres, and H. S. Mueller, Biochem. Biophys. Res. Commun., 104(4),
15321536 (1982).
193. B. H. J. Bielski, D. E. Cabelli, R. L. Arudi, and A. B. Ross, J. Phys. Chem. Ref. Data,
14(4), 10411100 (1985).
194. M. Erben-Russ, C. Michael, W. Bors, and M. Saran, Radiat. Environ. Biophys., 26,
289294 (1987).
195. L. K. Patterson and K. Hasegawa, Ber. Bunsenges. Phys. Chem., 82, 951956 (1978).
196. K. Hasegawa and L. K. Patterson, Photochem. Photobiol., 28, 817823 (1978).
197. R. D. Small, Jr., J. C. Scaiano, and L. K. Patterson, Photochem. Photobiol., 29, 4951
(1979).
198. M. Erben-Russ, C. Michael, W. Bors, and M. Saran, J. Phys. Chem., 91, 23622365
(1987).
199. M. V. Piretti, C. Cavani, and F. Zeli, Revue francaise des Corps Gras, 25(2), 7378
(1978).
200. J. M. Gebicki and B. H. J. Bielski, J. Amer. Chem. Soc., 103, 70207022 (1981).
201. R. J. Hamilton, C. Kalu, E. Prisk, F. B. Padley, and H. Pierce, Food Chem., 60(2),
193199 (1997).
202. M. J. Thomas, K. S. Mehl, and W. A. Pryor, J. Biol. Chem., 257(14), 83438347 (1982).
203. D. H. Giamalva, D. F. Church, and W. A. Pryor, J. Amer. Chem. Soc., 108, 66466651
(1986).
204. C. Schoneich and K-D. Asmus, Radiat. Environ. Biophys., 29, 263271 (1990).
205. F. H. Doleiden, S. R. Fahrenholtz, A. A. Lamola, and A. M. Trozzolo, Photochem.
Photobiol., 20(6), 519521 (1974).
206. W. A. Pryor, J. W. Lightsey, and D. F. Church, J. Amer. Chem. Soc., 104, 66856692
(1982).
207. M. G. J. Heijman, H. Nauta, and Y. K. Levine, Radiat. Phys. Chem., 26(1), 7382 (1985).
208. M. Al-Sheikhly and M. G. Simic, J. Phy. Chem., 93, 31033106 (1989).
209. Y. S. Ilan, G. Czapski, and D. Meisel, Biochim. Biophys. Acta, 430, 209224 (1976).
210. J. M. C. Gutteridge, FEBS Lett., 172(2), 245249 (1984).
211. B. H. J. Bielski, R. L. Arudi, and M. W. Sutherland, J. Biol. Chem., 258, 47594761
(1983).
212. R. Overend and G. Paraskevopoulos, J. Phys. Chem., 82(12), 13291333 (1978).
213. Y. A. Ilan, J. Rabani, and A. Henglein, J. Phys. Chem., 80, 15581562 (1976).
214. H. W. Gardner, R. Kleiman, and D. Weisleder, Lipids, 9(9), 696706 (1974).
215. P. Schieberle, Y. Trebert, J. Firl, and W. Grosch, Chem. Phys. Lipids, 41, 101116 (1986).
216. P. Schieberle, Y. Trebert, J. Firl, and W. Grosch, Chem. Phys. Lipids, 48, 281288 (1988).
217. Y. Yamamoto, S. Haga, E. Niki, and Y. Kamiya, Bull. Chem. Soc. Jpn., 57(5), 12601265
(1984).
218. K. M. Schaich and D. C. Borg, Lipids, 23(6), 570579 (1988).
219. K. M. Schaich, Unpublished data.
220. K. M. Schaich and D. C. Borg, in M. G. Simic and M. Karel, eds., Autoxidation in Food
and Biological Systems, Plenum Publishing Corp, New York, 1980, pp. 4570.
221. J. A. Kerr, Chem. Rev., 66, 465500 (1966).
222. G. Scott, Atmospheric Oxidation and Antioxidants, Elsevier, London, 1965.
223. K. U. Ingold, Acc. Chem. Res., 2, 19 (1969).
224. E. Schauenstein, J. Lipid Res., 8, 417428 (1967).
225. E. N. Frankel, in H. Schultz, E. D. Day, and W. O. Sinnhuber, eds., Symposium on Foods:
Lipids and Their Oxidation, AVI Publishing, Westport, Connecticut, 1962, pp. 5176.
226. K. A. Tallman, D. A. Pratt, and N. A. Porter, J. Amer. Chem. Soc, 123, 1182711828
(2001).
227. D. T. Coxon, K. R. Price, and H. W-S. Chan, Chem. Phys. Lipids, 28, 365378 (1981).
228. W. E. Neff, E. N. Frankel, and D. Weisleder, Lipids, 16, 439448 (1981).
229. W. A. Pryor and J. P. Stanley, J. Org. Chem., 40(24), 36153617 (1975).
230. N. A. Porter, M. O. Funk, D. Gilmore, R. Isaac, and J. Nixon, J. Amer. Chem. Soc.,
98(19), 60006005 (1976).
231. R. F. Garwood, B. P. S. Khambay, B. C. L. Weedon, and E. N. Frankel, J. Chem. Soc.
Chem. Commun., 364365 (1977).
232. E. N. Frankel, W. E. Neff, and T. R. Bessler, Lipids, 14(12), 961967 (1979).
233. J. Terao and S. Matsushita, J. Food Process. Preserv., 3, 329337 (1980).
234. F. Haslbeck and W. Grosch, Lipids, 18(1), 706713 (1983).
REFERENCES 351
235. H. W-S. Chan and G. Levett, Lipids, 12(10), 837840 (1977).

236. J. Terao and S. Matsushita, Agric. Biol. Chem., 45(3), 587593 (1981).
237. M. VanRollins and R. C. Murphy, J. Lipid Res., 25, 507517 (1984).
238. L. J. Parr and P. A. T. Swoboda, J. Food Technol., 11, 112 (1976).
239. S. L. Boyd, R. J. Boyd, Z. Shi, R. C. Barclay, and N. A. Porter, J. Amer. Chem. Soc., 115,
687693 (1993).
240. J. A. Howard and K. U. Ingold, Can J. Chem., 46, 26552660 (1968).
241. E. G. Janzen, F. J. Johnston, and C. A. Ayers, J. Amer. Chem. Soc., 89, 11761183 (1967).
242. I. M. Campbell, R. B. Caton, and D. N. Crozier, Lipids, 9(11), 916920 (1974).
243. H. W-S. Chan, G. Levett, and J. A. Matthew, Lipids, 24, 245256 (1979).
244. N. A. Porter and J. R. Nixon, J. Amer. Chem. Soc., 100(22), 71167117 (1978).
245. N. A. Porter, D. H. Roberts, and C. B. Ziegler, Jr., J. Amer. Chem. Soc., 102, 59125913
(1980).
246. N. A. Porter, B. A. Weber, H. Weenan, and J. A. Kahn, J. Amer. Chem. Soc., 102,
55975601 (1980).
247. N. A. Porter, L. S. Lehman, B. A. Weber, and K. J. Smith, J. Amer. Chem. Soc., 103,
64476455 (1981).
248. H. Weenan and N. A. Porter, J. Amer. Chem. Soc., 104, 52165221 (1982).
249. A. N. Roe, A. T. McPhail, and N. A. Porter, J. Amer. Chem. Soc., 105, 11991203 (1983).
250. N. A. Porter and D. G. Wujek, J. Amer. Chem. Soc., 106, 26262629 (1984).
251. N. A. Porter, L. S. Lehman, and D. G. Wujek, in W. Bors, M. Saran, and D. Tait, eds.,
Oxygen Radicals in Chemistry and Biology, Walter de Gruyter, Berlin, 1984,
pp. 235237.
252. N. A. Porter, Acc. Chem. Res., 19, 262268 (1986).
253. N. A. Porter and J. S. Wujek, J. Org. Chem., 52, 50855089 (1987).
254. K. A. Mills, S. E. Caldwell, G. R. Dubay, and N. A. Porter, J. Amer. Chem. Soc., 114,
96899691 (1992).
255. N. A. Porter, K. A. Mills, S. E. Caldwell, and G. R. Dubay, J. Amer. Chem. Soc., 116,
66976705 (1994).
256. N. A. Porter, K. A. Mills, and R. L. Carter, J. Amer. Chem. Soc., 116, 66906696 (1994).
257. D. A. Pratt, J. H. Mills, and N. A. Porter, J. Amer. Chem. Soc, 125, 58015810 (2003).
258. J. A. Howard and K. U. Ingold, Can. J. Chem., 45, 793802 (1967).
259. R. A. Sheldon and J. K. Kochi, Metal-Catalyzed Oxidations of Organic Compounds,
Academic Press, New York, 1981.
260. M. Simic, in G. E. Adams E. M. Fielden and B. D. Michael, eds., in Fast Processes in
Radiation Chemistry and Biology, Wiley, New York, 1975, pp. 162179.
261. W. Bors, M. Erben-Russ, and M. Saran, J. Electroanal. Chem., 232, 3749 (1987).
262. T. M. Bockman, S. M. Hubig, and J. K. Kochi, J. Amer. Chem. Soc., 120, 28262830 (1998).
263. C. Walling and A. Padwa, J. Amer. Chem. Soc., 85, 15931597 (1963).
264. A. Factor, C. A. Russell, and T. G. Traylor, J. Amer. Chem. Soc., 87, 36923696 (1965).
265. H. W-S. Chan, J. A. Matthew, and D. T. Coxon, J. Chem Soc. Chem. Commun., 235236
(1980).
266. R. K. Haynes and S. C. Vonwiller, J. Chem. Soc. Chem. Commun., 11021104 (1990).
267. H. W. Gardner, in H. W-S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic

Press, London, 1987, pp. 5193.
268. N. A. Porter, P. J. Zuraw, and J. A. Sullivan, J. Org. Chem., 49(8), 13451348
(1984).
269. J. E. Baldwin, J. Chem. Soc. Chem. Commun., 734736 (1975).
270. M. O. Funk, R. Isaac, and N. A. Porter, J. Amer. Chem. Soc., 97, 12811282 (1975).
271. M. Hamberg and B. Samuelsson, J. Amer. Chem. Soc., 88, 23492350 (1966).
272. N. A. Porter and M. O. Funk, J. Org. Chem., 40, 36143615 (1975).
273. E. N. Frankel, W. E. Neff, E. Selke, and D. Weisleder, Lipids, 17(1), 1118 (1982).
274. E. D. Mihelich, J. Amer. Chem. Soc., 102, 71437145 (1980).
275. D. E. Van Sickle, F. R. Mayo, E. S. Gould, and R. M. Arluck, J. Amer. Chem. Soc., 89(4),
977984 (1967).
276. W. E. Neff, E. N. Frankel, and K. Fujimoto, J. Amer. Oil Chem. Soc., 65(4), 616623
(1988).
277. R. Hiatt and T. McCarrick, J. Amer. Chem. Soc., 97, 52345237 (1975).
278. F. R. Mayo, J. Amer. Chem. Soc., 80, 25002507 (1958).
279. F. R. Mayo, Acc. Chem. Res., 1(7), 193201 (1968).
280. I. H. Elson, S. W. Mao, and J. K. Kochi, J. Amer. Chem. Soc., 97(2), 335341 (1975).
281. S. E. Lewis and F. R. Mayo, J. Amer. Chem. Soc., 70(4), 15331536 (1958).
282. J. K. Kochi, J. Amer. Chem. Soc., 84, 27852793 (1962).
283. D. G. Hendry and D. Schuetzle, J. Org. Chem., 41, 31793182 (1976).
284. I. A. Witting, S. S. Chang, and F. A. Kummerow, J. Amer. Oil Chem. Soc., 34, 470473
(1957).
285. R. P. Sims and W. H. Hoffman, in W. O. Lundberg, ed., Autoxidation and Antioxidants
vol. 2, Interscience, London, 1962, pp. 629694.
286. O. S. Privett, J. Amer. Oil Chem. Soc., 36(10), 507512 (1959).
287. R. Hiatt, C. W. Gould, and F. R. Mayo, J. Org. Chem., 29, 34613466 (1964).
288. K. Miyashita, K. Fujimoto, and T. Kaneda, Agric. Biol. Chem., 46(3), 751755 (1982).
289. T. L. Mounts, D. J. McWeeny, C. D. Evans, and H. J. Dutton, Chem. Phys. Lipids, 4,
197202 (1970).
290. E. N. Frankel, W. E. Neff, E. Selke, and D. D. Brooks, Lipids, 23(4), 295298 (1988).
291. J. R. Thomas, J. Amer. Chem. Soc., 87, 39353940 (1965).
292. D. A. Lindsay, J. A. Howard, E. C. Horswill, L. Iton, K. U. Ingold, T. Cobbley, and A. Li,
Can. J. Chem., 51(6), 870880 (1973).
293. K. Adamic, J. A. Howard, and K. U. Ingold, Can. J. Chem., 47, 38033808 (1969).
294. T. G. Traylor and C. A. Russell, J. Amer. Chem. Soc., 87(16), 36983706 (1965).
295. C. Walling, H. P. Waits, J. Milanovic, and C. G. Pappiaonnou, J. Amer. Chem. Soc.,
92(16), 49274932 (1970).
296. R. Hiatt and T. G. Traylor, J. Amer. Chem. Soc., 87(16), 37663768 (1965).
297. H. W. S. Chan, F. A. A. Prescott, and P. A. T. Swoboda, J. Amer. Oil Chem. Soc., 53(9),
572576 (1976).
298. S. S. Kim, S. Y. Kim, S. S. Ryou, C. S. Lee, and K. H. Yoo, J. Org. Chem., 58, 192196
(1993).
REFERENCES 353
299. W. Bors, C. Michael, and M. Saran, Biochim. Biophys. Acta, 796, 312319 (1984).
300. E. Bascetta, F. D. Gunstone, C. M. Scrimgeour, and J. C. Walton, J. Chem. Soc. Chem.
Commun., 110112 (1982).
301. R. Hiatt and L. Zigmund, Can. J. Chem., 48, 39673969 (1970).
302. R. Hiatt and S. Szilagyi, Can. J. Chem., 48, 616627 (1970).
303. B. C. Gilbert, R. G. G. Holmes, and R. O. C. Norman, J. Chem. Res., 7, 19491973 (1977).
304. C. Walling and A. Padwa, J. Amer. Chem. Soc., 1963, 85, 15971601 (1963).
305. Y. P. Tsentalovich, L. V. Kulik, N. P. Gritsan, and A. V. Yurkovskaya, J. Phys. Chem.,
102, 79757980 (1998).
306. D. V. Avila, K. U. Ingold, and J. Kusztyk, J. Amer. Chem. Soc., 117, 29292930 (1995).
307. H. Paul, R. D. Small, Jr., and J. C. Scaiano, J. Amer. Chem. Soc., 100(14), 45204527
(1978).
308. W. Bors, D. Tait, C. Michel, M. Saran, and M. Erben-Russ, Isr. J. Chem., 24, 1724
(1984).
309. J. K. Kochi, J. Amer. Chem. Soc., 84, 1193 (1962).
310. C. Walling and P. J. Wagner, J. Amer. Chem. Soc., 86, 33683375 (1964).
311. W. E. Neff, E. N. Frankel, C. R. Schofield, and D. Weisleder, Lipids, 13(6), 415421
(1978).
312. E. Selke, E. N. Frankel, and W. E. Neff, Lipids, 13(7), 511513 (1978).
313. F. R. Mayo, J. Amer. Chem. Soc., 80, 24972500 (1958).
315. G-S. Wu, R. A. Stein, and J. F. Mead, Lipids, 12(11), 971978 (1977).
316. G-S. Wu, R. A. Stein, and J. F. Mead, Lipids, 13(8), 517524 (1978).
317. B. Acott and A. L. J. Beckwith, Austral. J. Chem., 17(12), 13421353 (1964).
318. H. W. Gardner, K. Eskins, G. W. Grams, and G. R. Inglett, Lipids, 7(5), 32434 (1972).
319. H. W. Gardner and R. Kleiman, Biochim. Biophys. Acta, 665, 113125 (1981).
320. H. W. Gardner and R. Kleiman, Biochim. Biophys. Acta, 665, 126133 (1981).
321. W. E. Neff, Lipids, 15(8), 587590 (1980).
322. H. W. Gardner and R. Kleiman, Lipids, 14(10), 848851 (1979).
323. J. K. Kochi, ed., Free Radicals, vol. 1, John Wiley and Sons, New York, 1973.
324. C. Walling and W. Thaler, J. Amer. Chem. Soc., 83, 38773884 (1961).
325. C. A. Russell, J. Amer. Chem. Soc., 21, 300302 (1959).
326. C. Walling and P. J. Wagner, J. Amer. Chem. Soc., 85, 23332334 (1963).
327. D. V. Avila, C. E. Brown, K. U. Ingold, and J. Lusztyk, J. Amer. Chem. Soc., 115, 466
470 (1993).
328. B. C. Gilbert, P. D. R. Marshall, R. O. C. Norman, N. Pineda, and P. S. Williams, J.
Chem. Soc., 13921400 (1981).
329. E. H. Farmer and D. A. Sutton, J. Chem. Soc., 119122 (1943).
330. P. Neta, M. Dizdaroglu, and M. Simic, Sr. J. Chem., 24, 2528.
331. A. Baignee, J. A. Howard, J. C. Scaiano, and L. C. Stewart, J. Amer. Chem. Soc., 105,
61206123 (1983).
332. A. G. Neville, C. E. Brown, D. M. Rayner, J. Lusztyk, and K. U. Ingold, J. Amer. Chem.
Soc., 111, 92699270 (1989).
333. E. Baciocchi, M. Bietti, M. Salamone, and S. Steenken, J. Org. Chem, 67, 22662270
(2002).
334. A. Loidl-Stahlhofen and G. Spiteller, Biochim. Biophys. Acta, 1211, 156160 (1994).
335. W. I. Kimoto and A. M. Gaddis, J. Amer. Oil Chem. Soc., 51(7), 307311 (1974).
336. A. Boto, D. Hernandez, R. Hernandez, and E. Suarez, J. Org. Chem, 68, 53105319
(2003).
337. R. Ellis, A. M. Gaddis, G. T. Currie, and S. L. Powell, J. Amer. Oil Chem. Soc., 45(8),
553559 (1968).
338. J. H. Horner, S-Y. Choi, and M. Newcomb, Org. Lett., 2(21), 33693372 (2000).
339. W. H. Richardson, M. B. Yelvington, A. H. Andrist, E. W. Ertley, R. S. Smith, and
T. D. Johnson, J. Org. Chem., 38, 42194225 (1973).
340. W. Grosch, in H. W-S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic Press,
London, 1987, pp. 95139.
341. E. N. Frankel, Chem. Phys. Lipids, 44, 7385 (1987).
342. S. Fukuzumi and J. K. Kochi, J. Amer. Chem. Soc., 103, 27832791 (1981).
343. S. Wilsey, P. Dowd, and K. N. Houk, J. Org. Chem., 64, 88018811 (1999).
344. G. A. Russell, J. Amer. Chem. Soc., 78, 10471054 (1956).
345. E. N. Frankel, W. E. Neff, and E. Selke, Lipids, 19(10), 790800 (1984).
346. S. B. El-Magoli, M. Karel, and S. Yong, J. Food Biochemistry, 3, 111123 (1980).
347. E. N. Frankel, J. Amer. Oil Chem. Soc., 61(12), 19081917 (1984).
348. E. N. Frankel, W. E. Neff, W. K. Rodwedder, B. P. S. Khambay, R. F. Garwood, and
B. C. L. Weedon, Lipids, 12, 10551061 (1977).
349. E. N. Frankel, W. E. Neff, W. K. Rohwedder, B. P. S. Khambay, R. F. Garwood, and
B. C. L. Weedon, Lipids, 12(11), 901907 (1977).
350. L. Bateman, H. Hughes, and A. L. Morris, Discuss. Faraday Soc., 14, 190199 (1953).
351. J. Sliwiok, T. Kowalska, W. Kowalski, and A. Biernat, Microchem. J., 19(4), 362372
(1974).
352. M. Hicks and J. M. Gebicki, Int. J. Radiat. Biol., 64(2), 143148 (1993).
353. C. Walling and L. Heaton, J. Amer. Oil Chem. Soc., 87, 4851 (1965).
354. J. A. Howard and J. E. Bennett, Can. J. Chem., 50(14), 23742377 (1972).
355. M. H. Brodnitz, Agric. Food Chem., 16(6), 994999 (1968).
356. G. A. Russell, J. Chem. Ed., 36, 111118 (1959).
357. D. E. Van Sickle, F. R. Mayo, R. M. Arluck, and M. A. Syz, J. Amer. Chem. Soc., 89(4),
967977 (1967).
358. L. K. Patterson and J. L. Redpath, in K. L. Mittal, ed., Micellization, Solubilization, and
Microemulsions, vol. 2, Plenum Pressm, New York, 1977, pp. 589601.
359. W. Bors, M. Erben-Russ, and M. Saran, Bioelectrochem. Bioenerg., 18, 3749 (1987).
360. J. A. Howard, W. J. Schwalm, and K. U. Ingold, Adv. Chem. Series, 75, 623 (1968).
361. J. A. Howard and K. U. Ingold, Can. J. Chem., 46, 26612666 (1968).
362. J. A. Howard and K. U. Ingold, J. Amer. Chem. Soc., 90, 10581059 (1968).
363. H. De Groot and T. Noll, Chem. Phys. Lipids, 44, 209226 (1987).
364. J. A. Howard, in R. F. Gould, ed., Organic Free Radicals, American Chemical Society,
Washington, D.C., 1978, pp. 413432.
REFERENCES 355
365. J. A. Howard, K. Adamic, and K. U. Ingold, Can. J. Chem., 47, 37933795 (1969).
366. J. A. Howard and K. U. Ingold, J. Amer. Chem. Soc., 90, 10561058 (1968).
367. P. Scheiberle, B. Tsoukalas, and W. Grosch, Z. Lebensm. Unters. Forsch., 168, 448456
(1979).
368. P. Scheiberle, W. Grosch, H. Kexel, and H-L. Schmidt, Biochim. Biophys. Acta, 666,
322326 (1981).
369. W. Grosch and J. Megele, in W. Bors, M. Saran, and D. Tait, ed., Oxygen Radicals in
Chemistry and Biology, Walter de Gruyter & Co., Berlin, 1984, pp. 249256.
370. P. Schieberle and W. Grosch, Z. Lebensm. Unters Forsch, 173, 192198 (1981).
371. P. Schieberle and W. Grosch, Z. Lebensm. Unters Forsch, 173, 199203 (1981).
372. H. W-S. Chan, G. Levett, and J. A. Matthew, J. Chem. Soc. Chem. Commun., 756758
(1978).
373. A. U. Khan, J. Phys. Chem., 80(20), 22192228 (1976).
374. E. N. Frankel and W. E. Neff, Biochim. Biophys. Acta, 754, 264270 (1983).
375. W. E. Neff, E. N. Frankel, and D. Weisleder, Lipids, 17(11), 780790 (1982).
376. W. A. Pryor, J. P. Stanley, and E. Blair, Lipids, 11(5), 370379 (1976).
377. L. K. Dahle, E. G. Hill, and R. T. Holman, Arch. Biochem. Biophys., 98(2), 253261
(1962).
378. R. J. Shamberger, B. A. Shamberger, and C. E. Willis, J. Nutr., 107, 14041409 (1977).
379. H. W. Gardner and D. Weisleder, Lipids, 11(2), 12734 (1976).
380. H. W. Gardner, R. Kleiman, D. Weisleder, and G. E. Inglett, Lipids, 12(8), 655660
(1977).
381. S. H. Yong and M. Karel, J. Amer. Oil Chem. Soc., 55(3), 352357 (1978).
382. S. H. Yong and M. Karel, J. Food Sci., 22(2), 568574 (1979).
383. K. M. Schaich and M. Karel, J. Food Sci., 40, 456459 (1975).
384. K. M. Schaich and M. Karel, Lipids, 11, 392400 (1976).
385. M-H.Yang, and K. M. Schaich, Free Radic. Biol. Med., 20, 225236 (1996).
386. K. M. Schaich and D. C. Borg, in W. Bors, M. Saran, and D. Tait, eds., Oxygen Radicals
in Chemistry and Biology, Walter de Gruyter, Berlin, 1984, pp. 603606.
387. H. W. Gardner, in J. W. Finley, ed., Xenobiotics in Foods and Feeds, American Chemical
Society, Washington, D.C., 1983, pp. 6384.
388. D. C. Borg and K. M. Schaich, Isr. J. Chem., 24, 3853 (1984).
389. W. A. Pryor, Free Radic. Biol. Med., 7, 177178 (1989).
390. M. Karel, K. M. Schaich, and R. B. Roy, Agric. Food Chem., 23, 159164 (1975).
391. J. Terao, T. Ogawa, and S. Matsushita, Agric. Biol. Chem., 39(2), 397402 (1975).
392. E. Bothe, M. N. Schuchmann, D. Schulte-Frohlinde, and C. von Sonntag, Photochem.
Photobiol., 28, 639644 (1978).
8
Lipid Oxidation:
Measurement Methods
Fereidoon Shahidi and Ying Zhong
Memorial University of Newfoundland,
St. Johns, Newfoundland, Canada
1. INTRODUCTION
Dietary lipids, naturally occurring in raw food materials or added during food
processing, play an important role in food nutrition and flavor. Meanwhile, lipid
oxidation is a major cause of food quality deterioration, and has been a challenge
for manufacturers and food scientists alike. Lipids are susceptible to oxidative
processes in the presence of catalytic systems such as light, heat, enzymes, metals,
metalloproteins, and micro-organisms, giving rise to the development of off-flavors
and loss of essential amino acids, fat-soluble vitamins, and other bioactives. Lipids
may undergo autoxidation, photo-oxidation, thermal oxidation, and enzymatic
oxidation under different conditions, most of which involve some type of free radi-
cal or oxygen species (1, 2). Among these, only autoxidation and thermal oxidation
are discussed here in detail.
Autoxidation is the most common process leading to oxidative deterioration and
is defined as the spontaneous reaction of atmospheric oxygen with lipids (3). The
process can be accelerated at higher temperatures, such as those experienced during
deep-fat frying, which is called thermal oxidation, with increases in free fatty acid
and polar matter contents, foaming, color, and viscosity (4). Unsaturated fatty acids
357
358 LIPID OXIDATION: MEASUREMENT METHODS
are generally the reactants affected by such reactions, whether they are present as
free fatty acids, triacylglycerols (as well as diacyglycerols or monoacylglycerols),
or phospholipids (3). It has been accepted that both autoxidation and thermal
oxidation of unsaturated fatty acids occurs via a free radical chain reaction that
proceeds through three steps of initiation, propagation, and termination (5). A
simplified scheme explaining the mechanism of autoxidation is given below:
Initiation:
initiator
LH L + H
Propagation:
L + O2 LOO
LOO + LH LOOH + L
Termination:
2 LOO
LOO + L Nonradical products
L + L
As oxidation normally proceeds very slowly at the initial stage, the time to reach a
sudden increase in oxidation rate is referred to as the induction period (6). Lipid
hydroperoxides have been identified as primary products of autoxidation; decom-
position of hydroperoxides yields aldehydes, ketones, alcohols, hydrocarbons, vola-
tile organic acids, and epoxy compounds, known as secondary oxidation products.
These compounds, together with free radicals, constitute the bases for measurement
of oxidative deterioration of food lipids. This chapter aims to explore current
methods for measuring lipid oxidation in food lipids.
2. METHODS FOR MEASURING LIPID OXIDATION
Numerous analytical methods are routinely used for measuring lipid oxidation in
foods. However, there is no uniform and standard method for detecting all oxidative
changes in all food systems (7). Therefore, it is necessary to select a proper and
adequate method for a particular application. The available methods to monitor
lipid oxidation in foods can be classified into five groups based on what they mea-
sure: the absorption of oxygen, the loss of initial substrates, the formation of free
radicals, and the formation of primary and secondary oxidation products (8). A
number of physical and chemical tests, including instrumental analyses, have
been employed in laboratories and the industry for measurement of various lipid
oxidation parameters. These include the weight-gain and headspace oxygen uptake
method for oxygen absorption; chromatographic analysis for changes in reactants;
MEASUREMENT OF OXYGEN ABSORPTION 359
iodometric titration, ferric ion complexes, and Fourier transform infrared (FTIR)
method for peroxide value; spectrometry for conjugated dienes and trienes, 2-thio-
barbituric acid (TBA) value, p-anisidine value (p-AnV), and carbonyl value;
Rancimat and Oxidative Stability Instrument (OSI) method for oil stability index;
and electron spin resonance (ESR) spectrometric assay for free-radical type and
concentration. Other techniques based on different principles, such as differential
scanning calorimetry (DSC) and nuclear magnetic resonance (NMR), have also
been used for measuring lipid oxidation. In addition, sensory tests provide subjec-
tive or objective evaluation of oxidative deterioration, depending on certain details.
3. MEASUREMENT OF OXYGEN ABSORPTION
3.1. Weight Gain

Consumption of oxygen during the initial stage of autoxidation results in an
increase in the weight of fat or oil, which theoretically reflects its oxidation level.
Heating an oil and periodically testing for weight gain is one of the oldest methods
for evaluating oxidative stability (9). This method requires simple equipment and
directly indicates oxygen absorption through mass change. Oil samples are weighed
and stored in an oven at a set temperature with no air circulation. To avoid the influ-
ence of mass change by volatiles, samples can be preheated in an inert atmosphere.
Samples are then taken out of the oven at different time intervals, cooled to ambient
temperature, and reweighed; the weight gain is then recorded. The induction period
can be obtained by plotting weight gain against storage time. In some cases, the
time required to attain a 0.5% weight increase is taken as an index of oil stability
(7, 9, 10).
As a physical method for measuring lipid oxidation, the weight-gain method has
several drawbacks such as discontinuous heating of the sample, which may give rise
to non-reproducible results, and requiring long analysis time and intensive human
participation (7). Nevertheless, this method offers advantages such as low instru-
mentation cost as well as a high capacity and processing speed of samples without
limitation (7). Antolovich et al. (9) suggested that this technique may be extended
to more sophisticated continuous monitoring of mass and energy changes as in ther-
mogravimetry (TG)/differential scanning calorimetry (DSC). The weight-gain
method can also be used for measuring antioxidant activity by comparing the
results in the presence and absence of an antioxidant. Nevertheless, this method
is useful only when highly unsaturated oils, such as marine oils and vegetable
oils containing a high content of polyunsaturated fatty acids, are examined.
3.2. Headspace Oxygen Uptake

In addition to the weight-gain method, oxygen consumption can be measured
directly by monitoring the drop of oxygen pressure. Using headspace oxygen meth-
od, an oil sample is placed in a closed vessel also containing certain amount of oxy-
gen at elevated temperatures, commonly around 100 C. The pressure reduction in
the vessel, which is due to the oxygen consumption, is monitored continuously and
recorded automatically. The induction period as the point of maximum change in
rate of oxygen uptake can be calculated (11). A commercial instrument for this
method, known as Oxidograph, is available. In the Oxidograph, the pressure change
in the reaction vessel is measured electronically by means of pressure transducers
(7, 12).
Oxygen consumption can also be measured by electrochemical detection of
changes in oxygen concentration. However, the analysis of the graphical data
obtained has been the bottleneck for this technique. The use of a semiautomatic
polarographic method has been proposed as an improvement for evaluation of lipid
oxidation by determination of oxygen consumption (13). As described by Genot
et al. (13), this method is based on use of two oxygen meters with microcathode
oxygen electrodes, coupled to a computerized data collection and processing unit.
The headspace oxygen method is simple and reproducible and may be the best
analytical method to evaluate the oxidative stability of fats and oils (14). Its appli-
cation in measurement of lipid oxidation in food products other than fats and oils,
however, is limited because protein oxidation also absorbs oxygen (15).
4. MEASUREMENT OF REACTANT CHANGE
Lipid oxidation can also be assessed by quantitatively measuring the loss of initial
substrates. In foods containing fats or oils, unsaturated fatty acids are the main
reactants whose composition changes significantly during oxidation. Changes in
fatty acid composition provide an indirect measure of the extent of lipid oxidation
(15). In this method, lipids are extracted from food, if necessary, and subsequently
converted into derivatives suitable for chromatographic analysis (7). Fatty acid
methyl esters (FAME) are the derivatives frequently used for determination of fatty
acid composition, usually by gas chromatography (GC) (16). Similarly, iodine
value, which reflects the loss of unsaturation, can also be used as an index of lipid
oxidation (17).
Measurement of changes in fatty acid composition is useful for identification of
lipid class and fatty acids that are involved in oxidation reactions (7). However,
because the distribution of unsaturated fatty acids varies in different food systems,
for instance, the highly unsaturated fatty acids being located predominantly in
phospholipids of muscle foods, separation of lipids into neutral, glycolipid, phos-
pholipid, and other classes may be necessary (7, 15). Moreover, it is an insensitive
way of assessing oxidative deterioration. For comparison through calculation, oxi-
dation of 0.4% polyunsaturated fatty acids to monohydroperoxides would represent
a change of 16 meq oxygen/kg oil in peroxide value, whereas a change of less than
1.0 meq oxygen/kg oil could readily be detected by measuring peroxide value (12).
Additionally, the application of this method is limited because of its inability
to serve as an indicator of oxidation of more saturated lipids (7). Nevertheless,
its usefulness for measuring oxidation of highly unsaturated oils cannot be
underestimated.
MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION 361
5. MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION
5.1. Peroxide Value (PV)

Lipid oxidation involves the continuous formation of hydroperoxides as primary
oxidation products that may break down to a variety of nonvolatile and volatile
secondary products (8, 15). The formation rate of hydroperoxides outweighs their
rate of decomposition during the initial stage of oxidation, and this becomes
reversed at later stages. Therefore, the peroxide value (PV) is an indicator of the
initial stages of oxidative change (18). However, one can assess whether a lipid
is in the growth or decay portion of the hydroperoxide concentration by monitoring
the amount of hydroperoxides as a function of time (7).
Analytical methods for measuring hydroperoxides in fats and oils can be classi-
fied as those determining the total amount of hydroperoxides and those based on
chromatographic techniques giving detailed information on the structure and the
amount of specific hydroperoxides present in a certain oil sample (8). The PV repre-
sents the total hydroperoxide content and is one of the most common quality indi-
cators of fats and oils during production and storage (9, 18). A number of methods
have been developed for determination of PV, among which the iodometric titra-
tion, ferric ion complex measurement spectrophotometry, and infrared spectroscopy
are most frequently used (19).
5.1.1. Iodometric Titration Method Iodometric titration assay, which is based

on the oxidation of the iodide ion (I) by hydroperoxides (ROOH), is the basis of
current standard methods for determination of PV (9). In this method, a saturated
solution of potassium iodide is added to oil samples to react with hydroperoxides.
The liberated iodine (I2) is then titrated with a standardized solution of sodium thio-
sulfate and starch as an endpoint indicator (7, 9, 20). The PV is obtained by calcu-
lation and reported as milliequivalents of oxygen per kilogram of sample (meq/kg).
The official determination is described by IUPAC (21). Chemical reactions involved
are given below:
ROOH 2H 2KI ! I2 ROH H2 O 2K

I2 2NaS2 O3 ! Na2 S2 O6 2NaI
Although iodometric titration is the most common method for measurement of PV,
it suffers from several disadvantages. The procedure is time-consuming and labor-
intensive (18). As described by Ruiz et al. (18), the assay includes six steps: accu-
rate weighing of the sample, dissolution of lipids in chloroform, acidification with
acetic acid, addition of potassium iodide, incubation for exactly 5 minutes, and
titration with sodium thiosulfate. This technique requires a large amount of sample
and generates a significant amount of waste (18, 22, 23). Furthermore, possible
absorption of iodine across unsaturated bonds and oxidation of iodide by dissolved
oxygen are among potential drawbacks of this method (7, 9). Besides, lack of sen-
sitivity, possible interferences, and difficulties in determining the titration endpoint
are also the main limitations (8, 23). To overcome these drawbacks, novel methods
based on the same reaction have been developed, in which some other techniques
are adopted as modification of the classical iodometric assay. Techniques such as
colorimetric determination at 560 nm (24), potentiometric endpoint determination
(25), and spectrophotometric determination of the I 3 chromophore at 290 nm or
360 nm (26, 27) have been proposed. In addition, an electrochemical technique
has been used as an alternative to the titration step in order to increase the sensitiv-
ity for determination of low PV by reduction of the released iodine at a platinum
electrode maintained at a constant potential (7).
5.1.2. Ferric Ion Complexes Other chemical methods based on the oxidation of
ferrous ion (Fe2) to ferric ion (Fe3) in an acidic medium and the formation of
iron complexes have also been widely accepted. These methods spectrophotometri-
cally measure the ability of lipid hydroperoxides to oxidize ferrous ions to ferric
ions, which are complexed by either thiocyanate or xylenol orange (23, 28, 29).
Ferric thiocyanate is a red-violet complex that shows strong absorption at 500
510 nm (8). The method of determining PV by coloremetric detection of ferric thio-
cyanate is simple, reproducible, and more sensitive than the standard iodometric
assay, and has been used to measure lipid oxidation in milk products, fats, oils,
and liposomes (8, 23).
The ferrous oxidation of xylenol orange (FOX) assay uses dye xylenol orange to
form a blue-purple complex with a maximum absorption at 550600 nm (8). This
method is rapid, inexpensive, and not sensitive to ambient oxygen or light (30). It
can consistently quantify lower hydroperoxide levels; and good agreement exists
between the FOX assay and the iodometric method (30). The FOX method has
been successfully adapted to a variety of applications. However, because many fac-
tors, such as the amount of sample, solvent used, and source of xylenol orange, may
affect the absorption coefficient, knowledge of the nature of hydroperoxides present
in the sample, and careful control of the conditions used are required for accurate
measurements (8).
5.1.3. Fourier Transform Infrared Spectroscopy (FTIR) It has been recog-

nized that hydroperoxides can quantitatively be determined by IR spectroscopy
via measurement of their characteristic O-H stretching absorption band (31). An
absorption band at about 2.93 mm indicates the generation of hydroperoxides,
whereas the replacement of a hydrogen atom on a double bond or polymerization
accounts for the disappearance of a band at 3.20 mm. The formation of aldehydes,
ketones, or acids gives rise to an extra band at 5.72 mm. Furthermore, cis-, trans-
isomerization and formation of conjugated dienes can be detected through the
changes in the absorption band in the range of 10 mm to 11 mm (7).
A rapid Fourier transform infrared spectroscopy (FTIR) method based on the
stoichiometric reaction of triphenylphosphine (TPP) with hydroperoxides has
been developed and successfully applied to determination of PV of edible oils
(32). The hydroperoxides present in oil samples react stoichiometrically with
TPP to produce triphenylphosphine oxide (TPPO), which has an intense absorption
MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION 363
band at 542 cm1 in the mid-IR spectrum (8, 18). The band intensity is measured
and converted to peroxide value. The chemical reaction involved is given below:
ROOH TPP ! TPP O ROH
By using tert-butyl hydroperoxide spiked oil standards and evaluation of the band
formed at 542 cm1, a linear calibration graph covering the range of 1100 PV was
obtained (18). More recently, disposable polymer IR (PIR) cards have been used as
sample holders where unsaturated oil samples oxidize at a fairly rapid rate (33). In
the FTIR/PIR card method, warm air continuously flows over the sample allowing
oxidation to be monitored at moderate temperatures. At periodic intervals, indivi-
dual cards are removed and the FTIR spectra scanned (33). Another new FTIR
approach uses flow injection analysis (FIA), which offers exact and highly repro-
ducible timing of sample manipulation and reaction as well as a closed environment
with oxygen and light being easily excluded (18).
The FTIR spectroscopy is a simple, rapid, and highly precise method. It shows
excellent correlation with the iodometric method and avoids the solvent and reagent
disposal problems associated with the standard wet chemical method (18, 32). The
FTIR method provides an automated, efficient and low-cost means of evaluating
oxidation in oils undergoing thermal stress and has gained considerable interest
for quality control in the industry (8, 20, 34). However, there is a need to charac-
terize the spectral changes, assign wavelengths to more common molecular species
produced, and access potential spectral cross interferences (20). Recently, an
improved Fourier transform infrared attenuated total reflectance (FTR-ATR) meth-
od using the whole FTIR spectral data instead of particular wavenumbers has been
proposed (34).
In addition to the three major methods discussed above, other techniques have
also been employed in determination of PV, such as chemiluminescence and chro-
matography. Chemiluminescence method is based on detecting the chemilumines-
cent products generated during the reaction of hydroperoxides with substances such
as luminol and dichlorofluorescein (7, 35). This method was reviewd by Jimenez
et al. (36). High correlations have been found between chemiluminescence and
other standard methods, indicating that chemuliminescence could serve as an accu-
rate tool for determination of PV (37). However, this method has low sensitivity to
tert-butyl hydroperoxide, tert-butyl perbenzoate, diacyl peroxides, and dialkyl per-
oxides (35). Chromatographic techniques, mainly gas chromatography (GC) and
high-performance liquid chromatography (HPLC), have also been employed for
evaluation of lipid oxidation. These methods provide information about specific
hydroperoxides, whereas other assays measure their total amount. Chromatographic
methods require small amounts of sample, and interference from minor compounds
other than hydroperoxides can be easily excluded (8). HPLC shows advantages over
GC and has become a popular technique for hydroperoxide analysis. It operates at
room temperature, thus decreases the risk of artifact formation, and no prior deri-
vatization is required (8). A wide range of hydroperoxides can be analyzed using
either normal or reverse-phase HPLC. Thus, hydroperoxides, the primary products
and intermediates in lipid oxidation reaction, provide an important parameter for

evaluation of oxidation level. In addition, the inhibition of formation or action of
these unstable species by antioxidants can be used as a means of assessing antiox-
idant activity (9). Measurement of hydroperoxides is also carried out in accelerated
tests to establish the oxidative stability of a given oil. A case in point is the active
oxygen method (AOM), in which air is bubbled through fat or oil held at 98100 C
and PV is determined periodically (7, 38). The time required to reach a PV of 100
meq/kg is the AOM stability of the oil sample (7). This method is now considered
outdated and is replaced by other standard methods in the industry, although
product specifications still routinely give AOM values (38).
5.2. Conjugated Dienes and Trienes

It was discovered in 1933 that the formation of conjugated dienes in fats or oils
gives rise to an absorption peak at 230235 nm in the ultraviolet (UV) region. In
the 1960s, monitoring diene conjugation emerged as a useful technique for the
study of lipid oxidation (9). During the formation of hydroperoxides from unsatu-
rated fatty acids conjugated dienes are typically produced, due to the rearrangement
of the double bonds. The resulting conjugated dienes exhibit an intense absorption
at 234 nm; similarly conjugated trienes absorb at 268 nm (7). An increase in UV
absorption theoretically reflects the formation of primary oxidation products in fats
and oils. Good correlations between conjugated dienes and peroxide value have
been found (39, 40).
Ultraviolet detection of conjugated dienes is simple, fast, and requires no
chemical reagents and only small amounts of samples are needed. However, this
hydroperoxydiene oxodiene
O O
O
H
Reduction
hydroxydiene
O
H
conjugated triene
and
conjugated tetraene
Figure 1. Chemical reaction steps in conjugable oxidation products (COP) assay.

TABLE 1. Summary of Methods for Analysis of Primary Oxidation Products.
Method Principle Measurement Sensitivity Applications
Iodometric titration (PV) Reduction of ROOH with KI Titration with Na2S2O3 0.5-meq/kg fat Fats and oils
and measurement of I2
Ferric ion complexes (PV) Reduction of ROOH with Fe2 Absorption at 500510 nm 0.1-meq/kg fat Fats, oils and food lipids
and formation of Fe3 of the red complex with
complexes SCN
Absorption at 560 nm of the 0.5-meq/kg sample All samples
blue-purple complex with
xylenol orange
FTIR (PV) Reduction of ROOH with TPP Absorption at 542 cm1 of 0.2-meq/kg fat Fats and oils
TPPO
Chemiluminescence (PV) Reaction with luminol in the Chemiluminescence 1 pmol Fats and oils
presence of heme catalyst emission of oxidized
luminol
GC-MS (PV) Reduction of ROOH to ROH ROH derivatives From ng to fg depending on All samples
and quantitation of ROH technical details, amount
derivatives of sample and detection
system
UV spectrometry Estimation of conjugated Absorption at 230234 nm 0.2 meq/kg lipid All samples
(conjugated dienes and dienes and trienes and 268 nm
trienes)
NOTE: The oxygen absorption measurement and loss of double bonds for fatty acid analysis are not considered as primary changes in this table.
Adapted from (8).
method has less specificity and sensitivity than PV measurement (9, 12). Further-
more, the result may be affected by the presence of compounds absorbing in the
same region, such as carotenoids (7). To avoid these interferences, an alternative
spectroscopic method measuring conjugable oxidation products (COPs) has
been proposed. In this method, hydroperoxides and some decomposition products
are converted to more conjugated chromophores by reduction and subsequent
dehydration (Figure 1). The concentrations of the resultant conjugated trienes
and tetraenes are determined from their respective absorption at 268 nm and
301 nm and expressed as COP values (7, 12).
Table 1 summarizes different methods available for analysis of primary oxida-
tion products. Both chemical and instrumental methods are included in this
table.
6. MEASUREMENT OF SECONDARY PRODUCTS OF OXIDATION
The primary oxidation products (hydroperoxides) are unstable and susceptible to

decomposistion. A complex mixture of volatile, nonvolatile, and polymeric second-
ary oxidation products is formed through decomposition reactions, providing var-
ious indices of lipid oxidation (5). Secondary oxidation products include aldehydes,
ketones, alcohols, hydrocarbons, volatile organic acids, and epoxy compounds,
among others. Methods for assessing lipid oxidation based on their formation are
discussed in this section.
6.1. Thiobarbituric Acid (TBA) Test

The thiobarbituric acid (TBA) test was proposed over 40 years ago and is now
one of the most extensively used methods to detect oxidative deterioration of
fat-containing foods (41). During lipid oxidation, malonaldehyde (MA), a minor
component of fatty acids with 3 or more double bonds, is formed as a result of
the degradation of polyunsaturated fatty acids. It is usually used as an indicator
of the lipid oxidation process, both for the early appearance as oxidation occurs
and for the sensitivity of the analytical method (42). In this assay, the MA is reacted
with thiobarbituric acid (TBA) to form a pink MA-TBA complex that is measured
spectrophotometrically at its absorption maximum at 530535 nm (Figure 2) (9, 43, 44).
The extent of oxidation is reported as the TBA value and is expressed as milligrams
OH O OH
O O
N HN NH
+ H C CH2 C H
HS N OH S N OH O N S
H
TBA MA
TBA-MA adduct
Figure 2. Reaction of 2-thiobarbituric acid (TBA) and malonaldehyde (MA).
MEASUREMENT OF SECONDARY PRODUCTS OF OXIDATION 367
of MA equivalents per kilogram sample or as micromoles of MA equivalents per

gram of sample. It must, however, be noted that alkenals and alkadienals also react
with the TBA reagent and produce a pink color. Thus, the term thiobarbituric acid
reactive substances (TBARS) is now used instead of MA.
The TBA test can be performed by various procedures, among which four major
types have frequently been employed. These include test on the whole sample, test
on an aqueous or acid extract of sample, test on a steam distillate, and test on
extracted lipid from a sample (45). The test on a steam distillate (distillation meth-
od) is the most commonly used method for determining TBA value. Tarladgis et al.
(46) found that the distillation of an acidified sample was essential to liberate MA
from precursor or bound forms, to produce maximal color development and, espe-
cially, to separate TBARS from the food matrix (44). Although the distillation
method is the most popular TBA method, it is generally considered less accurate
and reproducible than the method using food extracts (15). However, trends
obtained in comparative studies always provide useful information that correspond
with other measurements. Comparison of different TBA test procedures has
been made by Hoyland et al. (46), Shahidi et al. (47), Pikul et al. (48), and
Wang et al. (49).
The TBA test is used frequently to assess the oxidative state of a variety of food
systems, despite its limitations, such as lack of specificity and sensitivity (44). As
already noted, many other substances may react with the TBA reagent and contri-
bute to absorption, causing an overestimation of the intensity of color complex (44).
Interferences may come from additional absorption of other alkanals, 2-alkenals,
2,4-alkdienals, ketones, ketosteroids, acids, esters, proteins, sucrose, urea, pyri-
dines, and pyrimidines, also referred to as TBARS (43, 50). For instance, the
reaction of TBA with various aldehydes leads to the development of a yellow
chromogen (aldehyde-TBA adduct) with an absorption maximum at 450 nm, which
overlaps with the pink peak at 532 nm resulting in erroneously high TBA values in
certain cases (43, 45, 51). Furthermore, the presence of barbituric acid impurities in
the TBA reagent may produce TBA-MA-barbituric acid and MA-barbituric acid
adducts that absorb at 513 nm and 490 nm, respectively, indicating that thiobarbi-
turic acid should be purified before use (43). In addition, nitrite can interfere in the
TBA test, whereas sulfanilamide could be added to samples to avoid the interfer-
ence when residual nitrite is present (52). In order to improve the specificity and
sensitivity of the TBA test, several modifications to the original TBA procedures
have been proposed, including reduction of the heating temperature to stabilize
the yellow color aldehyde-TBA complex (53), addition of antioxidants to sample
in an attempt to prevent oxidation during the test (54), extraction of the MA prior
to the formation of the chromogen (43), direct FTIR analysis of TBARS, and use of
HPLC to separate the complex before measurement or to characterize the individual
species of TBARS (9, 43).
Despite it limitations, the TBA test provides an excellent means for evaluating
lipid oxidation in foods, especially on a comparative basis. However, its use in bulk
oils is less common than the so-called para-anisidine value (p-AnV) detailed
below.
6.2. p -Anisidine Value (p -AnV)

The p-anisidine value (p-AnV) method measures the content of aldehydes (princi-
pally 2-alkenals and 2,4-alkadienals) generated during the decomposition of hydro-
peroxides. It is based on the color reaction of p-methoxyaniline (anisidine) and the
aldehydic compounds (55). The reaction of p-anisidine reagent with aldehydes
under acidic conditions affords yellowish products that absorb at 350 nm
(Figure 3) (7, 12). The color is quantified and converted to p-AnV. The p-AnV is
defined as the absorbance of a solution resulting from the reaction of 1 g of fat in
isooctane solution (100 ml) with p-anisidine (0.25% in glacial acetic acid) (12).
This test is more sensitive to unsaturated aldehydes than to saturated aldehydes
because the colored products from unsaturated aldehydes absorb more strongly at
this wavelength (12). However, it correlates well with the amount of total volatile
substances (55). The p-AnV is a reliable indicator of oxidative rancidity in fats and
oils and fatty foods (56). A highly significant correlation between p-AnV and flavor
scores and PV has been found (57). Nevertheless, some authors have indicated that
p-AnV is comparable only within the same oil type because initial AnV varies
among oil sources (58). For instance, oils with high levels of polyunsaturated fatty
acids might have higher AnV even when fresh (59).
This method is used less frequently in North America, but is widely employed in
Europe (38), particularly as a part of the Totox number, as explained below. Caution
O OH NH2
C C +
H C H CH3O
H
Malonaldehyde p-Methoxyaniline
(enolic form) (p-anisidine)
N OH
CH3O
NH2
CH3O
N NH
CH3O OCH3
Figure 3. Possible reactions between p-anisidine reagent and malonaldehyde.

must be exercised when performing this test because of toxicity of the anisidine
reagent (55).
6.3. Totox Value

The Totox value is a measure of the total oxidation, including primary and second-
ary oxidation products. It is a combination of PV and p-AnV:
Totox value 2 PV p-AnV
During lipid oxidation, it is often observed that PV first rises, then falls as hydro-
peroxides decompose (38). PV and p-AnV reflect the oxidation level at early and
later stages of oxidation reaction, respectively. Totox value measures both hydro-
peroxides and their beakdown products, and provides a better estimation of the
progressive oxidative deterioration of fats and oils (38). However, Totox value
has no scientific basis because it is a combination of two indicators with different
dimensions (7). Recently, Wanasundara and Shahidi used TBA values and defined
TotoxTBA as 2PV TBA using the TBA test in place of the p-AnV assay (60).
6.4. Carbonyls
The carbonyl compounds, including aldehydes and ketones, are the secondary oxi-
dation products generated from degradation of hydroperoxides, and are suggested
to be the major contributors to off-flavors associated with the rancidity of many
food products (9). The analysis of total carbonyl compounds, which is based on
the absorbance of the carbonyl derivatives, provides another approach to measure
the extent of lipid oxidation in fats and oils. In this method, the total carbonyl
content is measured by a colorimetric 2,4-dinitrophenylhydrazone procedure. The
carbonyl compounds formed during lipid oxidation are reacted with 2,4-dinitrophe-
nylhydrazine (DNPH) followed by the reaction of the resulting hydrazones with
alkali (Figure 4). The final colored products are then analyzed spectrophotometrically
R H R H
R C O + H2N N NO2 R C N N NO2
NO2 NO2
OH
H2O
R

R C N N NO2
NO2
Figure 4. Reactions between carbonyls and 2,4-dinitrophenylhydrazine.

at a given wavelength (7, 15). Many variations of this method using an alternative
solvent, reagent, wavelength, or workup have been reported. The determination of
total content of carbonyls has been used in different oxidative stability studies.
However, it has been criticized because the determination conditions cause degra-
dation of hydroperoxides into carbonyl derivatives, giving erroneous results (58).
Carbonyls produced from protein oxidation may also give rise to higher values
than those expected from lipid oxidation alone. The addition of triphenylphosphine
(TPP) prior to carbonyl determination has been proposed to avoid the interference
from hydroperoxides. Hydroperoxides are reduced by TPP, and neither TPP nor
TPPO, the oxidation products of TPP, interfere with the measurement of carbonyl
content (61). In quality assessment of used frying fats, where short-chain carbonyls
are already removed by distillation at the high temperature of the deep-frying,
selectivity can be improved by determination of higher carbonyl compounds instead
of the total carbonyls. HPLC is used to separate the DNPH derivatives of higher
carbonyls from those of short-chain carbonyl compounds (62).
Apart from detection of total carbonyl content, the analysis of individual carbo-
nyl compounds has gained popularity for following lipid oxidation. Hexanal, one of
the major secondary products formed during the oxidation of linoleic and other o6
fatty acids, serves as a reliable indicator of lipid oxidatin in foods rich in o6 fatty
acids (7). A strong linear relationship was reported between hexanal content, sen-
sory scores, and TBA values (63). Moreover, measurement of hexanal offers the
advantage of analyzing a single, well-defined end product for antioxidant efficiency
studies (9). Hexanal can be quantified by chromatography (64) or as the intensity of
the carbonyl band by NIR spectroscopy (65). Nevertheless, these methods may
require volatilization of hexanal, whereas hexanal volatilization may be hindered
due to covalent or other types of binding between hexanal and proteins in foods
and, thus, may affect accurate hexanal quantifications (66). More recently, an
indirect enzyme-linked immunosorbanct assay (ELISA) has been developed for
monitoring lipid oxidation through quantification of hexanal-protein adducts, which
are recognized by polyclonal or monoclonal antibodies (66).
Other carbonyl compounds, including propanal, pentanal, decadienal, etc., are
also used for evaluating lipid oxidation in foods. For instance, propanal is a recom-
mended indicator for lipid oxidation in foods that are high in o3 fatty acids, such as
marine oils (67, 68). In general, it is essential to use appropriate indicators when
assessing the oxidative deterioration of different food systems.
6.5. Oil Stability Index (OSI)

During lipid oxidation, volatile organic acids, mainly formic acid and acetic acid,
are produced as secondary volatile oxidation products at high temperatures, simul-
taneously with hydroperoxides (20, 69). In addition, other secondary products,
including alcohols and carbonyl compounds, can be further oxidized to carboxylic
acids (20). The oil stability index (OSI) method measures the formation of volatile
acids by monitoring the change in electrical conductivity when effluent from
oxidizing oils is passed through water (12). The OSI value is defined as the point
of maximal change of the rate of oxidation, attributed to the increase of conductivity
by the formation of volatile organic acids during lipid oxidation (70). However, this
method requires a somewhat higher level of oxidation (PV > 100) to obtain mea-
surable results than other methods in which hydroperoxides are the most important
products formed and detected (71). Therefore, to determine oil stability in the
laboratory, especially for some oils that are stable under normal conditions, the oxi-
dation process is accelerated by exposing oil samples to elevated temperatures in
the presence of an excess amount of air or oxygen (72, 73). The OSI method differs
from ambient storage conditions by using a flow of air and high temperatures
to accelerate oxidation (71). The OSI is an automated development of the active-
oxygen method (AOM), because both employ the principle of accelerated oxida-
tion. Nevertheless, the OSI test measures the changes in conductivity caused by
ionic volatile acids, whereas PV is determined in the AOM (7).
Two pieces of commercially available equipment, the Rancimat (Metrohm Ltd.)
and the Oxidative Stability Instrument (Omnion Inc.), are employed for determin-
ing the OSI value. Rancimat is a rapid automated method, which agrees well with
the AOM (71). In the Rancimat assay, a flow of air is bubbled through a heated oil,
usually at 100 C or above. For marine oils, temperatures as low as 80 C are often
used. Volatile compounds formed during accelerated oxidation are collected in dis-
tilled water, increasing the water conductivity. The change of conductivity is plotted
automatically and the induction period of the oil or the time taken to reach a fixed
level of conductivity is recorded (20, 74). The Rancimat assay enables continuous
monitoring of the oxidation process. As reported by Farooq et al. (75), analysis by
the Rancimat method is four to five times more rapid than that by the AOM. Excel-
lent correlation between Rancimat and conjugated dienes has been found (72).
However, the main shortcoming of this method is that only eight samples can be
included in each batch. Another appatatus, the Oxidative Stability Instrument, oper-
ates on the same principle as the Rancimat, and has the capacity of simultaneously
analyzing up to 24 samples (20). Various modifications have been proposed for
assessing lipid oxidation by the OSI method. These include the use of auxiliary
energies, such as microwaves to shorten the analysis time (72) and a combination
of the OSI method with chromatography to obtain specific information about vola-
tile products (76). The volatiles trapped during measurement by the Rancimat assay
can be analyzed by headspace-GC (HS-GC) with FID and GC-MS for quantifica-
tion of individual volatiles, thus improving the specificity of the assessment (76).
Although the OSI method is useful for quality control of oils, it is not recom-
mended for measurement of antioxidant activity for certain reasons. The high tem-
peratures used do not allow reliable predictions of antioxidant effectiveness at
lower temperatures. Volatile antioxidants may be swept out of the oil by the air
flow under test conditions, and also the oils are severely deteriorated when endpoint
is reached (12).
6.6. Hydrocarbons and Fluorescence Assay

Formation of saturated hydrocarbons, especially short-chain (C1-C5) hydrocarbons
such as ethane, propane, and pentane, can be measured for monitoring lipid oxida-
tion when aldehydes are either absent or undetectable (7, 15). Pentane content,
O RNH2
H C CH CHOH RN CH CH CHOH
Amine-Malonaldehyde Adduct
(Non-fluorescent)
RNH2
O O
H C CH2 C H RN CH CH CH NHR
Malonaldehyde Conjugated Schiff Base
(Fluorescent)
Figure 5. Reaction of lipid oxidation products such as malonaldehyde and amines.
determined by GC techniques, has been a useful parameter to assess rancidity of

fats and oils as well as freeze-dried muscle foods (7, 15). Significant correlations
existed between pentane levels and rancid odor scores (15).
It has been observed that the content of secondary oxidation products, such as
malonaldehyde (MA), decreases with increased lipid oxidation, which can be
explained by further reaction of MA with proteins. MA reacts with compounds con-
taining primary amino groups (proteins, amino acids, DNA, phospholipids) to form
fluorescent products (Figure 5) (37). A fluorescence assay has been successfully
used to assess lipid oxidation in muscle foods and biological tissues.
In addition to MA, hydroperoxides and other aldehydes also react with amino
compounds generating various fluorescent products with different excitation and
emission maxima (37). Significant correlations existed between this method and
the TBA value as well as oxygen absorption level, and appears to be a reliable
TABLE 2. Summary of Methods for Analysis of Secondary Oxidation Products.
Method Compounds Comments Applications
TBA TBARS, mainly Spectrometry technique All samples,

malonaldehyde It can be carried out especially fish
on whole sample oils
p-Anisidine Aldehydes, mainly Absorption at 350-nm Fats and oils
alkenals Standard method
Carbonyls Total carbonyls or Spectrometry technique Fats and oils
specific carbonyl and HPLC for total
compound formed or specific carbonyl
compounds
OSI method Volatile organic acids Monitoring changes in Fats and oils
(Rancimat & conductivity Rapid
Oxidative Stability and automated
Instrument)
Gas Chromatography Volatile carbonyls and Direct headspace Rapid All samples
hydrocarbons analysis
Adapted from (8).

MEASUREMENT OF FREE RADICALS 373
indicator of oxidative deterioration in muscle foods, especially in freeze-dried

products (37, 77).
Table 2 exhibits the summary of methods for analysis of secondary oxidation
products.
7. MEASUREMENT OF FREE RADICALS
The initial steps of lipid oxidation involve chain reactions of free radicals as impor-
tant short-lived intermediates. Oxidation level of fats and oils can be measured
directly by detecting the formation of radicals. Methods based on the detection
of radicals or on the tendency for the formation of radicals provide a good indica-
tion of initiation of lipid oxidation (78, 79).
Electron spin resonance (ESR), also referred to as electron paramagnetic reso-
nance (EPR) spectroscopy, relies on the paramagnetic properties of the unpaired
electrons in radicals and has been developed for assessing the formation of free
radicals originating in the early stages of oxidation and the onset of primary oxida-
tion (6, 78). The assay measures the absorption of microwave energy when a
sample is placed in a varied magnetic field (7). Quantification of radical concentra-
tions is complicated by comparison with stable paramagnetic compounds, such as
transition metals and nitroxyl radicals (78). However, the short lifetimes and low
steady-state concentration of the highly reactive lipid-derived radicals make it dif-
ficult to detect these radicals at concentrations lower than the minimum detectable
concentration of 109 M (78). To overcome this problem, various approaches have
been used, including pulse radiolysis and UV photolysis, continuous flow systems
and spin trapping, among which spin trapping has been the most widely employed
procedure (9). Spin trapping technique allows the accumulation of detectable con-
centrations of longer-lived radicals by addition to samples of a spin trapping agent,
which reacts with free radicals to form more stable spin adducts, but often at the
expense of the ability to identify the original radical (6, 9, 78). Nitroso compounds
and nitrones are the most common spin traps, both leading to nitroxyl type spin
adducts, such as a-phenyl-tert-butylnitrone (PBN) adducts (Figure 6) (78).
O
H H O
N+ + R Ph N
Ph CMe3 R CMe3
PBN
O
Me3C N O + R R N
CMe3
MNP
Figure 6. Formation of nitroxyl radical spin adducts.
ESR spectroscopy is of great value for the study of the early stages of lipid oxi-
dation and prediction of oxidative stability of fats and oils. It has high sensitivity
and allows mild conditions by applying significantly low temperatures and requires
little sample preparation (6, 78, 80). Strong linear correlations were found between
ESR and Rancimat and oxygen consumption analyses (6, 79). ESR has also been
used for evaluation of antioxidant activity (81). Nevertheless, spin traps used in the
ESR assay have been reported to exhibit widely differing trapping efficiencies for
different radicals and show both pro-oxidant and antioxidant effects (9, 82, 83).
Moreover, spin adducts can act as antioxidants, giving erroneous results of oxida-
tive stability of samples (9). However, even with these limitations, the ESR spectro-
scopy is a suitable method for measuring lipid oxidation in foods and in biological
tissues.
8. OTHER METHODS
8.1. Differential Scanning Calorimetry (DSC)

During lipid oxidation, fat or oil materials reveal a number of thermally induced
transitions, such as the transfer of oxygen molecules to unsaturated fatty acids
(exothermic process) (84). Therefore, thermal analysis can be applied in accelerated
oil stability tests. The differential scanning calorimetry (DSC) technique, which is
based on thermal release of oxidation reactions, has the potential as a nonchemical
method for assessing oxidative stability of fats and oils, indicating the onset of
advanced oxidation (termination) (6). It provides unique energy profile information,
which specifically measures the temperature and heat flows associated with lipid
oxidation as a function of time and temperature (85). The method uses isothermal
or nonisothermal conditions and a flow of oxygen as purge gas, with a calorimeter
measuring the heat flow into (endothermic) or out of (exothermic) an oil sample
undergoing oxidation changes (6, 84). The oxidation curves of the sample are
obtained with different heating time, and a dramatic increase for the evolved
heat can be observed with the appearance of a sharp exothermic curve during initia-
tion of oxidation. The endpoint is taken at the time where a rapid exothermic reac-
tion between oil and oxygen occurs and induction period (IP) determined
automatically by intersection of extrapolated baseline and tangent line (leading
edge) of the exotherm (Figure 7) (6, 84). The DSC also measures oxidation onset
temperature, the temperature at maximum reaction, and the ending temperature
(84). The isothermal and nonisothermal DSC show good agreement, suggesting
that both isothermal and nonisothermal DSC are suitable for oxidation studies of
oils (86). The DSC technique has recently been reviewed by Tan et al. (84, 87).
The DSC is a sensitive, effective, and consistent method for characterization of
the quality of oils at different stages of oxidation (20). It is simple and rapid, and it
requires no solvent or chemical reagent. As reported by Hassel et al. (89), oils sam-
ples, which required 14 days via AOM, could be evaluated in less than 4 hours
by DSC. Thus, DSC is a reliable alternative to current methods for monitoring lipid
OTHER METHODS 375
Exothermic
IP
50 100 150 200 250

Time (min)
Figure 7. Determination of induction period (IP) by DSC.
oxidation (85). The results from DSC show excellent correlations with other
accelerated methods and chemical analyses (6, 73, 85).
8.2. Nuclear Magnetic Resonance (NMR) Spectroscopy

High-resolution 1H NMR spectroscopy, in which hydrogen atoms (proton, 1H) with
various locations in the triacylglycerol (TAG) molecules are determined, has been
used to evaluate oxidative deterioration of fats and oils (7). The principle of NMR is
that hydrogen atoms in a strong magnetic field absorb energy, in the radiofrequency
range, depending on their molecular environment, in which changes occur during
the oxidation process (7). These changes may be monitored by NMR spectroscopy
as a reflection of oxidation level of food lipids. The oil sample is dissolved in
CDCl3 to avoid inference from solvent, and its NMR spectrum recorded, with tetra-
methylsilane (TMS) as an internal standard (7). The spectrum shows several groups
of signals, corresponding to the hydrogen atoms in different locations in the TAG
molecules (Figure 8). The total number of each of these differently located protons
can be calculated, from which ratios of aliphatic to olefinic protons (Rao) and ali-
phatic to diallylmethylene protons (Rad) may be obtained (7). Both ratios increase
steadily during lipid oxidation and may serve as an index of oxidative deterioration
of oil samples. This method was reviewed by Guillen et al. (90). NMR spectroscopy
has been used by many researchers, and the changes in Rao and Rad measured by
NMR correlated well with Totox values (91, 92), conjugated diene values, and TBA
values (93). In addition to 1H NMR, 13C NMR and 31P NMR are also powerful tools
to predict oxidative stability of oils (9496). 13C NMR enables direct observation
of carbon atoms. The selectivity and dispersion of 13C NMR spectra are very
high (96). 13C NMR assesses lipid oxidation by monitoring the changes of carbon
chains in TAG molecules, revealing the specific sites that oxidative degradation
b
H d g a a c a a e f g h
b H C O CO CH2 (CH2)n CH CH CH2 CH CH CH2 CH2 (CH2)n CH3
a H C O CO CH2 (CH2)n CH3
b H C O CO CH2 (CH2)n CH3
H
b
TMS
h
a d
e
b
f
c
8 7 6 5 4 3 2 1 0 PPM
1
Figure 8. H NMR spectrum of oxidized canola oil.
occurs (94). However, because the abundance of the NMR active 13C nucleus iso-
tope is only 1.12% of 12C, the sensitivity of 13C NMR is usually much lower than
that of 1H NMR (96).
NMR spectroscopy is a rapid, nondestructive, and reliable technique for asses-
sing lipid oxidation. It simultaneously measures both the primary and the secondary
oxidative changes in oils, and provides specific information on oxidative regions in
the TAG molecules. Thus, NMR spectroscopy is considered a more suitable means
for estimating lipid oxidation than chemical determinations.
8.3. Sensory Evaluation

For the food industry, the detection of oxidative off-flavors by taste or smell is the
main method of deciding when a lipid-containing food is no longer fit for consump-
tion (12). Terminologies and methodologies have been developed for sensory
evaluation of specific food products such as meats, peanuts, and vegetable oils
(97). In the edible oil industry, the AOCS (American Oil Chemists Society) Flavor
Quality Scale (revised) with separate grading and flavor intensity has been
employed for describing lipid oxidation (97), as summarized in Table 3. The
descriptive analysis, including the detection and the description of both the quali-
tative and quantitative sensory aspects of a product, is performed by a trained panel,
as the sensitivity to the off-flavors varies among different individuals (12, 97). The
sensory induction period of the product can be determined.
MEASUREMENT OF FRYING FAT DETERIORATION 377
TABLE 3. A Partial List of Terms Used to Describe

Oxidized Oil.
Flavor-Related Terms Process-Oriented Terms
Buttery Hydrogenated
Nutty Oxidized
Beany Reverted
Grassy Light-struck
Watermelon Rancid
Painty
Fishy
Adapted from (97).
Sensory evaluation of lipid oxidation has been conducted by many researchers

(98100). However, as a subjective method, the reproducibility of sensory analysis
is generally considered worse than that of chemical or instrumental methods. More
recently, use of an electronic nose to monitor the formation of volatile compounds
associated with off-flavors from lipid oxidation has been proposed to supplement
information from human sensory panels (101).
9. MEASUREMENT OF FRYING FAT DETERIORATION
Deep-fat frying is a popular method for food preparation, in which vegetable oils
not only are used as a heat-exchange medium, but also contribute to the quality of
fried products (7). However, lipid oxidation easily occurs at relatively high tem-
peratures, producing a complex series of compounds that exerts undesirable effects
on food flavor and quality (4). The measurement of lipid oxidation, therefore, is
essential to determine its effect on food and oil quality, as well as the useful life
of fats or oils subjected to frying. The oxidative changes in frying fats are charac-
terized by a decrease in the total unsaturation of the fat with increases in the free
fatty acid content, foaming, color, and viscosity as well as the content of polar com-
pounds and polymeric material (4). Quality evaluation of frying fats, may be carried
out in different ways. Physical methods estimate oxidative degradation by monitor-
ing changes in physical properties of frying fats, such as molecular weight, specific
gravity, smoke point, refractive index, chromatic parameter, viscosity, surface
tension, and dielectric constant (4). Generally, rejection point of frying fat is estab-
lished by sensory assessment. Chemical methods include the iodine value, saponi-
fication value, free fatty acid content, peroxide value, TBA value, or p-anisidine
value, among others. PV is less useful because hydroperoxides decompose at about
150 C, and no accumulation of peroxides can be detected.
The extent of oxidation can also be assessed by the analysis of oxidized fatty
acids by spectroscopic means such as IR and NMR techniques (102). Moreover,
GC-MS for volatile profile analysis (103) and HPLC for determination of DNPH
derivatives of nonvolatile higher carbonyl compounds (62) provide qualitative
9 7 5 3
COOH
CH3
Double bond may present on the positions of C4, C5, C7, C8, or C9.
(CH2)nH (CH2)nH (CH2)nH
R R R
R = (CH2)11nCOOH 1 < n < 11
(CH2)nH
(CH2)nH (CH2)12nCOOH
(CH2)8nCOOH
n=1&2 n=3&4
Figure 9. Chemical structures of cyclic fatty acids formed during deep frying.
and quantitative evaluation of oxidation in frying fats. Cyclic fatty acids (Figure 9),
which may contain hydroxy and keto groups, are formed during deep frying and can
be measured by chromatography after derivatization (4, 7). Furthermore, determi-
nation of polar material in frying fats is a reliable approach for oil quality evalua-
tion and is an official method in Europe. This method involves separation of fat into
a polar and nonpolar fraction via silica gel chromatography. Nonpolar fat can be
weighed and the total polar material calculated or determined directly by their elu-
tion from the silica gel column (4,7).
Routine analysis for frying fat deterioration has been reviewed by Gertz (104).
Usually, more than two methods are required when using chemical analysis because
no single group of compounds has been identified as a key indicator of oxidative
degradation of frying fats.
10. METHODS FOR MEASURING ANTIOXIDANT ACTIVITY
A variety of natural and synthetic antioxidants are used in fat-containing foods in

order to inhibit lipid oxidation with a wide range of efficiencies, depending on their
properties, concentrations, and processing conditions. The need to measure antiox-
idant activity is well documented. Although numerous methods have been proposed
for measurement of antioxidant activity, the essential features of any test are a sui-
table substrate, an oxidation initiator, and an appropriate measure of endpoint (9).
Therefore, certain aspects should be taken into consideration when selecting a test
for measuring antioxidant activity. These include the model food system used for
METHODS FOR MEASURING ANTIOXIDANT ACTIVITY 379
the test, and the means by which oxidation is accelerated and monitored (12). Nor-
mally, most assessments of antioxidant activity are performed in oil, or other model
systems, giving sensible prediction for the activity in oil or water-in-oil emulsions,
whereas the results may be misleading for oil-in-water emulsions (12). Further-
more, stripping of oils may be necessary in such evaluations because the endogen-
ous antioxidants in nonstripped oils are found to enhance the oxidative stability of
oils, thus giving rise to erroneous results in the efficiency of antioxidants under
investigation (105107). In addition to oils and fats, lipid substrates used for testing
antioxidant activity could be fatty acids, fatty acid ethyl esters or triacylglycerols
(9), and b-carotene (108110). In some cases, such as radical scavenging methods,
no substrate is used. Most test procedures involve initiators to accelerate oxidation.
The combination of increased temperature and oxygen supply, addition of metal
catalysts, and exposure of the reactants to light can reduce the oxidative stability
by a large amount (9, 12). Nevertheless, the elevated temperature may bring about
changes in the oxidation mechanism, thus causing difficulties in the prediction of
TABLE 4. Methods of Expressing Results of Antioxidant Activity Tests.
Method Dimensions
Induction period h, d
Time to reach a set level of oxidation (pre- h, d
induction period)
Rate of oxidation (pre-induction period) mol kg1 hr1, gL1 d1
Concentration to produce equivalent effect to mol kg1, gL1
reference antioxidant (pre-induction period)
Concentration of ROOH functional group after mequiv. kg1
set time period
Concentration of oxidation product after set mg kg1 (ppm w/w)
time period
Scale reading after set time period Absorbance, conductivity, etc.
Free stable radical quenching (DPPH) Percentage inhibition
EC50, concentration to decrease concentration
of test free radical by 50%
TEC50, time to decrease concentration of test
free radical by 50%
Total radical-trapping antioxidant parameter mmol peroxy radical deactivated L1
(TRAP)
ABTS assay, phycoerythrin assay TEAC (mM Trolox equivalent to 1-mM test
substance)
Phycoerythrin assay ORAC, oxygen radical absorbance capacity;
mmol of Trolox equivalents
FRAP assay Absorbance of Fe2 complex at 593 nm
produced by antioxidant reduction of
corresponding tripyridyltriazine Fe3
complex
Metal chelating assay Percentage of inhibition of ferrozine-Fe2
complex formation
NOTE: Also see Tables 1 and 2 for other tests applicable to antioxidant activity determination.
Adapted from (9).
antioxidant activity at low temperatures as compared with those at high tempera-

tures (9, 12). After the substrate is oxidized under standard conditions, the oxidation
is monitored by chemical, instrumental, or sensory methods. An appropriate mea-
sure of endpoint is essential for assessing antioxidant activity. Analytical strategies
for endpoint determination include measurement at a fixed time point, measurement
of reaction rate, lag phase measurement, and integrated rate measurement (9). The
resulting antioxidant activity is expressed using a wide range of parameters (Table 4).
Approaches proposed for testing antioxidant activity include measuring of the
current state of oil samples, as discussed above, and radical scavenging assays, which
are gaining popularity in the evaluation of antioxidant activity. Radical scavenging
methods measure the relative abilities of antioxidants to scavenge synthetic radicals
or natural in comparison with the antioxidant potency of a standard antioxidant
compound (111). Trolox (6-hydroxy-2,5,7,8-tetramethylchroma-2-carboxylic acid),
ascorbic acid, and quercetin are among the standard antioxidants frequently
used. The most commonly used synthetic radicals are DPPH (2,2-diphenyl-1-
picrylhydrazyl) and ABTS (3-ethylbenzthiazoline-sulfonic acid) radicals. DPPH test
(112116) and ABTS assay (117122) are simple, rapid, and involve no substrate.
However, it has been suggested that these artificial substrate-free methods do not
always adequately mimic the processes in food systems, which sometimes makes
them less valuable for predicting the effectiveness of the antioxidant in foods (9).
Other measurements of antioxidant activity include FRAP (ferric reducing-
antioxidant power) (123126), TRAP (total radical-trapping antioxidant parameter)
(123, 127), phycoerythrin assay (128, 129), and test of metal chelating capacity
(130, 131), among others. Reviews on methods for testing antioxidant activity
have been published (9, 12).
11. CONCLUSIONS AND RECOMMENDATIONS
Lipid oxidation may be assessed in many ways, among which changes in the initial
reactants and formation of oxidation products are most commonly assessed. Mean-
while, sensory analysis assesses both the subjective and, in some cases, objective
measurements of oxidative changes in foods. Each method shows both advantages
and disadvantages, thus it is important to select the most adequate method, depend-
ing on the system under investigation and the state of oxidation itself. The use of
two or more methods assessing both primary and secondary oxidation products is
highly recommended.
REFERENCES
1. F. Shahidi, Nahrung., 44, 158163 (2000).

2. J. R. Vercellotti, A. J. St. Angelo, and A. M. Spanier, in A. J. St. Angelo, ed., Lipid
Oxidation in Food, ACS Symposium Series 500, American Chemical Society,
Washington, D.C., 1992, pp. 111.
REFERENCES 381
3. M. Gordon, in J. Pokorny, N. Yanishlieva, and M. Gordon, eds., Antioxidants in Food:

Practical Applications, Woodhead Publishing, Ltd., Cambridge, England, 2001,
pp. 721.
4. E. G. Perkins, in A. J. St. Angelo, ed., Lipid Oxidation in Food, American Chemical
5. A. Kamal-Eldin, M. Makinen, and A. M. Lampi, in A. Kamal-Eldin, ed., Lipid Oxidation
Pathways, AOCS Press, Champaign, Illinois, 2003, pp. 136.
6. J. Velasco, M. L. Anderson, and L. H. Skibsted, Food Chem., 85, 623632 (2004).
7. F. Shahidi and U. N. Wanasundara, in C. C. Akoh and D. B. Min, eds., Food Lipids:
Chemistry, Nutrition and Biotechnology, Marcel Dekker, Inc., New York, 2002,
pp. 465487.
8. M. C. Dobarganes and J. Velasco, Eur. J. Lipid Sci. Technol., 104, 420428 (2002).
9. M. Antolovich, P. D. Prenzler, E. Patsalides, S. Mcdonald, and K. Robards, Analyst, 127,
183198 (2002).
10. I. Niklova, S. Schmidt, K. Habalova, and S. Sekretar, Eur. J. Lipid Sci. Technol., 103,
299306 (2001).
11. J. Velasco and C. Dobarganes, Eur. J. Lipid Sci. Technol., 104, 661676 (2002).
12. M. Gordon, in J. Pokorny, N. Yanishlieva, and M. Gordon, eds., Antioxidants in
Food: Practical Applications, Woodhead Publishing, Ltd., Cambridge, England,
2001, pp. 7184.
13. C. Genot, G. Kansci, and M. Laroche, Sciences des Aliments., 14, 673682 (1994).
14. H. J. Chung, A. S. Colakoglu, and D. B. Min, J. Food Sci., 69, 8388 (2004).
15. S. L. Melton, Food Technol., 37, 105111 (1983).
16. G. O. Fruhwirth, T. Wenzl, R. El-Toukhy, F. S. Wagner, and A. Hermetter, Eur. J. Lipid
Sci. Technol., 105, 266274 (2003).
17. B. J. F. Hudson, in J. C. Allen and J. Hamilton, eds., Rancidity of Foods, Applied Science
Publishers, London, England, 1983, pp. 4758.
18. A. Riuz, M. J. Ayora-Canada, and B. Lendl, Analyst, 126, 242246 (2001).
19. G. Yildiz, R. L. Wehling, and S. L. Cuppett, J. Amer. Oil Chem. Soc., 80, 103107 (2003).
20. A. Kiritsakis, A. Kanavouras, and K. Kiritsakis, Eur. J. Lipid Sci. Technol., 104, 628638
(2002).
21. Determination of Peroxide Value, in IUPAC: Standard Methods for the Analysis of Oils,
Fats and Derivatives, IUPAC Standard Method 2.501, International Union of Pure and
Applied Chemistry, Pergamon, Oxford, England, 1992.
22. D. L. Berner, INFORM, 1, 884886 (1996).
23. S. Eymard and C. Genot, Eur. J. Lipid. Sci. Technol., 105, 497501 (2003).
24. T. Asakawa and S. Matsushita, J. Amer. Oil. Chem. Soc., 55, 691620 (1978).
25. S. Hara and Y. Totani, J. Amer. Oil Chem. Soc., 65, 19481950 (1988).
26. M. Hicks and J. M. Gebicki, Anal. Biochem., 99, 249253 (1979).
27. J. M. Gebicki and J. Guille, Anal. Bichem., 176, 360364 (1989).
28. Z. Y. Jiang, A. C. S. Woollard, and S. P. Wolff, Lipids, 26, 853856 (1991).
29. S. P. Wolff, Methods Enzymol., 233, 182189 (1994).
30. J. M. DeLong, R. K. Prange, D. M. Hodges, C. F. Forney, M. C. Bishop, and M. Quillian,
J. Agric. Food Chem., 50, 248254 (2002).
31. J. Sedman, F. R. van de Voort, and A. A. Ismail, in R. E. McDonald and M. M. Mossoba,

eds., New Techniques and Applications in Lipid Analysis, AOCS Press, Champaign,
Illinois, 1997, pp. 283324.
32. J. Dong, K. Ma, F. R. van de Voort, and A. A. Ismail, J. AOAC Int., 80, 345352 (1997).
33. T. A. Russin, F. R. van de Voort, and J. Sedman, J. Amer. Oil Chem. Soc., 80, 635641
(2003).
34. B. Innawong, P. Mallikarjunan, J. Irudayaraj, and J. E. Marcy, Lebensm. Wiss. U.
Technol., 37, 2328 (2004).
35. T. Miyazawa, K. Fujimoto, and T. Kandeda, Agric. Biol. Chem., 51, 25692573 (1987).
36. A. M. Jimenez and M. J. Navas, Grasas y Aceites, 53, 6475 (2002).
37. J. P. Wold and M. Mielnik, J. Food Sci., 65, 8795 (2000).
38. C. Stauffer, in Fats & Oils, Eagan Press, St. Paul, Minnesota, 1996, pp. 1527.
39. F. Shahidi, U. Wanasundara, and N. Brunet, Food Res. Int., 27, 555562 (1994).
40. U. N. Wanasundara, F. Shahidi, and C. R. Jablonski, Food Chem., 52, 249253 (1995).
41. E. Kishida, S. Tokumaru, Y. Ishitani, M. Yamamoto, M. Oribe, H. Iguchi, and S. Kojo,
J. Agric. Food Chem., 41, 15981600 (1993).
42. S. Cesa, J. Agric. Food Chem., 52, 21192122 (2004).
43. D. Jardine, M. Antolovich, P. D. Prenzler, and K. Robards, J. Agric. Food Chem., 50,
17201724 (2002).
44. A. de las Heras, A. Schoch, M. Gibis, and A. Fischer, Eur. Food Res. Technol., 217, 180
184 (2003).
45. D. V. Hoyland and A. J. Taylor, Food Chem., 40, 271291 (1991).
46. B. G. Tarladgis, B. M. Watts, M. T. Younathan, and L. Dugan, J. Amer. Oil Chem. Soc.,
37, 4448 (1960).
47. F. Shahidi and C. Hong, J. Food Biochem., 15, 97105 (1991).
48. J. Pikul, D. E. Leszczynski, and F. A. Kummerow, J. Agric. Food Chem., 37, 13091313
(1989).
49. C. Wang, L. Zhu, and M. S. Brewer, in 1996 IFT annual meeting: Book of Abstracts,
pp. 161.
50. R. Guillen-Sans and M. Guzman-Chozas, Crit. Rev. Food Sci. Nutr., 38, 315330
(1998).
51. Q. Sun, C. Faustman, A. Senecal, A. L. Wilkinson, and H. Furr, Meat Sci., 57, 5560
(2001).
52. F. Shahidi, L. J. Rubin, L. L. Diosady, and D. F. Wood, J. Food Sci., 55, 274275 (1985).
53. R. Marcuse and J. Pokorny, Fett. Wiss. Technol., 96, 185187 (1994).
54. H. de A. Gomes, E. N. da Silva, M. R. Lopes do Nascimento, and H. T. Fukuma, Food
Chem., 80, 433437 (2003).
55. F. Doleschall, Z. Kemeny, K. Recseg, and K. Kovari, Eur. J. Lipid Sci. Technol., 104, 14
18 (2002).
56. G. H. van der Merwe, L. M. du Plessis, and J. RN. Taylor, J. Sci. Food Agric., 84, 5258
(2003).
57. G. R. List, C. D. Evans, W. F. Kwolek, K. Warner, and B. K. Boundy, J. Amer. Oil Chem.
Soc., 51, 1721 (1974).
58. M. D. Guillen and N. Cabo, Food Chem., 77, 503510 (2002).
REFERENCES 383
59. P. J. White, in K. Warner and N. A. M. Eskin, eds., Methods to Assess Quality and
Stability of Oils and Fat-containing Foods, AOCS Press, Chamaign, Illinois, 1995,
pp. 159178.
60. U. N. Wanasundara and F. Shahidi, J. Food Lipids, 2, 7386 (1995).
61. T. Chiba, M. Takazawa, and K. Fujinoto, J. Amer. Oil Chem. Soc., 66, 15881592 (1989).
62. E. Schulte, Anal. Bioanal. Chem., 372, 644648 (2002).
63. F. Shahidi and R. B. Pegg, J. Food Lipids, 1, 177186 (1994).
64. C. F. Goodridge, R. M. Beaudry, J. J. Pestka, and D. M. Smith, J. Agric. Food Chem., 51,
41854190 (2003).
65. H. Z. Zhang and T. C. Lee, IFT Annual Meeting, 1995, pp. 161.
66. C. F. Goodridge, R. M. Beaudry, J. J. Pestka, and D. M. Smith, J. Agric. Food Chem., 51,
75337539 (2003).
67. U. N. Wanasundara and F. Shahidi, Paper presented at the Canadian Section of American
Oil Chemists Societys Annual Meeting, Guelph, Ontario, Canada, November 1516,
1995.
68. F. Shahidi and S. A. Spurvey, J. Food Lipids, 3, 1325 (1996).
69. K. Schwarz, G. Bertelsen, L. R. Nissen, P. T. Gardner, M. I. Heinonen, A. Hopia,
T. Huynh-Ba, P. Lambelet, D. McPhail, L. H. Skibsted, and L. Tijburg, Eur. Food Res.
Technol., 212, 319328 (2001).
70. K. Miura, H. Kikuzaki, and N. Nakatani, J. Agric. Food Chem., 50, 18451851 (2002).
71. S. de la Presa-Owens, M. C. Lopez-Sabater, and M. Rivero-Urgell, J. Agric. Food Chem.,
43, 28792882 (1995).
72. M. P. Canizares-Macias, J. A. Garcia-Mesa, and M. D. Luque de Castro, Anal. Bioanal.
Chem., 378, 479483 (2004).
73. C. P. Tan, Y. B. Che-Man, J. Selamat, and M. S. A. Yusoff, Food Chem., 76, 385389
(2002).
74. R. Aparicio, L. Roda, M. A. Albi, and F. Gutierrez, J. Agric. Food Chem., 47, 41504155
(1999).
75. A. Faroop, M. I. Bhanger, and T. G. Kazi, J. Amer. Oil Chem. Soc., 80, 151155 (2003).
76. L. C. Boyd, V. C. Nwosu, C. L. Young, and L. MacMillian, J. Food Lipids, 5, 269282
(1998).
77. A. L. Wilkinson, Q. Sun, A. Senecal, and C. Faustman, J. Food Sci., 66, 2024 (2001).
78. M. L. Andersen and L. H. Skibsted, Eur. J. Lipid Sci. Technol., 104, 6568 (2002).
79. C. U. Carlsen, M. L. Andersen, and L. H. Skibsted, Eur. Food Res. Technol., 213, 170
173 (2001).
80. M. K. Thomsen, D. Kristensen, and L. H. Skibsted, J. Amer. Oil Chem. Soc., 77, 725730
(2000).
81. L. R. Nissen, Tuong-Huynh-Ba, M. A. Petersen, G. Bertelsen, and L. H. Skibsted, Food
Chem., 79, 387394 (2002).
82. B. Halliwell and J. M. C. Gutteridge, in Free Radicals in Biology and Medicine, Oxford
University Press, New York, 1999.
83. D. M. Lee, Methods Enzymol., 234, 513523 (1994).
84. C. P. Tan and Y. B. Che Man, Trends Food Sci. Technol., 13, 312318 (2002).
85. C. P. Tan and Y. B. Che Man, Food Chem., 67, 177184 (1999).
86. G. Litwinienko, A. Daniluk, and T. Kasprzycka-Guttman, J. Amer. Oil Chem. Soc., 76,
655657 (1999).
87. C. P. Tan and Y. B. Che Man, Lipid Technol., 14, 112115 (2002).
88. B. Kowalski, K. Ratusz, D. Kowalska, and W. Bekas, Eur. J. Lipid Sci. Technol., 106,
165169 (2004).
89. R. L. Hassel, J. Amer. Oil Chem. Soc., 53, 179181 (1976).
90. M. D. Guillen and A. Ruiz, Trends Food Sci. Technol., 12, 328338 (2001).
91. F. Shahidi, U. Wanasundara, and N. Brunet, Food Res. Int., 27, 555562 (1994).
92. U. Wanasundara and F. Shahidi, J. Food Lipids, 1, 1524 (1993).
93. S. P. J. Namal-Senanayake and F. Shahidi, J. Food Lipids, 6, 195203 (1999).
94. I. Medina, R. Sacchi, I. Giudicianni, and S. Aubourg, J. Amer. Oil Chem. Soc., 75, 147
154 (1998).
95. F. J. Hidalgo, G. Gomez, J. L. Navarro, and R. Zamora, J. Agric. Food Chem., 50, 5825
5831 (2002).
96. B. W. L. Diehl, in R. J. Hamilton, ed., Lipid Analysis in Oils and Fats, Blackie Academic
& Professional, London, 1998, pp. 87135.
97. G. V. Cinille and C. A. Dus, in A. J. St. Angelo ed., Lipid Oxidation in Food, American
Chemical Society, Washington, D.C., 1992, pp. 279289.
98. C. J. Broadbent and O. A. Pike, J. Amer. Oil Chem. Soc., 80, 5963 (2003).
99. E. A. Coppin and O. A. Pike, J. Amer. Oil Chem. Soc., 78, 1318 (2001).
100. M. Timm-Heinrich, X. Xu, N. S. Nielsen, and C. Jacobsen, Eur. J. Lipid Sci. Technol.,
105, 436448 (2003).
101. N. Shen, S. Moizuddin, L. Wilson, S. Duvick, P. White, and L. Pollak, J. Amer. Oil Chem.
Soc., 78, 937940 (2001).
102. S. Khatoon and A. G. G. Krishna, J. Food Lipids, 5, 247267 (1998).
103. V. Rosales and J. de D, Dissertation Abstracts International, B., 57, 7290 (1997).
104. C. Gertz, Lipid Technol., 13, 4447 (2001).
105. T. Keceli and M. H. Gordon, J. Sci. Food Agric., 81, 13911396 (2001).
106. M. A. Khan and F. Shahidi, J. Amer. Oil Chem. Soc., 77, 963968 (2000).
107. V. Povilaityte, M. E. Cuvelier, and C. Berset, J. Food Lipids, 8, 4564 (2001).
108. F. Shahidi, C. Desilva, and R. Amarowicz, J. Amer. Oil Chem. Soc., 80, 443450
(2003).
109. F. Shahidi and R. Amarowicz, J. Amer. Oil Chem. Soc., 73, 11971199 (1996).
110. R. Amarowicz, M. Naczk, and F. Shahidi, J. Amer. Oil Chem. Soc., 77, 957961 (2000).
111. O. A. Zaporozhets, O. A. Krushynska, N. A. Lipkovska, and V. N. Barvinchenko,
J. Agric. Food Chem., 52, 2125 (2004).
112. D. Bandoniene, M. Murkovic, W. Pfannhauser, P. R. Venskutonis, and D. Gruzdiene, Eur.
Food Res. Technol., 214, 143147 (2002).
113. N. Nenadis and M. Tsimidou, J. Amer. Oil Chem. Soc., 79, 11911195 (2002).
114. H. Matsuura, H. Chiji, C. Asakawa, M. Amano, T. Yoshihara, and J. Mizutani, Biosci.
Biotechnol. Biochem., 67, 23112316 (2003).
115. P. J. Park, F. Shahidi, and Y. J. Jeon, J. Food Lipids, 11, 1528 (2004).
116. R. Amarowicz, B. Raab, and F. Shahidi, J. Food Lipids, 10, 5162 (2003).
REFERENCES 385
117. A. Podsedek, D. Sosnowska, and B. Anders, Eur. Food Res. Technol., 217, 296300
(2003).
118. A. M. Alonso, D. A. Guillen, and C. G. Barroso, Eur. Food Res. Technol., 216, 445448
(2003).
119. N. J. Miller and C. Rice-Evan, Redox Rep., 2, 161171 (1996).
120. C. E. Rice-Evans and N. J. Miller, Methods Enzymol., 234, 279 (1994).
121. L. Bramati, F. Aquilano, and P. Pietta, J. Agric. Food Chem., 51, 74727474 (2003).
122. N. J. Miller and C. A. Rice-Evans, Free Rad. Res., 26, 195199 (1997).
123. N. Pellegrini, M. Serafini, B. Colobi, D. del Rio, S. Salvatore, M. Bianchi, and
F. Brighenti, J. Nutr., 133, 28122819 (2003).
124. I. E. F. Benzie and J. J. Strain, Anal. Biochem., 239, 70 (1996).
125. I. E. F. Benzie and J. J. Strain, Methods Enzymol., 299, 15 (1999).
126. J. Chen, H. Lindmark-Mansson, L. Gorton, and B. Akesson, Int. Dairy J., 13, 927935
(2003).
127. A. Ghiselli, M. Serafini, G. Maiani, E. Azzini, and A. Fero-Luzzi, Free Rad. Biol. Med.,
18, 2936 (1995).
128. O. Boxin, M. Hampsch-Woodill, and R. L. Prior, J. Agric. Food Chem., 49, 46194626
(2001).
129. S. McDonald, P. D. Prenzler, M. Antolovich, and K. Robards, Food Chem., 73, 7384
(2001).
130. S. N. El and S. Karakaya, Int. J. Food Sci. Nutr., 55, 6774 (2004).
131. M. Oktay, I. Gulcin, and O. I. Kufrevioglu, Lebensm. Wiss. U. Technol., 36, 263271
(2003).
9
Flavor Components
of Fats and Oils
Chi-Tang Ho1 and Fereidoon Shahidi2
1
Rutgers University
New Brunswick, New Jersey
2
Most of the flavor compounds in fats and oils are produced by the reaction of
oxygen with unsaturated fatty acids in triacylglycerols or polar lipids. On the other
hand, some flavor compounds such as those present in cocoa butter, roasted sesame
oil, or roasted peanut oil are generated by the interaction of reducing sugars with
amino compounds during thermal processing.
The development of objectionable flavor compounds by oxidation has significant
detrimental effects on consumer acceptability of edible oils. In the last four dec-
ades, much progress has been made in the chemistry of volatile products of lipid
oxidation, mainly as a result of advances in separation techniques and analytical
methodology, particularly gas chromatography-mass spectrometry.
1. FREE RADICAL AUTOXIDATION OF LIPIDS
The reaction of unsaturated lipids with oxygen to form hydroperoxides is generally

a free radical process involving three basic steps of initiation, propagation, and
termination (1, 2).
387
388 FLAVOR COMPONENTS OF FATS AND OILS
Initiation:
heat, light
RH R + H
metal
ROOH + Mn+ RO + M(n+1)+ + OH
ROOH + M(n+1)+ ROO + Mn+ + H+
2ROOH ROO + RO + H2O
Propagation:
R + O2 ROO
slow
ROO + RH ROOH
RO + RH ROH + R
Termination:
R + R R R
R + ROO ROOR
ROO + ROO ROOR + O2
RH, R, RO, ROO, ROOH, and M represent an unsaturated fatty acid or ester
with H attached to the allylic carbon atom, alkyl radical, alkoxy radical, peroxy
radical, hydroperoxide, and transition metal, respectively.
The initiation reaction is the hemolytic abstraction of hydrogen to form a car-
bon-centered alkyl radical in the presence of an initiator. Under normal oxygen
pressure, the alkyl radical reacts rapidly with oxygen to form the peroxy radical,
which in turn reacts with more unsaturated lipids to form hydroperoxides. The
lipid-free radical thus formed can further react with oxygen to form a peroxy radi-
cal. Hence, the autoxidation is a free radical chain reaction. Because the rate of
reaction between the alkyl radical and oxygen is fast, most of the free radicals
are in the form of the peroxy radical. Consequently, the major termination takes
place via the interaction between two peroxy radicals.
The rate of autoxidation increases with the degree of unsaturation. In neat sys-
tems without adding initiators, linoleate having two double bonds was 40 times
more reactive than oleate, which has only one double bond, linolenate having three
double bonds was 2.4 times more reactive than linoleate, and arachidonate having
four double bonds was 2 times more reactive than linolenate (2, 3).
2. HYDROPEROXIDES OF FATTY ACIDS OR THEIR ESTERS
It is well known that the free radical mechanism of hydroperoxide formation

involves the abstraction of a hydrogen atom from the a-methylene group of a lipid
HYDROPEROXIDES OF FATTY ACIDS OR THEIR ESTERS 389
molecule. This result is favored because of the formation of a very stable allyl radi-
cal in which the electrons are localized over either three carbon atoms such as in the
case of oleate or five carbon atoms such as in the case of linoleate and other poly-
unsaturated fatty esters. The mechanisms for the formation of isomeric hydroper-
oxides by autoxidation have been reviewed extensively (4, 5).
For oleate, the hydrogen abstraction on C-8 and C-11 produces two allylic radi-
cals. These intermediates react with oxygen to produce a mixture of 8-, 9-, 10-, and
11-allylic hydroperoxides. Autoxidation of linoleate involves hydrogen abstraction
on the doubly reactive allylic C-11, with the formation of a pentadienyl radical. The
intermediate radical reacts with oxygen to produce a mixture of conjugated 9- and
13-diene hydroperoxides. In the case of linolenate in which there are two separate
1,4-diene systems, hydrogen abstraction will take place on the two methylene
groups, C-11 and C-14. These intermediate free radicals react with oxygen to
form conjugated dienes with hydroperoxides on C-9 and C-13, or C-12 and C-
16, with the third double bond remaining unaffected.
2.1. Decomposition of Hydroperoxides

Hydroperoxides of unsaturated fatty acids formed by autoxidation are very unstable
and break down into a wide variety of volatile flavor compounds as well as nonvo-
latile products. It is widely accepted that hydroperoxide decomposition involves
homolytic cleavage of the OOH group, giving rise to an alkoxy radical and a
hydroxy radical (5).
R CH R R CH R
+ OH
OOH O
The alkoxy radical undergoes b-scission on the C C bond, with the formation of an
aldehyde and alkyl or vinyl radical. A general reaction scheme with the formation
of volatile aldehyde, alkene, and alcohol is illustrated in Figure 1 (6).
2.1.1. Aldehydes Of the volatiles produced by the breakdown of the alkoxy

radicals, aldehydes are the most significant flavor compounds. Aldehydes can be
produced by scission of the lipid molecules on either side of the radical. The pro-
ducts formed by these scission reactions depend on the fatty acids present, the
hydroperoxide isomers formed, and the stability of the decomposition products.
Temperature, time of heating, and degree of autoxidation are variables that affect
thermal oxidation (7).
Some volatile aldehydes formed by autoxidation of unsaturated fatty acids are
listed in Table 1. The aromas of aldehydes are generally described as green, painty,
metallic, beany, and rancid, and they are often responsible for the undesirable fla-
vors in fats and oils. Hexanal has long been used as an index of oxidative deteriora-
tion in foods. Some aldehydes, particularly the unsaturated aldehydes, are very
potent flavor compounds. Table 2 lists aroma characteristics of some common alde-
hydes found in fats and oils (8).
R1 CH CH CH R2
O OH
OH
(A) (B)
R1 CH CH CH R2
O
Scission A Scission B
R1 CH CH + R2 CHO R1 CH CH CHO + R2
O2 RH RH O2
R1 CH CH O O R2 O O
R1 CH CH2 R2H
RH RH
R1 CH CH O OH R2 O OH
OH OH
R1 CH CH O R2 O
RH
RH
R1 CH CH OH R2 OH
R1 CH2 CHO
Figure 1. General reaction pathway for the hemolytic cleavage of hydroperoxides of unsaturated
fats (6).
Hexanal and 2,4-decadienal are the primary oxidation products of linoleate. The
autoxidation of linoleate generates 9- and 13-hydroperoxides of linoleate. Cleavage
of 13-hydroperoxide will lead to hexanal and breakdown of 9-hydroperoxide will
lead to 2,4-decadienal (9). Subsequent moisture-mediated retro-aldol reaction of
2,4-decadienal will produce 2-octenal, hexanal, and acetaldehyde (10). 2,4-Deca-
dienal is known to be one of the most important flavor contributors to deep-fat fried
foods (11).
2,4-Decadienal can undergo further oxidation to produce trans-epoxy-trans-
decenal. This compound was recently characterized as one of the most potent odor-
ants of soybean oil stored in the dark and has a low odor threshold of approximately
1.5 pg/L (air) (12).
TABLE 1. Some Volatile Aldehydes Obtained from Autoxidation

of Unsaturated Fatty Acids (6).
Fatty Acid Monohydroperoxides Aldehydes Formed
Oleate 8-OOH 2-Undecenal

Decanal
9-OOH 2-Decenal
Nonanal
10-OOH Nonanal
11-OOH Octanal
Linoleate 9-OOH 2,4-Decadienal
3-Nonenaal
13-OOH Hexanal
Linolenate 9-OOH 2,4,7-Decatrienal
3,6-Nonadienal
12-OOH 2,4-Heptadienal
3-Hexenal
13-OOH 3-Hexenal
16-OOH Propanal
Arachidonate 5-OOH 2,4,7,10-Hexadecatetraenal
3,6,9-Pentadecatrienal
8-OOH 2,4,7-Tridecatrienal
3,6-Dodecadienal
9-OOH 3,6-Dodecadienal
11-OOH 2,4-Decadienal
3-Nonenal
12-OOH 3-Nonenal
15-OOH Hexanal
Eicosapentaenoate 5-OOH 2,4,7,10,13-Hexadecapentaenal
3,6,9,12-Pentadecatetraenal
8-OOH 2,4,7,10-Tridecatetraenal
3,6,9-Dodecatrienal
9-OOH 3,6,9-Dodecatrienal
11-OOH 2,4,7-Decatrienal
3,6-Nonadienal
12-OOH 3,6-Nonadienal
3-Hexenal
15-OOH 3-Hexenal
18-OOH Propanal
2.1.2. Ketones Aliphatic ketones formed by autoxidation of lipids also contri-

bute to the flavor of oils and food products. For example, Guth and Grosch (13)
identified 1-octen-3-one as one of the odor-active compounds in reverted soybean
oil. This compound was described as metallic and mushroom-like. The reaction
pathway for the formation of 1-octen-3-one from the linoleate-10-hydroperoxide
via the b-scission route is illustrated in Figure 2. 10-Hydroperoxide of linoleate
is not the usual hydroperoxide formed by autoxidation of linoleate; however, it is
one of the major hydroperoxides formed by the photosensitized oxidation (singlet
oxygen reaction) of linoleate (14).
TABLE 2. Sensory Characteristics of Selected Aldehydes.
Volatile Aldehyde Odor Characteristics
Pentanal Woody, bitter, oil

Hexanal Fatty, powerful, oily, grassy
Heptanal Oily, fatty, heavy, woody, penetrating, nutty
Octanal Fatty, sharp, citrus
Nonanal Fatty, waxy, painty, citrus
Decanal Penetrating, sweet, waxy, painty
Undecanal Fatty, tallowy
2-Hexenal Sweet, fragrant, almond. Fruity, green, leafy
2-Heptenal Oxidized, tallowy, pungent
2-Octenal Brown beans, herbaceous, spicy
2-Nonenal Penetrating, Fatty, waxy, nutty, rancid
2-Decenal Painty. Fishy, fatty
2-Undecenal Fresh, fruity, orange-peel like
2,4-Heptadienal Fatty, rancid, hazelnut-like
2,4-Decadienal Powerful, fatty, citrus
Modified from Morales et al. (8).
Linoleate
10
CH3 (CH2)4 CH CH CH2 CH CH CH (CH2)6 COOR
O OH
OH
CH3 (CH2)4 CH CH CH2 CH3 (CH2)4 CH CH CH2
O2
RH
O OH O O
OH
RH
O OH
(O)
CH3 (CH2)4 CH CH CH2

O
Figure 2. Mechanism for the formation of 1-octen-3-one from 10-hydroperoxide of linoleate.
CH3 (CH2)4 CH CH CH2 CH CH (CH2)7 COOR
O
CH3 (CH2)4 CH CH CH CH CH (CH2)7 COOR
O2, RH
CH3 (CH2)4 CH CH CH CH CH3 (CH2)4 CH CH CH CH
O OH
OH
CH3 (CH2)4 CH CH CH CH
O
H
H3C (CH2)4 H3C (CH2)4
O O
Figure 3. Mechanism for the formation of 2-pentylfuran (6).
2.1.3. Furan 2-Pentylfuran has been identified in many fats and oils. It is a well-
known autoxidation product of linoleate and has been known as one of the com-
pounds responsible for the reversion of soybean oil (15). Figure 3 shows the prob-
able mechanism for its formation. The conjugated diene radical generated from the
cleavage of the 9-hydroxy radical of linoleate may react with oxygen to produce
vinyl hydroperoxide. The vinyl hydroperoxide will then undergo cyclization via
the alkoxy radical to yield 2-pentylfuran (7).
2.1.4. Alcohols and Other Compounds Cleavage of lipid hydroperoxides will

also lead to alcohols, alkanes, alkenes, and alkynes. The mechanism for the forma-
tion of 1-octen-3-ol, which has a strong mushroom flavor, is also shown in Figure 2.
Because of their relative high odor threshold, alcohols and hydrocarbons are gen-
erally not considered to be important contributors to the flavors of fats and oils and
lipid-containing foods.
2.2. Singlet Oxygen Oxidation of Lipids

Oxidation of lipids occurs in the presence of molecular oxygen in both the singlet
and triplet states. Atmospheric oxygen that is in the triplet state contains two
unpaired electrons, whereas oxygen in the singlet state has no unpaired electrons
(16, 17). The electron arrangement of triplet oxygen does not allow for a direct
reaction of lipid molecules that exist in the singlet state. Singlet oxygen can be
1O
R1 2 R1
R2 R2
H H O O
6-membered ring
OOH
R1
R2
Figure 4. The ene reaction of singlet oxygen with unsaturated fatty acid.
generated by the interaction of light, photosensitizers, and oxygen. Singlet oxygen

has been suggested to be responsible for initiating lipid oxidation of food products
because of its ability to directly react with linoleic acid at least 1480 times faster
than triplet oxygen (17).
The ene reaction of singlet oxygen is important in the photoxidation process
in edible oils. In this reaction, singlet oxygen reacts with olefins to form allyl hydro-
peroxides through a six-membered ring transition state as shown in Figure 4. The
lipid hydroperoxides produced from singlet oxygen are different from those gener-
ated by autoxidation. In singlet oxygen oxidation, both conjugated and nonconju-
gated hydroperoxides are formed; in free radical autoxidation, nonconjugated
hydroperoxides are not usually formed (17).
One of the major volatile compounds formed from cottonseed oil was 1-decyne,
which was reported as a predominant volatile of photooxidized cottonseed oil
formed from its precursor, sterculic acid, as shown in Figure 5. The endogenous
trace amount of chlorophyll acts as a photosensitizer to produce singlet oxygen.
The singlet oxygen then attacks the cyclopropenoid fatty acid, namely, sterculic
acid, followed by degradative cleavage and molecular rearrangement to yield
1-decyne (18).
CH2 1
CH2
O2
CH3(CH2)6CH2C C (CH2)7COOH CH3(CH2)6CH C C (CH2)7COOH
OOH
Sterculic acid
CH2 CH
CH3(CH2)6CH C O CH3(CH2)6CH2 C O
C (CH2)7COOH C (CH2)7COOH
OH OH
CH3(CH2)7 C CH + HOOC(CH2)7 COOH
1-decyne
Figure 5. Proposed mechanism for the formation of 1-decyne from the photooxidation of
sterculic acid (18).
MAJOR VOLATILE COMPOUNDS OF COMMERCIAL FATS AND OILS 395
3. MAJOR VOLATILE COMPOUNDS OF COMMERCIAL

FATS AND OILS
The volatile compounds of eight different vegetable oils, namely, canola, corn, cot-
tonseed, olive, peanut, safflower, soybean, and sunflowerseed oil, have been ana-
lyzed and reported by Snyder et al. (19). Table 3 shows the quantitative data of
the volatile compounds identified in these vegetable oils after eight days of storage
at 60 C.
The volatile compounds in each of the stored vegetable oil samples were related
to the main fatty acid components of the oil. Safflower, sunflowerseed, corn, and
TABLE 3. Volatile Compounds in Vegetable Oils After 8 Days Storage at 60 C (19).
Volatile GC Peak Area

Compounds Canola Corn Cottonseed Olive Peanut Safflower Soybean Sunflowerseed
Ethane 1.0 1.0 0.8 0.2 8.0 0.2 0.9 1.0

Propane 0.7 10.6 12.6 0.6 10.4 9.1 12.7
Propenal 1.5 2.3 2.1 1.7 2.4 2.3 4.5 2.9
Pentene 0.2
Pentane 0.5 24.3 32.7 14.9 19.0 54.1 11.1 41.5
Propanal 1.0 0.3 0.6 0.7
Pentene 0.1 0.1 0.1 0.1
Hexane 1.1 0.4 0.3 0.3 0.2 0.4 0.3 0.1
2-Butenal 0.5 0.2 0.2 0.3 0.5 0.1
1-Penten-3-ol 0.4 0.8 0.1
Pentanal 0.5 1.0 1.3 1.8 2.8 1.9 2.9 2.4
Heptane 0.2 0.2 2.2 0.5 0.2 0.1 0.4
Pentenal 0.3 0.1 0.1 0.6
Pentanol 0.3 0.5 0.8 0.5 1.4 2.0 1.5 1.3
Octene 0.1 0.1 0.2 0.5 0.1 0.2
Hexanal 3.1 9.8 10.3 5.8 7.6 11.1 10.7 10.5
Octane 3.7 2.2 3.9 4.1 0.3 0.2 0.3 4.5
Octene 0.1 0.1 0.2 0.1 0.2
t-2-Hexenal 0.2 0.3 0.5 0.2 0.8 0.3 0.5 0.7
Heptanal 0.8 0.6 2.2 1.8 2.0 3.2 2.1 1.8
c-2-Heptenal 0.2 0.2 0.1 0.1 0.3 0.2
t-2-Heptenal 0.9 1.3 3.6 2.6 1.5 5.6 5.1 3.4
1-Octen-3-ol 0.1 0.3 0.2 0.3 0.3 0.1 0.3 0.3
2-Pentylfuran 0.2 0.5 0.8 0.2 0.7 0.7 1.0 0.6
t,c-2,4-Heptadienal 0.5 0.6
Octanal 0.5 0.3 0.4 1.1 0.6 0.4 0.5 0.3
t,t-2,4-Heptadienal 0.8 0.8
Octenal 0.2 0.5 0.5 0.6 0.7 0.6 0.4 0.5
Nonanal 1.2 0.5 0.6 2.8 0.5 0.4 0.6 0.4
t-2-Decenal 0.2 0.2 0.3 0.7 0.2 0.2 0.3 0.3
Decenol 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1
t,c-2,4-Decadienal 0.3 0.3 0.4 0.1 0.3 0.3 0.6 0.4
t,t-2,4-Decadienal 0.8 1.0 1.2 0.8 0.8 1.5 1.4 1.4
Undecenal 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.2
Total area 25.8 63.1 77.6 49.2 53.6 99.6 40.2 70.4
TABLE 4. Odor Profiles of the Butter Samples (29).
Butter Samples Odor Quality Intensity a
Irish sour cream (ISC) Buttery, creamy, sweet 3

Cultured butter (CB) Typical butter-like, sweet 23
Sour cream (SC) Mild, weakly buttery, sour 12
Sweet cream (SwC) Slightly sour mild 1
Farmer sour cream (FSC) Rancid, like butanoic acid 3
a
Intensity: 1, weak; 2, medium; 3, strong.
cottonseed oils, with the highest amount of linoleate, tended to produce the greatest
amount of volatiles, especially pentane and hexanal. Canola and soybean oils,
which contain linolenate, both formed measurable amounts of 2,4-heptadienal.
Olive oil, with the largest quantity of oleate, produced the most octanal and
nonanal.
3.1. Flavor Compounds of Selected Fats and Oils

3.1.1. Butter Because of their commercial significance, the flavor of butter and
butter oil has been studied extensively. More than 230 volatile compounds have
been identified in different types of butter as well as butter oil (20). The typical
flavor of fresh butter is influenced by carbonyl compounds formed by oxidation
of unsaturated fatty acids in milk. Critical flavors in butter have recently been
reviewed (21, 22).
Using odor activity value (ratio of concentration to odor threshold), Forss et al.
(23), Urbach et al. (24), and Stark and co-worker (2527) reported d-decalactone,
d-octalactone, decanoic acid, dodecanoic acid, skatole, and indole as important
contributors to the flavor of butter oil. In addition, the data of Siek et al. (28) indi-
cated that in fresh butter, the levels of butanoic acid, caproic acid, d-decalactone
were above their taste threshold.
By using aroma extract dilution analysis (AEDA) of the volatile fractions of
fresh and stored butter oil, Widder et al. (29) determined diacetyl, butanoic acid,
d-octalactone, skatole, d-decalactone, cis-6-dodeceno-d-decalactone, 1-octen-3-
one, and 1-hexen-3-one as potent contributors to the flavor of butter oil. The con-
centration of 1-octen-3-one, trans-2-nonenal, and cis-1,5-octadien-3-one increased
during the storage of the butter oil at room temperature.
Table 5 shows the sensory evaluation by Schieberle et al. (30) of the different
kinds of butter, namely, Irish sour cream (ISC), cultured butter (CB), sour cream
(SC), sweet cream (SwC), and farmer sour cream (FSC). It revealed ISC butter
and FSC butter with the highest overall odor intensities. Table 5 shows that 19
odor-active compounds were detected by aroma extract dilution analysis (AEDA)
in a distillate of the ISC butter. The highest flavor dilution (FD) factors have been
found for d-decalactone, skatole, cis-6-dodeceno-g-lactone, and diacetyl followed
by trans-2-nonenal, cis,cis-3,6-nonadienal, cis-2-nonenal, and 1-octen-3-one.
TABLE 5. Potent Odorants in an Irish Sour Cream Butter (30).
Flavor Dilution
Compound (FD) Factor Odor Description
Diacetyl 256 Buttery

1-Penten-3-one 32 Vegetable-like
Hexanal 8 Green
1-Octen-3-one 8 Mushroom-like
cis,cis-3,6-Nonadienal 64 Soapy
cis-2-Nonenal 64 Fatty, green
trans-2-Nonenal 128 Green, tallowy
trans,trans-2,4-Nonadienal 8 Fatty, waxy
trans,trans-2,4-Decadienal 32 Fatty, waxy
g-Octalactone 64 Coconut-like
trans-4,5-Epoxy-trans-2-decenal 32 Metallic
Skatole 512 Mothball-like
d-Decalactone 4096 Coconut-like
cis-6-Dodeceno-g-lactone 512 Peach-like
Acetic acid 128 Pungent
Butanoic acid 512 Buttery, sweaty
Hexanoic acid 32 Pungent, musty
Aroma extract dilution analysis has also been applied to heated butter (31). Key
aroma compounds with highest FD factors have been identified as d-octalactone,
skatole, methional, d-decalactone, 2,5-dimethyl-4-hydroxy-3(2H)-furanone (fura-
neol), and 1-octen-3-one followed by 1-hexen-3one, cis-2-nonenal, trans-2-
nonenal, trans, trans-2,4-decadienal, trans-4,5-epoxy-trans-2-decenal, and g-octa-
lactone. Unsaturated triacylglycerols in butterfat are presumed to generate these
potent odorants during heating. Additionally, thermal degradation of proteins and
Maillard reaction can account for the formation of skatole, methional, and furaneol.
3.1.2. Cocoa Butter Cocoa butter is one of the most liked and highly prized
food ingredients because of its desirable flavor and unique melting behavior. As
early as 1961, van Elzakker and van Zutphen (32) studied and identified 23 volatile
compounds in the vacuum steam distillate of cocoa butter. Later, Rizzi (33) identi-
fied nine alkylpyrazines including methylpyrazine, 2,5-dimethylpyrazine, 2,6-
dimethylpyrazine, 2,3-dimethylpyrazine, 2-ethyl-5-methylpyrazine, trimethylpyra-
zine, 2,5-dimethyl-3-ethylpyrazine, 2,6-dimethyl-3-ethylpyrazine, and tetramethyl-
pyrazine in the basic fraction of a vacuum steam distillate of cocoa butter.
The flavor of cocoa butter depends on the processing conditions to which the
cocoa beans are subjected. Cocoa butter obtained from roasted cocoa beans has a
strong flavor reminiscent of cocoa. Cocoa butter obtained from unroasted cocoa
beans that have been given a steam treatment has a considerable milder, yet distinc-
tive, flavor.
The most comprehensive study on the flavor compounds of cocoa butter was that
of Carlin et al. (3437). They compared the volatile compounds of cocoa butters
TABLE 6. Some Major Pyrazines Identified in Cocoa Butters from Roasted and Unroasted
Cocoa Beans (37).
Relative Concentration

Pyrazines Roasted Unroasted
Tetramethylpyrazine 270 3400

Trimethylpyrazine 2620 380
2,5-Dimethyl-3-ethylpyrazine 750 40
2,5-Dimethylpyrazine 710 60
2-Isopropyl-3-methylpyrazine 510 trace
2,6-Dimethylpyrazine 490
2-Acetyl-3-methylpyrazine 390 20
2-Ethyl-3,5,6-trimethylpyrazine 310 20
2,6-Diethyl-3-methylpyrazine 10 150
6,7-Dihydro-5H-cyclopentapyrazine 110
Methylpyrazine 90 10
2-Acetyl-3-ethylpyrazine 90 10
2,5-Diethyl-3-methylpyrazine 90
2-Butyl-3,6-dimethylpyrazine 80 trace
2-Methyl-6,7-Dihydro-5H-cyclopentapyrazine 80 trace
2-Methyl-5-vinylpyrazine 70
Isopropenylpyrazine 70
2-Methyl-3-pentylpyrazine 50 c
2-Methyl-6-vinylpyrazine 40 30
2,3-Dimethylpyrazine trace 30
2,3-Dimethyl-5-ethylpyrazine 30
2,3-Dimethyl-5-butylpyrazine 30
2,5-Dimethyl-5-isobutylpyrazine 30
5,6,7,8-Tetrahydroquinoxaline 30
from roasted and unroasted cocoa beans. Pyrazines were present in greater numbers
and at higher concentrations in the roasted cocoa butter. Of the 62 pyrazines iden-
tified, 57 were identified in the roasted cocoa butter and only 27 in the unroasted
samples. Table 6 lists the comparison of the major pyrazines identified in cocoa
butters from roasted and unroasted cocoa beans.
The most abundant pyrazine identified in cocoa butters was tetramethylpyrazine,
which existed at an extremely high concentration in the unroasted cocoa butter but
only a moderate level in the roasted cocoa butter. Tetramethylpyrazine accounted
for over 90% of the pyrazine content of the unroasted cocoa butter. Besides thermal
generation, tetramethylpyrzine could be formed in cocoa beans through biosyn-
thetic reactions. Kosuge and Kamiya (38) identified tetramethylpyrazine as a meta-
bolic product of a strain of Bacillus subtilis. Several species of this organism were
identified in a fermenting mass of cocoa beans by Ostovar (39).
Table 7 lists the oxazoles and thiazoles identified in the sample of cocoa butter.
They were present only in roasted cocoa butter. The sensory characteristics of these
compounds shown in Table 7 indicated that oxazoles and thiazoles possessed inter-
esting green, fatty, sweet, and nutty sensory qualities and were high-impact flavor
TABLE 7. Oxazoles and Thiazoles Identified in Roasted Cocoa Butter (36).
Compounds Odor Description
2-Pentylthiazole Strong, green, fatty, sweet

2-Acetyl-5-methylthiazole
2-Isopropyl-4,5-dimethylthiazole
2-Isopropyl-4-methylthizaole Strong, camphorus
2-Acetyloxazole
2-Isopropyl-4,5-dimethyloxazole
2-Isopropyl-4-ethyl-5-methyloxazole Sweet, fruity
2-Methyl-4,5-dibutyloxazole
4,5-Dimethyloxazole
2-Methyl-4-ethyl-5-propyloxazole Green fatty, vegetable-like
2,5-Dimethyl-4-butyloxazole Fresh acidic, green, pickle-like
2-Methyl-4-ethyl-5-butyloxazole Acidic, fatty, sweet, flowery
2-Butyl-4-methyl-5-ethyloxazole Green, sweet
2-Butyl-4-ethyl-5-methyloxazole Green, herbal, weak, acidic, slight buttery
4,5-Dibutyloxazole
2,5-Dibutyl-4-methyloxazole Sweet, fruity, green
compounds in roasted cocoa butter (36, 37). Of particular interest, 2-pentylthiazole

identified had strong fatty, green, and sweet notes and may be an important contri-
butor to cocoa butter flavor. Oxazoles could possibly be formed through the
Strecker degradation of aminoketones, which result from the condensation of
a-dicarbonyl compounds with amino acids (40). They might also form through
reaction between amino acids (41). Maga (42) has reviewed the occurrence of thia-
zoles in foods and possible pathways of formation. Thiazoles could possibly form
through the interaction of sulfur-containing amino acids and carbonyl-containing
compounds.
3.1.3. Lard Lard is a traditional edible fat for Chinese people. Lard is generally
prepared either by dry-rendering or by wet-rendering. The dry-rendered lard with
pork back fat as the raw material usually has better flavor than the wet-rendered lard
and is used as cooking fat or shortening. The wet-rendered lard with pork belly fat
as the raw material usually has an undesirable flavor and must be refined before
further use.
The volatile flavor compounds of lard have been studied by Watanabe and Sato
(4348). They heated the lard at 160170 C under a stream of air and collected the
volatile compounds. They found that 2,4-decadienal and lactones contributed sig-
nificantly to the flavor of lard. Hwang and Chen (49) compared the volatile flavor
compounds generated by heating the crude and refined samples of both dry-ren-
dered and wet-rendered lard at 190 C for 2 hours. Table 8 summarizes the amounts
of some flavor-contributing volatiles in different samples of lard. Crude dry-ren-
dered lard showed the highest content of these compounds followed by crude
wet-rendered lard, refined dry-rendered lard, and refined wet-rendered lard. Appar-
ently, dry-rendering can yield lard with a stronger flavor.
TABLE 8. The Amounts of Volatile Flavor Contributing Compounds from Lards

of Different Treatments (47).
CDLa RDL CWL RWL

Compounds (mg/100-g lard)
1-Octen-3-ol 2.21 1.83 1.89 1.40

n-Hexanal 10.87 8.13 10.38 7.60
n-Octanal 2.67 2.12 2.07 1.86
trans-2-Heptenal 11.04 8.45 10.44 7.23
n-Nonanal 10.92 8.15 8.70 6.99
trans-2-Octenal 3.24 2.66 2.77 2.40
trans-2-Nonenal 3.14 2.46 2.59 1.89
2,4-Decadienal 17.01 14.34 15.06 12.10
2-Pentylfuran 0.80 0.43 0.43 0.38
g-Octalactone trace 0.38 trace trace
Total 61.90 48.95 54.78 41.85
a
CDL, crude dry-rendered lard; RDL, refined dry-rendered lard; CWL, crude wet-rendered lard; and RWL,
refined wet-rendered lard.
3.1.4. Soybean Oil Soybean oil is the highest volume vegetable oil produced in
the world, as well as in the United States. Because of its commercial importance,
the flavor chemistry of soybean oil has been extensively studied and reviewed (50).
The development of a characteristic, objectionable, beany, grassy, and hay-like
flavor in soybean oil, commonly known as reversion flavor, is a classic problem of
the food industry. Soybean oil tends to develop this objectionable flavor when its
peroxide value is still as low as a few meq/kg, whereas other vegetable oils, such as
cottonseed, corn, and sunflower, do not (15, 51). Smouse and Chang (52) identified
71 compounds in the volatiles of a typical reverted-but-not-rancid soybean oil. They
reported that 2-pentylfuran formed from the autoxidation of linoleic acid, which is
the major fatty acid of soybean oil, and contributes significantly to the beany and
grassy flavor of soybean oil. Other compounds identified in the reverted soybean oil
also have fatty acids as their precursors. For example, the green bean flavor is
caused by cis-3-hexenal, which is formed by the autoxidation of linolenic acid that
usually constitutes 211% in soybean oil. Linoleic acid oxidized to 1-octen-3-ol,
which is characterized by its mushroom-like flavor (53).
The most interesting studies on the flavor of soybean oil were those published
by Ullrich and Grosch (54) and Guth and Grosch (13, 55). By using aroma extract
dilution analysis, they determined some odor compounds that strongly contributed
to the off-flavor of soybean oil samples, which were stored at room temperature
either in daylight or in the dark. Table 9 lists the FD factor values of various
odor compounds in soybean oil samples. 3-Methyl-2,4-nonanedione, cis-3-hexenal,
cis-2-nonenal, cis-1,5-octadien-3-one, 1-octen-3-hydroperoxide, 4,5-epoxy-trans-
2-decenal, 1-octen-3-one, cis-1,5-octadien-3-hydroperoxide, and trans-2-nonenal
were identified as primary odorants of soybean oil, which were exposed to daylight.
They also observed that the major differences in the intensity of the reversion odor
TABLE 9. Aroma Extract Dilution Analysis of the Stored Soybean Oils (12).
Flavor Dilution (FD) Factor

Compound SBO (Daylight)a SBO (Dark)b
cis-3-Hexenal 2048 8
Hexanal 16 <1
Pentanoic acid 32 <1
trans-2-Heptenal 64 <1
1-Octen-3-one 256 16
cis-1,5-Octadien-3-one 512 <1
2,4-Heptadienal 8 4
Octanal 16 8
cis-2-Octenal 16 <1
trans-2-Octenal 32 8
cis-3-Nonenal 16 8
1-Octen-3-hydroperoxide 512 32
cis-1,5-Octadien-3-hydroperoxide 256 4
cis-2-Nonenal 1024 64
trans,cis-2,4-Nonadienal 16 8
trans-2-Nonenal 256 32
2,4-Nonadienal 32 16
trans,trans-2,4-Nonadienal 16 16
3-Methyl-2,4-nonanedione 16348 16
trans-4,5-Epoxy-trans-2-nonenal <1 8
trans,trans-2,4-Decadienal 32 16
trans-4,5-Epoxy-trans-2-decenal 512 256
a
The oil sample was stored for 30 days at room temperature and in daylight.
b
The oil sample was stored for the sample period in the dark.
of soybean oil samples were mainly caused by an increase in the concentration of

3-methyl-2,4-nonanedione during storage. Guth and Grosch (56) further identified
two furanoid fatty acids, namely, 10,13-epoxy-11,12-dimethyloctadeca-10,12-die-
noic acid and 12,15-epoxy-13,14-dimethyloctadeca-12,14-dienoic acid in soybean
oil as precursors for 3-methyl-2,4-nonanedione. The proposed mechanism for the
photogeneration of 3-methyl-2,4-nonanedione is shown in Figure 6. However, the
work of Kao et al. (57) cannot support the theory that furanoid fatty acids or
3-methyl-2,4-nonanedione contribute strongly to the reversion flavor of soybean
1O
2
O
(CH2)7COOH OH (CH2)7COOH
O O
O
+ (CH2)7COOH
OH O O O
Figure 6. Proposed mechanism for the photogeneration of 3-methyl-2,4-nonanedione (56).
TABLE 10. Odor Properties of the Hydroperoxides and the Epoxides Identified in Stored
Soybean Oils (12).
Odor Threshold

Compound Odor Description (ng/ L, air)
1-Octen-3-hydroperoxide Metallic, mushroom-like 0.61.2

1-Octen-3-one Mushroom-like 0.030.12
cis-1,5-Octadien-3-hydroperoxide Geranium-like, metallic 0.030.06
cis-1,5-Octadien-3-one Geranium-like, metallic 0.0030.006
trans-4,5-epoxy-trans-2-nonenal Metallic 0.251.0
trans-4,5-epoxy-trans-2-decenal Metallic, green 0.00050.005
oil. No significant flavor differences were found when soybean oils with high or low
contents of furanoid fatty acids were evaluated during storage for off-flavor inten-
sity of soybean oil (57).
The odor properties of two volatile hydroperoxides, 1-octen-3-hydroperoxide
and cis-1,5-octadien-3-hydroperoxide, are shown in Table 10. The odor thresholds
of these hydroperoxides were 10-fold higher than those of the corresponding
ketones. Precursors of both hydroperoxides are presumably the 10-hydroperoxide
of linoleic acid and linolenic acid, which are easily formed by photosensitized oxi-
dation of linoleic acid and linolenic acid (48). As shown in Figure 7, a b-scission of
the 10-hydroperoxy group, the rearrangement of the double bond and combination
of the allylic radical formed with oxygen, followed by abstraction of a hydrogen
atom would result in the two allyl hydroperoxides having eight carbon atoms.
As shown in Tables 9 and 10, on the basis of its high FD factor and its odor
properties, the trans-4,5-epoxy-trans-2-decenal contributed significantly to the
green, hay-like overall odor in soybean oil stored in the dark. Guth and Grosch sug-
gested (13) that an epoxyhydroperoxy fatty acid could be the precursor of such
epoxy aldehydes. Figure 8 shows the proposed pathway for the formation of
trans-4,5-epoxy-trans-2-decenal from linoleic acid via the trans-12,13-epoxy-9-
hydroperoxy-trans-10-octadecenoic acid intermediate.
3.1.5. Canola Oil Canola oil is obtained from low erucic acid, low glucosinolate
rapeseed. The unique polyunsaturated fatty acid and low saturated composition of
canola oil differentiates it from other oils. It has a higher oleic acid (18:1) content
(55%) and lower linoleic acid (18:2) content (26%) than most other vegetable oils,
but it contains 812% of linolenic acid (18:3) (58). Canola oil is most widely used
in Canada and is considered a nutritionally balanced oil because of its favorable
ratio of near 2:1 for linoleic to linolenic acid content. Unlike most other edible
oils, the major breakdown products of canola oil are the cis, trans- and trans, trans-
2,4-heptadienals with an odor character generally described as oily, fatty, and putty.
Stored canola oil shows a sharp increase in the content of its degradation products,
which are well above their odor detection thresholds. The aroma is dominated by
cis, trans-, trans, trans-2,4-heptadienals, hexanal, nonanal, and the cis, trans- and
O O
(CH2)6 COOH
R
O OH
H (CH2)6 COOH
O
R
O O
O OH
CH2
Figure 7. Reaction routes proposed for the formation of 1-octen-3-hydroperoxide (R
(CH2)4
) and cis-1,5-octadien-3-hydroperoxide (CH3
CH2 CH CH CH2) (13).
trans, trans-decadienals. Octen-3-one with a strong metallic flavor, generally con-

sidered an off-flavor in oils, occurs at a concentration of 75 ppb in aged oil, which is
approximately 750 times above its odor detection threshold. Despite its relatively
low concentration, the metallic note of octen-3-one contributes markedly to the
overall flavor character of the aged oil (59).
CH3 (CH2)4 CH CH CH2 CH CH (CH2)7 COOH
CH3 (CH2)4 CH CH CH CH CH (CH2)7 COOH

OOH
O2, H
OH
CH3 (CH2)4 CH CH CH CH CH (CH2)7 COOH
O O OH
CH3 (CH2)4 CH CH CH CH CHO

O
Figure 8. Proposed reaction route to the trans-4,5-epoxy-trans-2-decenal via the trans-12,13-
epoxy-9-hydroperoxyoctadec-(E)-10-enoic acid (13).
3.1.6. Olive Oil Olive oil is commonly used as a table and cooking oil because of
its unique flavor and stability. Olive oil comes from the fruits of the olive tree Olea
europea, which has been cultivated for many years in the Southern European coun-
tries bordering the Mediterranean and in North Africa. Virgin olive oil is extracted
under mild conditions and is normally consumed without further treatment, the nat-
ural flavor compounds that confer its characteristic aroma are preserved and are
uniquely recognized by consumers. In the intact fruit, a high percentage of fatty
acids, exists which are mainly bound as triacylglycerols; however, in the oil, there
is a resulting high percentage of free fatty acids. Two reasons could account for the
high levels of free fatty acid: A high moisture content that is favorable to lipase
action, and bruising of the olive fruit during harvest, transportation, and milling
of the fruits. The primary free fatty acids are the unsaturated oleic, linoleic, and
linolenic acids, with oleic acid constituting 80% of the free fatty acids. These
free fatty acids contribute significantly to the taste of the oil and serve as the pre-
cursors for the aroma compounds of olive oil.
The aroma compounds of olive oil have recently been reviewed by Kiritsakis
(60). The most abundant aroma compounds in virgin olive oil are C-6 aliphatic
compounds, trans-2-hexenal, trans-2-hexen-1-ol, hexan-1-ol, cis-3-hexen-1-ol,
cis-2-penten-1-ol, cis-3-hexenal, hexyl acetate, and hexanal, accounting for about
80% of total volatile compounds with the prominence of trans-2-hexenal (61
64). These C-6 compounds provide the green perception and unique aroma of olive
oil. Recently, 2,4-dimethylfuran has been found in olive oil with unpleasant sensory
quality (65). As a result, the ratio of trans-2-hexenal/2,4-dimethylfuran has been
proposed as a quality marker for olive oil and the ratio value of less than 1.5 indi-
cates lower quality olive oil (65).
The stability of olive oil compared with other vegetable oils is attributed to the
high-to-low ratio of oleic to linoleic acid, and to the degradation of the chlorophylls
to pheophytins (60). In addition, olive oil is also rich in antioxidative phenolic com-
pounds such as hydroxytyrosol (66).
3.1.7. Marine Oils Since the use of sardine oil as a food ingredient was discon-
tinued in the 1950s, the Food and Drug Administration (FDA) has determined that
fish oils were totally new ingredients for human foods. As a result, 90% of the fish
oil produced in the United States was exported to Europe as a food oil and 10% was
used domestically in nonfood applications. After lengthy petition, the FDA finally
affirmed the GRAS status of partially hydrogenated (PHMO) and hydrogenated
menhaden oil (HMO) for direct use as human food ingredients in 1989 (67).
Fish oils are a rich source of o-3 polyunsaturated fatty acids such as eicosapen-
taenoic acid (EPA) and docosahexaenoic acid (DHA). The supplementation of Wes-
tern diets with fish oils containing EPA and DHA has been recommended (68, 69),
and it can be beneficial for ischemic heart disease and thromboembolic events.
In order to explain the reason that marine oils are much more labile than com-
mon vegetable oils, basic lipid chemistry, such as oxidation rate, induction period,
and oxygen uptake were determined and reported by several researchers (7072).
The relative oxidation rates of fatty esters at 36.5 C were found to be highly
18 15 12 9 6 3
COOH
Initial H Abstraction Hydroperoxide Aldehydes Formed
C-18 20-OOH Propanal

C-15 17-OOH 3-Hexenal
13-OOH 2,4,7-Decatrienal
C-12 14-OOH 3,6-Nonadienal
10-OOH 2,4,7,10-Tridecatetraenal
C-9 11-OOH 3,6,9-Dodecatrienal
7-OOH 2,4,7,10,13-Hexadecapentaenal
C-6 8-OOH 3,6,9,12-Pentadecatetraenal
4-OOH 2,4,7,10,13,16-Nonadecahexaenal
Figure 9. Aldehydes derived from autoxidized decosahexaenoic acid.
correlated to their molecular unsaturation, i.e., oleate (1.0), linoleate (8.0), linole-
nate (21.7), and EPA DHA (39.1). Relative oxygen uptake (first two days in air)
of oleate, linoleate, linolenate, EPA, and DHA were <1, 1, 99, 743, and 948, respec-
tively. Induction periods (90 lux at 5 C) were also related to the degree of unsatura-
tion of fatty esters. The induction period of oleate was found to be more than
100 days (estimated) and those of linoleate, linolenate, and EPA DHA, were
50, 20, and 4 days, respectively (68).
The high oxidation rates of EPA and DHA and the instability of their hydroper-
oxides caused the rapid formation of secondary products such as volatile aldehydes
and other compounds, which, in turn, impart flavor reversion in fish oils (56). The
hydroperoxides produced from autoxidation of EPA (73) and DHA (74) have been
identified but not quantified. They form eight and ten isomers, respectively. Noble
and Nawar (75) analyzed the volatile compounds in autoxidized DHA and identi-
fied a number of aldehydes. Most of the aldehydes identified could be explained by
the b-scission of alkoxy radicals generated by the homolytic cleavage of each iso-
mer of the hydroperoxides as shown in Figure 9.
Meijboom and Stroink (76) found that 2-trans, 4-cis, 7-cis-decatrienal was the
compound responsible for the fishy off-flavors occurring in autoxidized oil contain-
ing o-3 fatty acids. This trienal was also found in autoxidized methyl DHA by
Noble and Nawar (75) and in autoxidized mackerel oil by Ke et al. (77).
The most detailed studies on the flavor of fish oil in recent years were probably
those of Hsieh et al. (78, 79), Lin et al. (80), and Lin (81). In their studies, a series
of alkanals, alkenals, alkadienals, and alkatrienals were determined by dynamic
headspace gas chromatography-mass spectrometry in crude menhaden oils (Table 11).
Most of these aldehydes contributed to the characteristic oxidized oily odors, such
as green grassy, waxy, and rancid in the crude oils. Alkatrienals, i.e., nonatrienal
and decatrienals, were also found at ppb levels in the dynamic headspace of
the crude oils. 2-trans,4-trans,7-cis-Decatrienal, 2-trans,4-cis,7-cis-decatrienal,
TABLE 11. Volatile Aldehydes Identified in the Gulf

Menhaden Oil (54).
Compounds GC Area (%)
n-Butanal 0.97
n-Pentanal 0.82
cis-2-Butenal 0.55
n-Hexanal 1.56
cis-2-Pentenal 1.48
n-Heptanal 1.42
trans-2-Hexenal 0.97
trans-4-Heptenal 0.24
n-Octanal 0.61
trans-2-Heptenal 0.31
n-Nonanal 0.51
2,4-Hexadienal trace
cis-2-Octeenal 0.13
trans-2-Octenal 0.43
2,4-Heptadienala 0.69
n-Decanal 0.15
2,4-Heptadienala 1.39
Benzaldehyde 0.34
trans-2-Nonenal 0.14
cis-4-Decenal 0.14
2,4-Octadienala 0.11
2,6-Nonadienala 0.06
2,4-Octadienala 0.34
trans-4-Decenal 0.02
2,4-Nonadienala trace
2,4-Nonadienala trace
2,4-Decadienal trace
2,4-Undecadienala trace
Nonatrienala trace
Nonatrienala trace
Decatrienala trace
Decatrienala trace
a
Configuration of geometric isomers were not determined.
and 4-cis-heptenal impart a strong fishy odor to oils. Aldehydes have a green
or plant-like note; ketones (1-octen-3-ol) have a metallic off-flavor. Besides the
aldehydes identified in crude menhaden oil, other compounds identified such as
short-chain unsaturated alcohol (1-penten-3-ol), had a medicinal odor, and others
had a green unpleasant odor and may also contribute to the flavor of fish oil.
Lin et al. (80) also reported that steam-deodorization can effectively remove a
total of 99% of most aldehydes in the oils. However, reversion flavor of fish oils
during storage can generate pentylfuran and aldehydes that have green and beany
odors.
Cadwallader and Shahidi (82) identified the potent odorants of seal blubber oil
by direct thermal desorption-gas chromatography-olfactometry (DTD-GCO). In
general, odorants were present at higher odor-potency in the crude oil. Predominant
odorants were (Z)-1,5-octadien-3-one (metallic), (E, E, Z)-2,4-7-decatrienal (fatty,
fishy), (Z)-3-hexenal (green, cut-leaf), and (E,Z)-2,6-nonadianal, (cucumber).
These compounds are among breakdown products of thermally labile hydroper-
oxides and are responsible for the typical off-flavors encountered in stored marine
oils.
Oxidative stability of o-3 fatty acids can be increased using free radical scaven-
gers. TBHQ (t-butylhydroquinone) at a concentration of 0.02% has successfully
slowed down the oxidation of menhaden oil for up to 40 days, compared with
3 days for the control group (83). a-Tocopherol and butylated hydroxytoluene
(BHT) alone or in combination increased the oxidative stability of EPA and
DHA (84). The most notable success in fish oil stabilization has been achieved
with ternary antioxidant systems, which contain a- or g-tocopherol concentrates,
ascorbic acid (or ascorbyl palmitate), and lecithin (85).
3.1.8. Sesame Oil Sesame oil has traditionally been used in eastern Asian coun-
tries, especially China, Japan, Korea, and Taiwan. It has been prized for its nutritive
and health-promoting values. Sesame oil, prepared from roasted sesame seeds, has
a distinctive flavor and a long shelflife (86). Several studies have been reported on
the flavor components of sesame oil (8791). The amount of volatile flavor com-
pounds in sesame oil is greatly affected by the roasting process. It has been reported
that the ratio of the amount of volatile components in deep-roasted oils was
increased by 27 times in deep-roasted oil as compared with that of light-roasted
oils (90).
Perhaps the most important compounds identified in the roasted sesame oils are
2-furfurylthiol and guaiacol. Using aroma extract dilution analysis method, these
two compounds have been characterized by Schieberle (92) to be the most odor-
active compounds in roasted sesame seeds. 2-Furfurylthiol, having an intense
coffee-like odor, increased from 16 ppb in roasted oil processed at 160 C for
30 min to 158 ppb in the oil processed at 200 C for 30 min (Table 12). Guaiacol
has a burnt and smoky odor with an extremely low-odor threshold of 0.02 ppt in
TABLE 12. Changes in the Content (ppb) of Selected Odor-Active Compounds in Sesame
Oils with Sesame Seeds Roasted at 160, 180, 200, and 220 C for 20 min (88, 89).
FD Factor in Roasted Content (ppb)

Compound Sesame Seed 160 C 180 C 200 C 220 C
2-Furfurylthiol 20480 16 51 105 158

Guaiacol 10240 147 365 537 718
Acetylpyrazine 1280 57 169 201 197
2-Ethyl-3,5-dimethylpyrazine 1280 73 161 276 287
2,3-Diethyl-5-methylpyrazine 640 13 29 57 50
trans,trans-2,4-Decadienal 640 121 169 287 333
2-Ethyl-5-methylpyrazine 320 96 234 301 301
water. The amount of guaiacol increased from 147 ppb in roasted oil processed at
160 C for 30 min to 718 ppb in the oil processed at 200 C for 30 min (Table 12).
The extremely high concentration of guaiacol in the high-temperature roasting
sample certainly contributes to its smoky and overburnt sensory quality.
REFERENCES
1. H. W. S. Chan, in H. W. S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic Press,

London, 1987, pp. 116.
2. E. N. Frankel, Lipid Oxidation, The Oily Press, Dundee, Scotland, 1998, pp. 1322.
3. R. T. Holman and O. C. Elmer, J. Amer. Oil Chem. Soc., 24, 127129 (1947).
4. E. N. Frankel, Lipid Oxidation, The Oily Press, Dundee, Scotland, 1998, pp. 2341.
5. H. W. S. Chan, in H. W. S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic Press,
London, 1987, pp. 1750.
6. C.-T. Ho and Q. Chen, in C.-T. Ho and T. G. Hartman, eds., Lipids in Food Flavors,
American Chemical Society, Washington, D. C., 1994, pp. 214.
7. E. N. Frankel, Progr. Lipid Res., 22, 133 (1982).
8. M. T. Morales, J. J. Rios and R. Aparicio. J. Agric. Food Chem., 45, 26662673 (1997).
9. C.-T. Ho, Y. Zhang, H. Shi, Huang, and J. Tang. Food Rev. Internat., 5, 253287 (1989).
10. D. B. Josephson and J. Glinka, in T. H. Parliaent, R. J. McGorrin and C.-T. Ho, eds.,
Thermal Generation of Aromas, American Chemical Society, Washington, D. C., 1989,
pp. 242.
11. C.-T. Ho, J. T. Carlin, T. C. Huang, l.S. Hwang, and L. B. Hau, in M. Martens, G. A. Dalen,
and H. Russwurm, Jr., eds., Flavour Science and Technology, Wiley Chichester, England,
1987, pp. 3542.
12. P. Schieberle and W. Grosch, Z. Lebensm. Unters. Forsch., 192, 130 (1991).
13. H. Guth and W. Grosch, Lebensm.-Wiss. U. Technol., 23, 59 (1990).
14. E. N. Frankel, W. E. Neff, and E. Selke, Lipids 16, 279285 (1981).
15. C.-T. Ho, M. S. Smagula, and S. S. Chang, J. Amer. Oil Chem. Soc., 55, 233237 (1978).
16. D. G. Bradley and D. B. Min, Crit. Rev. Food Sci. Nutr., 31, 211236 (1992).
17. D. B. Min and H. O. Lee, in R. Teranishi, E. L. Wick, and I. Hornstein, eds., Flavor
Chemistry: Thirty Years of Progress, Kluwer Academic/Plenum Publishers, New York,
1999, pp. 175187.
18. L. L. Fan, J. Y. Tang, and A. Wohlman, J. Amer. Oil Chem. Soc., 60, 1115 (1983).
19. J. M. Snyder, E. N. Frankel, and E. Selke, J. Amer. Oil Chem. Soc., 62, 16751679 (1985).
20. H. Maarse and C. A. Visscher, Volatile Compounds in Food. Qualitative Data Supplement,
TNO, Zeist, The Netherlands, 1988.
21. T. H. Parliment and R. J. McGorrin, in S. J. Risch and C.-T. Ho, eds., Flavor Chemistry:
Industrial and Academic Research, American Chemical Society, Washington, D. C., 2000,
pp. 4471.
22. R. J. McGorrin, in A. H. Spanier, F. Shahidi, T. H. Parliment, C. Mussinan, C.-T. Ho, and
E. T. Contis, eds., Food Flavors and Chemistry: Advances of The New Millennium, Royal
Society of Chemistry, Cambridge, U.K., pp. 6784.
REFERENCES 409
23. D. A. Forss, W. Stark, and G. Urbach, J. Dairy Res., 34, 131 (1967).
24. G. Urbach, W. Stark, and D. A. Forss, J. Dairy Res. 39, 35 (1972).
25. W. Stark, G. Urbach, and J. S. Hamilton, J. Dairy Res., 40, 39 (1973).
28. T. J. Siek, J. A. Albin, L. A. Sather, and R. C. Lindsay, J. Food Sci., 34, 265 (1976).
29. S. Widder, A. Sen, and W. Grosch, Z. Lebensm. Unters. Forsch., 193, 32 (1991).
30. P. Schieberle, K. Gassenmeier, H. Guth, A. Sen, and W. Grosch, Lebensm. Wissen.
u.-Technol., 26, 347356 (1993).
31. J. Budin, Ph.D. Thesis, University of Minnesota, Minneapolis, Minnesota, 1998.
32. A. H. M. van Elzakker and H. J. van Zutphen, Z. Lebensm. Unters. Forsch., 115, 222
(1961).
33. G. P. Rizzi, J. Agric. Food Chem., 15, 549551 (1967).
34. J. T. Carlin, M. S. Thesis, Rutgers University, New Brunswick, New Jersey, 1980.
35. J. T. Carlin, K. N. Lee, O. A. L. Hsieh, L. S. Hwang, C.-T. Ho, and S. S. Chang, Proc. 36th
P. M. C. A. Production Conf., 1982, pp. 95104.
36. C.-T. Ho, Q. Z. Jin, K. N. Lee, J. T. Carlin, and S. S. Chang, J. Food Sci., 48, 15701571
(1983).
37. J. T. Carlin, K. N. Lee, O. A. L. Hsieh, L. S. Hwang, C.-T. Ho, and S. S. Chang, J. Amer. Oil
Chem. Soc., 63, 10311036 (1986).
38. T. Kosuge and H. Kamiya, Nature, 193, 776 (1962).
39. K. Ostovar, Ph.D. Thesis, The Pennsylvania State University, University Park, PA, 1971.
40. O. G. Vitzum and P. Werkhoff, Z. Lebensm.-Unters. Forsch., 156, 300307 (1974).
41. G. Ohloff and I. Flament, Forschr. Chem. Org. Naturst., 36, 231283 (1979).
42. J. A. Maga, in T. E. Furia and N. Bellanca, eds., Fenarolis Handbook of Flavor
Ingredients, Vol. 1, 2nd ed., CRC Press Inc., Cleveland, Ohio, 1975, pp. 228238.
43. K. Watanabe and Y. Sato, Agric. Biol. Chem., 33, 242249 (1969).
48. K. Watanabe and Y. Sato, J. Zootech. Sci., 42, 393 (1971).
49. L. S. Hwang and C. W. Chen, C.-T. Ho and T. G. Hartman, eds., Lipids in Food Flavors,
American Chemical Society, Washington, D.C., 1994, pp. 244255.
50. T. H. Smouse, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of Fats and Oils,
American Oil Chemists Society, Champaign, Illinois, 1985, pp. 85116.
51. S. S. Chang, G. H. Shen, J. Tang, Q. Z. Jin, H. Shi, J. T. Carlin, and C.-T. Ho, J. Amer. Oil
Chem. Soc., 60, 553557 (1983).
52. T. H. Smouse and S. S. Chang, J. Amer. Oil Chem. Soc., 44, 509513 (1967).
53. B. S. Bistry and D. B. Min, in G. Charalambous, ed., Off-flavors in Foods and Beverages,
Elsevier Science Publishers, Amsterdam, The Netherlands, 1992, pp. 171209.
54. F. Ullrich and W. Grosch, Fat Sci. Technol., 91, 332 (1988).
55. H. Guth and W. Grosch, Fat Sci. Technol., 91, 225230 (1989).
56. H. Guth and W. Grosch. Fat Sci. Technol., 93, 249 (1991).
57. J. W. Kao, X. Wu, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 75, 831835
(1998).
58. Z. J. Hawrysh, in F. Shahidi, ed., Canola and Rapeseed: Production, Chemistry, Nutrition
and Processing Technology, Van Nostrand Reinhold, New York, 1990, pp. 99122.
59. S. K. Raghavan, D. R. Connell, and A. Khayat, in C-T. Ho, and T. G. Hartman, eds., Lipids
in Food Flavors. ACS Symposium Series 558. American Chemical Society, Washington,
D.C., 1994, pp. 292300.
60. A. K. Kiritsakis, J. Amer. Oil Chem. Soc., 75, 673681 (1998).
61. T. M. Morales, V. M. Alonso, J. J. Rios, and R. Aparicio, J. Agric. Food Chem., 43, 2925
2931 (1995).
62. F. Angerosa, C. Basti, and R. Vito, J. Agric. Food Chem., 46, 836839 (1998).
63. T. M. Morales and R. Aparicio, J. Amer. Oil Chem. Soc., 76, 295300 (1999).
64. J. Sanchez and J. L. Harwood, Eur. J. Lipid Sci. Technol., 104, 602613 (2002).
65. F. Angerosa, Eur. J. Lipid Sci. Technol., 104, 639660 (2002).
66. A. Ranalli and G. D. Mattia, J. Amer. Oil Chem. Soc., 74, 11051113 (1997).
67. C. F. Lin, in C.-T. Ho and T. G. Hartman, eds., Lipids in Food Flavors, American Chemical
68. K. Fujimoto, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of Lipid Foods,
69. F. Shahidi and K. R. Cadwallader, in Flavor and Chemistry of Seafoods, American
Chemical Society, Washington, D.C. 1997, pp. 18.
70. N. Ikeda and K. J. Fikuzumi, Jpn. Oil Chem. Soc., 27, 2126 (1978).
71. S. Y. Cho, K. J. Miyashita, T. Miyazawa, K. Fujimoto, and T. Kaneda, J. Amer. Oil Chem.
Soc., 64, 876879 (1987).
72. F. Shahidi, U. N. Wanasundara, Y. He, and V. K. S. Shukla, in C.-T. Ho and T. G. Hartman,
eds., Lipids in Food Flavors, American Chemical Society, Washington, D.C., 1994, pp.
186197.
73. R. Yamauchi, T. Yamada, K. Kato, and Y. Uevio, Agric. Biol. Chem., 47, 28972902 (1983).
74. M. Van Rollin and R. C. Murphy, J. Lipids Res., 25, 507517 (1984).
75. A. C. Noble and W. W. Nawar, J. Amer. Oil Chem. Soc., 48, 800 (1971).
76. P. W. Meijboom and J. B. A. Stroink, J. Amer. Oil Chem. Soc., 49, 555 (1972).
77. P. J. Ke, R. G. Ackman, and B. A. Linke, J. Amer. Oil Chem. Soc., 52, 349 (1975).
78. T. C. Y. Hsieh, C. F. Lin, J. B. Crowther, and J. P. Bimbo, Tropical and Subtropical
Fisheries Technological Society of the Americas, 1988, pp. 154161.
79. T. C. Y. Hsieh, S. S. Williams, W. Vejaphan, and S. P. Meyers, J. Amer. Oil Chem. Soc., 66,
114117 (1989).
80. C. F. Lin, T. C. Y. Hsieh, J. B. Crowther, and J. P. Bimbo, J. Food Sci., 55, 16691672
(1990).
81. C. F. Lin. in C.-T. Ho and T. G. Hartman, eds., Lipids in Food Flavors, American Chemical
82. K. R. Cadwallader and F. Shahidi, in F. Shahidi and J. W. Finley, eds., Omega-3 Fatty
Acids: Chemistry, Nutrition and Health Effects, ACS Symposium Series 788, American
Chemical Society, Washington, D.C., 2001, pp. 221235.
REFERENCES 411
83. J. K. Farkas, J. D. Floros, D. S. Lineback, and B. A. Watkins, J. Food Sci., 62, 505507
(1997).
84. J. Ogata, Y. Hagiwara, H. Hagiwara, and T. Shibamoto, J. Amer. Oil Chem. Soc., 73, 653
656 (1996).
85. R. J. Hamilton, C. Kalu, G. P. McNeill, F. B. Padley, and J. H. Pierce, J. Amer. Oil Chem.
Soc., 75, 813822 (1998).
86. K. Kikugawa, M. Arai, and T. Kurechi, J. Amer. Oil Chem. Soc., 60, 15281533 (1983).
87. G. C. Yen, J. Sci. Food Agric., 50, 563570 (1967).
88. G. C. Yen, S. L. Shyu, and T. C. Lin, Food Sci. (ROC), 13, 198211 (1986).
89. M. Shimoda, H. Shiratsuchi, Y. Nakada, Y. Wu, and Y. Osajima, J. Agric. Food Chem., 44,
39093912 (1996).
90. M. Shimoda, Y. Nakada, M. Nakashima, and Y. Osajima, J. Agric. Food Chem., 45, 3193
3196 (1997).
91. N. Ryu, S. M. Kim, J. Xi, and C.-T. Ho, in F. Shahidi, and C.-T. Ho, eds., Flavor Chemistry
of Ethnic Foods, Kluwer Academic/Plenum Publishers, New York, 1999, pp. 229237.
92. P. Schieberle, in P. Schreier, and P. Winterhalter, eds., Progress in Flavor Precursor
Studies, Allured Publishing, Carol Stream, Illinois, 1999, pp. 343360.
10
Flavor and
Sensory Aspects
Linda J. Malcolmson
Canadian International Grains Institute
Winnipeg, Manitoba, Canada
1. INTRODUCTION
Sensory evaluation is a scientific discipline that uses humans to measure the accept-
ability and sensory properties of food and other materials. Sensory properties
important in food products include attributes of appearance, odor, taste, and texture.
The use of humans as measuring devices is necessary because only humans can
define what is acceptable, and in many cases, no instrumental or chemical method
can adequately measure or replicate the human response. For this reason, sensory
evaluation is a vital component in any quality assessment program. In such pro-
grams, sensory evaluation can be used to monitor product quality; determine effects
of alternative processing, ingredients, or formulations; evaluate packaging; and
determine product shelf life.
2. SENSORY METHODS
Sensory methods are often criticized as being subjective techniques. Part of the pro-
blem lies with the failure to acknowledge that two distinct types of sensory tests
413
414 FLAVOR AND SENSORY ASPECTS
exist. Product-orientated tests involve the use of selected and trained panelists
under controlled testing conditions to evaluate the attributes of a product. These
products are objective because they meet the criteria of objectivity, namely freedom
from personal bias and repeatability. Consumer-oriented tests involve the use of
consumer panelists to determine degree of liking for a product or product accept-
ability. These tests, by their very nature, are subjective because it is the subjective
information (personal likes and dislikes) that is of interest. Thus, if sensory tests
are done under controlled testing conditions, using trained panelists and appropriate
sensory methodologies, the procedures are objective.
The sensory method chosen for a test is dependent on the type of information
required. The appropriate method can only be selected after the objectives of the
test are clearly defined. As illustrated in Table 1, there are a substantial number
of test methods available and it is important to be familiar with them in order to
apply the correct method to achieve the correct results. Familiarity with the test
methods also reduces the likelihood of relying on a single method. Difference
and descriptive tests are the main tests used to evaluate fat and oil quality. Both tests
require the use of trained panelists, but the level of training is considerably more for
descriptive testing. In difference testing, panelists only determine if there is a dif-
ference between samples. The degree of difference is not determined. In contrast,
descriptive tests allow panelists to rate the intensities of several attributes resulting
TABLE 1. Summary of Product-Oriented and Consumer-Oriented Sensory Tests.
Test Type Test Procedure and Objective Test Method
Product-oriented Thresholdmeasures the level at which Method of limits

a sensory characteristic can be detected Average error
Frequency
Differencemeasures difference or Paired comparison
similarity between products or attributes Duo-trio
of products Dual-standard
Triangle
Scalingmeasures intensity of a Ranking
product attribute Category
Magnitude estimation
Durationmeasures the duration for Time-intensity
which an attribute is detected
Descriptive analysismeasures the Flavor profile
intensity of several attributes Texture profile
associated with the product Quantitative descriptive
analysis
Deviation from reference
Consumer-oriented Acceptancemeasures acceptance
of product
Preferencemeasures which Paired comparison
product is preferred
Hedonicmeasures degree of Ranking
liking/disliking for a product Category
FACTORS AFFECTING SENSORY MEASUREMENTS 415
in a full description or profile of the samples. A more thorough discussion of test

methodologies is available elsewhere (1, 2).
3. FACTORS AFFECTING SENSORY MEASUREMENTS
Standard practices have been developed for conducting sensory panels in order to
minimize psychological errors and physical testing conditions that can influence
human judgment. By controlling these factors, reliable sensory results can be
achieved. The following sections outline the most common psychological errors
and how they can be minimized.
3.1. Expectation Error

Expectation error occurs when panelists are given information about the samples
prior to the test. This information can influence their judgment because panelists
usually find what they expect to find. To control for this error, panelists are only
given enough information for them to perform the task, samples are coded to con-
ceal identity, the panel is held in a separate room away from the preparation area,
and the person conducting the test does not serve as a panelist.
3.2. Suggestion Error

This error occurs when one panelists reaction to a sample causes other panel mem-
bers to change their judgments. To minimize this error, individual booths are used
and no discussion is permitted during testing.
3.3. Stimulus Error

Stimulus error occurs when panelists are influenced by irrelevant sample differ-
ences, such as differences in size, volume, and color of the samples. To minimize
this error, the samples must be prepared and served under identical conditions.
Unwanted color differences should be masked using colored lights.
3.4. Positional Bias or Order Effect

The order in which samples are evaluated can influence a panelists judgment. The
first sample is often scored higher regardless of the product being tested. To mini-
mize this effect, serve samples to panelists in a randomized order: this way posi-
tional bias is balanced for the panel as a whole. A warm-up sample can also be used
that calibrates the panel before beginning the test.
3.5. Central Tendency Error

This error is characterized by panelists avoiding the end points on a scale when rat-
ing samples. Use of reference samples and training can minimize this error.
3.6. Contrast and Convergence Errors

Contrast error occurs when two samples are very different from each other. Panel-
ists tend to exaggerate the difference in their scores. Convergence error is the oppo-
site effect whereby a sample may mask small differences between two samples
causing their scores to converge. To minimize both of these errors, sample and serv-
ing orders must be randomized among the panelists.
4. EXPERIMENTAL DESIGN AND STATISTICAL ANALYSIS
The experimental design that is selected and the statistical analysis that is applied to
the data will significantly influence the outcome of a test. In sensory testing, the
experimental design dictates the serving order of the samples. Several experimental
designs are available to choose from, including completely randomized, balanced
complete block, and balanced incomplete block designs. Key elements to good
experimental design include randomization, blocking, and replication.
Randomization is done to minimize the effects of uncontrolled sources of varia-
tion or error and to eliminate bias. It involves ordering sample treatments in such
a way that each treatment has an equal chance of being selected. In sensory testing,
the order of sample presentation to each panelist is randomized.
Blocking is used to increase the power of an experiment by removing known
sources of variability from the estimate of error. Blocks can include panelists, repli-
cations, treatment, or anything else that is a known source of variation. By grouping
experimental units into blocks, a more accurate measure of pure or experimental
error is obtained.
Replication involves repeating the experiment under identical conditions. It
improves the reliability and validity of test results and is necessary to provide an
estimate of experimental error. The number of replications completed is determined
by time, cost, and sample constraints. However, the more replications completed,
the better the estimate of experimental error and the more reliable the results.
Statistics are used in sensory testing to determine whether the responses from
panelists are sufficiently similar or represent a random occurrence. Knowing the
degree of similarity enables one to draw conclusions about the samples being tested
with some measure of confidence in the context of that population of subjects and in
the case of consumer-oriented tests to the population in general. Thus, it is impor-
tant that appropriate statistical methods be applied to sensory data to satisfy the test
objectives. For a more thorough discussion of the appropriate statistical methods for
each of the various sensory tests, refer to Stone and Sidel (1) and ASTM (2).
5. SENSORY TESTING FACILITY
An important requirement of conducting sensory tests is the availability of a facility

specifically designed for sensory testing. This is necessary to ensure that tests are
SENSORY TESTING FACILITY 417
run efficiently and conducted under controlled conditions with minimum distrac-
tions. Without proper controls, reliable results may not be achieved. Permanent test-
ing facilities provide the best testing environment. If this is not feasible, it is
possible to adapt existing facilities for sensory testing. The types of tests to be con-
ducted, the amount and frequency of testing, and the space and resources available
will be deciding factors in the facilities that are developed for sensory testing. In
general, space is needed for preparation of samples, testing, and training. A more
thorough discussion regarding the design of sensory testing facilities including
illustrations of possible layouts can be found elsewhere (1, 3).
The preparation area should provide sufficient counter space for preparation of
samples and layout of sample trays. The area should provide refrigerated and frozen
storage, electrical ranges and microwave ovens for heating, and sinks and dish-
washers for cleaning purposes. Installation of distilled or filtered water should be
considered for panelists rinsing water and preparation of solutions if tap water
imparts a flavor and odor. Ventilation hoods with exhaust fans should be installed
to remove odors in order to prevent the odors from spreading to the booth area.
The testing facility must provide panelists a setting that minimizes factors that
can influence their judgments. For this reason, individual booths are constructed so
that a panelist can evaluate the samples without being influenced by other panelists.
The booth area should be separate from the preparation area. Ideally, this room
should be adjacent to the preparation area to permit samples to be passed through
the wall of the preparation area to the panelist. Eight booths should be adequate to
run most sensory programs. Each booth should be equipped with a counter, chair, a
pass-through opening to the preparation area, individual lighting, and an electrical
outlet for warming trays. Individual sinks for expectoration can also be provided
but are problematic due to sanitation, noise, and odor problems. The use of covered
disposable containers for expectoration is a more acceptable alternative. Some
booths are computerized so that panelists utilize a computer screen, keyboard,
and mouse rather than a paper ballot to record their judgments.
A booth width of 6780 cm and a counter depth of 45 60 cm is recommended.
Counter height should be the same as the counter height on the other side of the
pass-through wall to permit ease of passing trays from one side to the other. Stan-
dard counter height is 90 cm. Partitions between booths should be at least 90 cm
high and should extend 45 cm beyond the edge of the countertop to provide full
privacy for the panelist (3). Walls and booths should be constructed of opaque,
nonreflecting material that is neutral in color. Countertops and flooring should be
odor-free and easily cleaned. Construction of a soffit above the booth counter is
recommended to house lighting fixtures and ventilation ducts (1). Adequate space
must be provided so that panelists can enter and exit the room without disturbing
others. Pass-throughs from the preparation area should be large enough to accom-
modate tray passage. The opening can be fitted with a hinged, sliding, flip-up,
or bread-box type door. However, more counter space will be required if hinged,
flip-up, or bread-box type openings are used.
Each booth should have its own lighting to ensure uniform light distribution. Use
of both incandescent and fluorescent lighting offers greater testing flexibility.
Fluorescent lights do not generate heat and are recommended when color evalua-
tion is critical because bulbs can be selected that more closely match natural day-
light. In situations when color differences between samples needs to be masked,
colored incandescent bulbs or colored filters are used. Red lights are recommended
for masking color differences in oil samples.
Room temperature control and adequate ventilation are the most critical factors
in designing the panel room. In order to maintain a comfortable temperature in the
room, air-conditioning should be installed. A slight positive air pressure should be
maintained in the booth area to prevent infiltration of external odors. Individual
exhaust ventilation ducts should be located in each booth. Air turnover in the
room should occur at least every 30 seconds (1). Recirculated and makeup air
should pass through activated carbon filters to remove odors.
If an area specifically designed for sensory testing is not available, a temporary
area such as a boardroom can be set up provided that noise, distractions, and odors
are minimized. Portable partitions made of heavy cardboard, wood, or metal can be
used to provide individual booths for panelists.
The training area should also be adjacent to the preparation area and should
allow for group discussion and testing of products. A comfortable, well-lit room
with good ventilation and a table for 812 panelists is required. A whiteboard or
flip chart should be provided to facilitate discussion.
Information pertaining to the design of a facility for evaluating room odor has
been published by Mounts and Warner (4). The most important feature of this facility
is the air lock room that a panelist must pass through before entering the actual odor
room. This helps prevent significant loss of volatile odor components from the test
room. Control of the amount of airflow into and out of the rooms is also required.
6. SAMPLE PREPARATION AND PRESENTATION
Samples should be prepared and presented to panelists under identical conditions.

It is recommended that at least 30 g of solid food or 15 mL of liquid food be
served (5). It may be necessary to serve some foods with a carrier or to use dilu-
tions. For example, crackers or white bread may be used as a carrier for the evalua-
tion of margarine. However, it is generally better to avoid the use of a carrier where
possible, because they have flavor and textural characteristics of their own that can
interfere with the panelists evaluation of the intended product. Through training,
panelists can usually overcome any aversions they might have to sampling a fat-
based product on its own. If crude oil is to be evaluated, dilution may be required
but the effect of the dilution medium needs to be considered. Warner et al. (6)
found that good quality freshly deodorized soybean oil could be used as a diluent
at a ratio of 95:5 in the evaluation of crude soybean oils. Stone and Hammond (7)
reported on a method for evaluating oils in emulsions stabilized with gum acacia.
The method allows for smaller quantities of oil to be evaluated because the total
volume is increased through dilution. This can be an issue when evaluating experi-
mental lines in breeding programs. The emulsion also clears the palate sooner and
SAMPLE PREPARATION AND PRESENTATION 419
more thoroughly than a pure oil, thereby reducing sample carryover from one oil
sample to the next.
Whenever human subjects are used, every effort must be made to ensure that the
panelists are not exposed to any risk associated with the samples. If there is some
concern, precautions, such as expectorating the samples or only evaluating odor,
should be implemented.
Generally, samples should be served at the temperature at which the food is
usually consumed. Soft margarines are served at 4 5 C, whereas hard margarines
can be served at 4 5 C or 2224 C. It is recommended that salad oils be evaluated
at a temperature of 50 C (8, 9) because this temperature brings out their character-
istic odor and flavor. A circulating waterbath can be used for heating and keeping
sample containers warm (9). If a waterbath is not available, an electric warming
tray equipped with a pan containing water can be used. Alternatively, AOCS (8)
recommends heating aluminum blocks containing the sample containers to the
desired temperature.
Glass containers are recommended in the evaluation of salad oils because glass
can be cleaned more thoroughly and therefore reused. Small glass beakers or jars
work the best, but test-tubes can also be used if necessary. The containers should
only be slightly filled with oil and should be capped to permit volatiles to build
in the headspace. AOCS (8) recommends using 10 mL of oil in a 50 mL glass
beaker covered with a watch glass. Disposable plastic or Styrofoam containers
can be used provided the samples are not being heated and they do not impart
any flavors or odors of their own. If samples are only being assessed for their
odor properties, 10 g of oil can be placed in a 125 mL covered glass jar containing
30 g of glass beads (10). The glass beads are used to enhance the release of odor
volatiles when the sample is swirled prior to sniffing. Disposable filter paper sticks
used in the perfume industry can also be employed for assessing the odor charac-
teristics of an oil by dipping the stick into the oil and smelling the oil absorbed onto
the filter paper.
Margarine can be served by preloading on to plastic serving spoons for flavor
evaluations. If margarines are to be evaluated for their spreadability properties,
they should be served in a small plastic or paper cup (2530 mL) with a plastic
knife and a slice of white bread. Bland crackers can also be used, although these
are less desirable than bread in evaluating spreadability due to their limited size
and ease of breaking.
Samples should be coded in such a way that they impart no information to the
panelists that might influence their judgment. It is therefore recommended that
samples be coded with three digit random numbers. Samples should be served in
a random order to each panelist to minimize positional bias or order effect.
The number of samples that can be evaluated in a session is dependent on the
amount of training and experience of the panelists and on the nature of the product
being evaluated. In the case of fats and oils, sensory fatigue can result from the
evaluation of too many samples. This is due, in part, to carryover effects caused
by the tendency of fats and oils to coat the mouth, making it difficult to cleanse
the palate. Warm water (38 40 C) is necessary to clear the mouth when fats and
oils are being evaluated. Unsalted crackers or white bread can also be used. In some
cases, it may be necessary to specify a waiting time between samples. As a result
of the nature of fats and oils, most experienced panelists expectorate samples rather
than swallow them. If samples are only being evaluated for odor, it is recommended
that panelists sniff a sample of water between samples or wait between samples
before proceeding to the next sample (2030 seconds is usually adequate). Panelists
need to mutually agree on how the samples should be evaluated and then should
consistently use this technique.
7. REFERENCE SAMPLES
The use of reference samples or standards is critical in both training and calibration
of panelists during testing. Reference samples are used during training to aid panel-
ists in understanding terminology used to describe product quality. They are also
used to anchor end-points and mid-points on attribute rating scales. The presence
of reference samples during testing permits calibration of panelists before evalua-
tion of the coded samples, thereby ensuring panelists score the coded samples more
consistently. This is especially critical when sensory testing takes place over a per-
iod of time, such as during storage testing. Reference samples can be either internal
or external references or standards. Internal references are selected from the sam-
ples that are to be tested. For example, fresh and oxidized oils are used to anchor the
bland and strong end-points on an oil intensity scale, respectively. An external
reference is a chemical or alternative food product that is selected to represent a
specific attribute. An example of an external reference would be the use of a bland
oil sample spiked with an artificial butter flavoring to anchor the high-intensity end-
point on a buttery intensity scale. Spiking good quality oil with oxidized oil using a
series of different ratios can be done to provide a range of oxidized samples
that can be used during training. Regardless of the reference used, it is important
that the product selected clearly represents the attribute and that it is stable and
will not change over the course of the sensory test (11). Terms, definitions, and re-
ference samples suitable for evaluating fats, oils, and fat-containing foods are pro-
vided in Table 2.
TABLE 2. Flavor and Odor Terms, Definitions, and Reference Standards Used
in the Evaluation of Fats, Oils, and Oil-Containing Foods.
Term Definition Reference Standard

Acrid Aromatic characteristic of burnt oil, Canola oil heated to 240 C, acrolein
pungent, sharp.
Beany Aromatic characteristic of soybeans. Cooked soybeans, not raw or green
Bland No aromatics or flavors perceived
Burnt Aromatic characteristic of overheating. Crude corn oil (odor only)
Buttery Aromatic characteristic of fresh, unsalted butter. Unsalted fresh butter diluted 1:99 in fresh oil
Cardboard Aromatic characteristic of slightly oxidized oil or Water that has had cardboard or paper
oil containing foods. soaked in it for 30 min and then filtered to
remove paper
REFERENCE SAMPLES 421
TABLE 2. (Continued )
Term Definition Reference Standard

Corny Aromatic characteristic of steeped corn and Crude corn oil diluted 5:95 in good quality
corn oil. deodorized corn oil
Fishy Aromatic characteristic of oxidized Cod liver oil or canola oil heated to 190 C
linolenate-containing oils such as linseed, for 30 min
canola, and soybean.
Fried Food Aromatic characteristic of food fried in hot oil. Deep fried potatoes
Fruity Aromatic characteristic of ripe non-citrus fruit. Olive oil
Grassy Aromatic characteristic of freshly mown Fresh cut green grass, hexanal
green grass.
Green Aromatic characteristic of raw, unripe Raw soybeans
vegetables.
Hay Aromatic characteristic of dried grass or hay. Crude soybean oil from heat-processed
beans diluted 5:95 in fresh oil
Hull-like Aromatic characteristic of grain hulls. Peanut or sunflower hulls
Hydrogenated Aromatics characteristic of typical hydrogenated Hydrogenated soybean oil with a iodine
oils comprised of attributes of paraffin, fruity, value of <100
and flowery.
Melon Aromatic characteristic of ripe melons. 0.002 ppm cis-6-nonenal in fresh oil
Metallic Aromatic characteristic of metal coins. 0.01% ferrous sulfate in distilled water
Musty Aromatic characteristic of damp basement. Geosmin (odor only)
Nutty Aromatic characteristic of fresh nuts. Freshly processed peanut oil
Overheated Aromatic characteristic of oil overheated Crude corn oil (odor only)
during processing
Oxidized General term used to describe aromatics
characteristic of oxidized oils from slight to
intense i.e. from buttery to painty. Not
recommended as a descriptor.
Painty Aromatic characteristic of oxidized Linseed oil
linolenate-containing oils such as linseed,
canola, and soybean with a peroxide value
>10.
Pine Aromatic characteristic of pine needles. Nondeodorized sunflower oil diluted 5:95
in fresh sunflower oil
Rancid Aromatic characteristic of oxidized oil. Safflower or sunflower oil with a peroxide
value >5
Raw Aromatic characteristic of unprocessed Whole wheat flour, raw potato
grain or vegetables.
Reverted General term used to describe an oxidative
process known as reversion. Not
recommended as a descriptor.
Rubbery Aromatic characteristic of old rubber. Rubber stoppers used for chemical
glassware
Smokey Aromatic characteristic of overheated oil. Oil heated to 240 C
Sour Basic taste characteristic of acids. 0.08% citric acid in water solution
Stale Aromatics characteristic of old, no longer Potato chips aged 2 weeks at 25 C
fresh product with loss of original flavors;
usually occurs in initial stages of oxidation.
Sulfur Aromatic characteristic of sulfur compounds. Nondeodorized canola oil diluted 5:95 in
fresh oil
Waxy Aromatic characteristic of melted paraffin. Paraffin oil
Weedy Aromatic characteristic of green non-grass weeds. Fresh cut green weeds
Woody Aromatic characteristic of dry wood. Sawdust, wood chips, popsicle sticks, or
peanut oil heated to 190 C for 30 min
Adapted from Warner (12).

8. SELECTION AND TRAINING OF PANELISTS
Panelists for trained panels can be recruited from the organization where the testing
is to be conducted. In many cases, approval to use panelists as human subjects
requires prior approval by an internal committee, so the agency can be in compli-
ance with national, regional, or organizational requirements for use of human sub-
jects. Provided there is management support, most employees are willing to
participate if they feel their contribution is important. Potential panelists should
be asked to complete a simple questionnaire to determine their likes and dislikes
of foods, any food restrictions, allergies and medications, and time available for
panels. This information is useful in screening individuals in terms of availability
and general health. Candidates should be in good health and not prone to frequent
colds. Certain medications may alter a panelists sensitivity to taste and odor com-
pounds. Age, sex, and smoking habits may be other factors to consider in selecting
panelists, although they are not critical.
The next step is to screen panelists for their sensory acuity. This is done by
having panelists complete tests designed to identify their ability to recognize basic
tastes, common odors, and textural characteristics. More information on these tests
is provided in guidelines published by ASTM (13).
The final step in the selection of panelists is based on their ability to detect dif-
ferences in the products that they will be evaluating and to be consistent in their
judgments. This is often done using a series of triangle tests. For example, samples
of vegetable oils of known differences are presented to potential panelists. Pane-
lists are asked to identify which sample is different to gain a measure of the
panelists ability to discriminate. The test is repeated several times to gain a mea-
sure of the panelists ability to reproduce judgments. A group of 20 25 potential
panelists are generally screened in order to select a panel of 812 members for
training.
The performance of individual panelists and the panel as a whole is improved
significantly through training as the goal is to help panelists make valid, reliable,
and objective judgments free of personal preferences. Training exercises are
designed to teach panelists about the terminology used to describe product quality,
the scoring system, and the task they are being asked to perform. They are
also exposed to the range of samples they are likely to encounter in the test
sessions and to reference samples. Through discussion, panelists learn to develop
standardized evaluations such that panelists responses are consistent and agree
with each other. Depending on the needs of the group, training sessions are
designed to last 30 60 minutes per day over a period of 12 weeks. The exact
length of time depends on the nature of the products to be tested, the number of
attributes to be measured, and the past experience of the panelists. Training is
complete when panelists are in agreement on the attributes to be measured and
in the placement of the reference samples on the scale, are comfortable with
the procedure, are consistent in their judgments, and are in agreement with each
other.
SENSORY EVALUATION OF OILS 423
9. MONITORING AND MOTIVATION OF PANELISTS
Panelist performance as a group and individually should be monitored throughout

the testing period to ensure the panel is performing and does not require recalibra-
tion. To determine this, a set of samples of known differences is evaluated by the
panel on several occasions.
It is important to maintain the interest and motivation of panelists throughout
training and testing because panelists who are highly motivated perform better.
Feedback about their performance, especially during training, is critical to keep
panelists motivated. Rewards of food or beverage at the conclusion of the session
can also be used. At the conclusion of the test, panelists should be provided with
information about the test and be assured that their contribution was appreciated.
This is important as it will assist in motivating panelists to participate in future
panels.
10. SENSORY EVALUATION OF OILS
Salad oils are evaluated for their initial odor and flavor and for their stability during
storage. During odor evaluations, the sample container should always be covered.
Panelists are instructed to gently swirl the container, remove the cover quickly,
and take three short sniffs before placing the lid back on the container. Odors dis-
appear from the headspace quickly, so it is important for panelists to make their
judgment as rapidly as possible. If retesting of the oil is required, panelists should
allow the volatiles to concentrate in the headspace of the covered sample. Generally
35 minutes is adequate. The exact waiting time is dependent on the amount of
headspace and the aperture of the container. Evaluation of the flavor characteristics
of oil requires taking 510 mL of oil into the mouth, pulling air through the oil,
and exhaling through the nose. This procedure enhances the flow of volatiles to
the retronasal area responsible for detection of odors through the oral cavity.
Although some researchers suggest that this method is more sensitive than the
nasal method of assessing volatiles, by carefully controlling testing conditions,
the nasal method can be successfully used to assess oil quality. Indeed, panelists
can generally evaluate more samples using this technique because there is no
buildup of oil residue as there is when oil is taken into the mouth.
The American Oil Chemists Society (AOCS) and the American Society for
Testing Material (ASTM) have published recommended practices with regard to
the serving containers and serving procedures that should be used when assessing
oil samples (8, 9). The AOCS practice includes two rating scales based on a
10-point scoring system for measuring the overall quality and intensity of liquid
vegetable oils. The intensity scale has a range from 10 bland to 1 extremely
strong (Figure 1). This scale is recommended for rating oils that have a bland
odor and flavor after processing, such as canola, soy, sunflower, safflower, and
cottonseed. Oils such as corn, peanut, and olive, which have a naturally distinct
Directions: Take 5-10 mL of warm oil into the mouth; pull air through the oil and exhale through
the nose. Rate samples for overall flavor intensity on the 10 point scale; identify flavors and rate
as weak (W), moderate (M), or strong (S).
Intensity Overall Intensity Scores
492 716 258 931
10 Bland ____ ____ ____ ____
9 Trace ____ ____ ____ ____
8 Faint ____ ____ ____ ____
7 Slight ____ ____ ____ ____
6 Mild ____ ____ ____ ____
5 Moderate ____ ____ ____ ____
4 Definite ____ ____ ____ ____
3 Strong ____ ____ ____ ____
2 Very Strong ____ ____ ____ ____
1 Extreme ____ ____ ____ ____
Descriptions Intensity
Nutty ____ ____ ____ ____

Buttery ____ ____ ____ ____
Corny ____ ____ ____ ____
Beany ____ ____ ____ ____
Hydrogenated ____ ____ ____ ____
Burnt ____ ____ ____ ____
Weedy ____ ____ ____ ____
Grassy ____ ____ ____ ____
Rubbery ____ ____ ____ ____
Melon ____ ____ ____ ____
Rancid ____ ____ ____ ____
Painty ____ ____ ____ ____
Fishy ____ ____ ____ ____
Other __________ ____ ____ ____ ____
Figure 1. Flavor intensity ballot (8).
yet desirable odor and flavor, should be rated on the quality scale (Figure 2). After
rating the overall flavor of the sample, panelists are instructed to rate the oil
for individual flavor characteristics on a 3-point scale from weak to strong using
a checklist of possible flavor attributes. The practice of having panelists rate the
overall intensity or quality, followed by rating individual attributes, suggests that
a single measurement is not always adequate to profile an oil. In such instances,
it may be more practical and informative to have panelists rate the intensity of
specific attributes using scales that are less restrictive than the 3-point scales
used in the AOCS method. Malcolmson et al. (14) have successfully used
SENSORY EVALUATION OF OILS 425
Directions: Take 5-10 mL of warm oil into the mouth; pull air through the oil and exhale through
the nose. Rate samples for overall flavor quality on the 10 point scale; identify flavors and rate
as weak (W), moderate (M), or strong (S).
Intensity Overall Quality Scores
492 716 258 931
10 Excellent ____ ____ ____ ____
9 Good ____ ____ ____ ____
8 ____ ____ ____ ____
7 Fair ____ ____ ____ ____
6 ____ ____ ____ ____
5 Poor ____ ____ ____ ____
4 ____ ____ ____ ____
3 Very Poor ____ ____ ____ ____
2 ____ ____ ____ ____
1 Bad ____ ____ ____ ____
Descriptions Intensity
Nutty ____ ____ ____ ____

Buttery ____ ____ ____ ____
Corny ____ ____ ____ ____
Beany ____ ____ ____ ____
Hydrogenated ____ ____ ____ ____
Burnt ____ ____ ____ ____
Weedy ____ ____ ____ ____
Grassy ____ ____ ____ ____
Rubbery ____ ____ ____ ____
Melon ____ ____ ____ ____
Rancid ____ ____ ____ ____
Painty ____ ____ ____ ____
Fishy ____ ____ ____ ____
Other __________ ____ ____ ____ ____
Figure 2. Flavor quality ballot (8).
unstructured line scales to rate the intensity of buttery and painty odors of various
canola oils stored under accelerated storage conditions.
Sensory quality of olive oil is currently determined by the European Union regu-
lation (15) or the International Olive Oil Council (IOOC) trade standard (16). Both
official methods used trained panelists but differ in the sensory descriptors and the
scales employed. The EU standard involves rating the oil for olfactory, gustatory,
and tactile attributes on intensity rating scales ranging from 0 (no perception) to 5
(extreme), followed by an overall rating for grade on a 9-point scale from 1 (lowest
quality) to 9 (maximum quality). Using these methods, Aparicio and Morales (17)
proposed a flavor sensory wheel to show relationships among attributes perceived
in olive oils evaluated by Spanish, Italian, Dutch, and British trained panels based
on principal component analyses.
Scales reportedly used by the oil industry combine intensity and quality charac-
teristics in one scale. For example, oil with a weak melon flavor is rated as a 5,
whereas oil with a weak painty flavor is rated as a 4. These scales are not
recommended because information cannot be accurately captured using scales of
this nature.
11. SENSORY EVALUATION OF OIL-CONTAINING FOODS
Evaluation of oil-containing foods, such as salad dressings, mayonnaise, and mar-

garines, requires highly trained panelists. During training, panelists agree on the
attributes important in a good quality product. Attributes of appearance, taste,
flavor, and texture need to be considered as well as physical attributes such as pour-
ability and spreadability. Definitions, references, and evaluation procedures must
also be established during training. The use of dilutions or carriers may be required.
However, trained panelists generally prefer to evaluate these products without the
use of carriers or dilutions as they find the task easier without the presence of inter-
fering ingredients. Attributes are rated for their intensities using either category/
structured scales or unstructured line scales. Using unstructured line scales, Kok
et al. (18) measured the spreadability, graininess, and waxiness of margarines
served at room temperature while Rousseau and Marangoni (19) measured cold
spreadability, butter flavor, off-flavors, and texture (oily, lingering coating in mouth
to clean, rapid melt-in-mouth). Structured scales were used by Jacobsen et al. (20)
to evaluate the aroma, flavor, and textural properties of mayonnaise.
12. SENSORY EVALUATION OF FRYING OILS/ROOM ODOR
Methods to evaluate the odor characteristics of heated oils were developed by

Evans et al. (21), which involved heating the oil to 190 C and having panelists enter
the room and rate the odor of the room for overall quality and individual odor char-
acteristics using 010 point intensity scales where 2 weak, 5 moderate, and
8 strong. This method was later upgraded by constructing a small room with
controlled air flow and temperature (4).
13. SENSORY EVALUATION OF FRIED FOODS
The sensory properties of fried foods are evaluated either immediately after frying,
e.g., French fries, chicken, fish, or after storage, e.g., potato chips and other snack
CONCLUSIONS 427
foods. Bread cubes have also been used as an alternative by some researchers
(22, 23). Similar to the evaluation of oils, fried foods can be evaluated for flavor
quality using a scale similar to the quality scale used for oils (Figure 1). However,
additional information can be obtained by having panelists rate individual flavor
or off-flavor characteristics using less restrictive category scales, i.e., scales grea-
ter than 3 points or by using unstructured line scales. Petukhov et al. (24) reported a
technique for evaluating the odor properties of stored potato chips that involved
presenting panelists with individual bags of chips. The bags were opened by the
panelist and immediately rated for painty and stale/musty odor using unstruc-
tured line scales. French fries are also evaluated for off-flavors/off-odors as well
as textural properties such as greasiness and crispiness (25).
14. ELECTRONIC NOSE
Recent advances in the technology of multisensor arrays and neural computing have
made the development of the electronic nose of great interest to the food industry
for discrimination between odors (26). Provided the instrument has been calibrated
properly, the technique is rapid, nondestructive, and objective. Shen et al. (27)
found the electronic nose was capable of measuring changes in volatile compounds
associated with lipid oxidation in canola, corn, and soybean oils stored under accel-
erated conditions and Aparicio et al. (28) found the electronic nose could be cali-
brated to detect rancidity levels in good quality olive oil spiked with rancid olive oil.
15. GAS CHROMATOGRAPHYOLFACTOMETERY
Gas chromatography olfactometery (GC-O) provides a sensory profile of odor

active compounds present in an aroma extract by sniffing the GC effluent. Several
techniques have been developed to collect and process GC-O data and to estimate
the sensory contribution of individual odor active compounds, including dilution
analysis (29, 30), time intensity (31), and detection frequency (32) methods.
GC-O has successfully been used to evaluate the odor active compounds of
olive oil (33), soybean oil (34), and fish oil enriched mayonnaise (35).
16. CONCLUSIONS
Sensory evaluation can play an important role in quality assessment programs.

By controlling and using proper testing conditions, and by using trained panelists
and appropriate sensory methodologies, sensory evaluation can provide essential
information necessary to make an informed decision on quality, and can contribute
to a better understanding of product behavior. Sensory evaluation is also necessary
in validating instrumental methods such as the electronic nose and for determining
key odorants using GC-O techniques.
REFERENCES
1. H. Stone and J. L. Sidel, Sensory Evaluation Practices, Academic Press, San Diego,
California, 1993.
2. American Society for Testing and Materials, MNL 26 Sensory Testing Methods, 2nd ed.,
ASTM, Philadelphia, Pennsylvania, 1996.
3. American Society for Testing and Materials, STP 913 Physical Requirement Guidelines
for Sensory Evaluation Laboratories, ASTM, Philadelphia, Pennsylvania, 1986.
4. T. L. Mounts, and K. Warner, in D. R. Erickson et al., eds., Handbook of Soy Oil
Processing and Utilization, American Soybean Association and American Oil Chemists
Society, Champaign, Illinois, 1980.
5. American Society for Testing and Materials, STP 434 Manual on Sensory Testing
Methods, ASTM, Philadelphia, Pennsylvania, 1968.
6. K. Warner, E. N. Frankel, and K. J. Moulton, J. Amer. Oil Chem. Soc., 65:386 (1988).
7. R. Stone, and E. G. Hammond, J. Amer. Oil Chem. Soc., 60:1277 (1983).
AOCS, Champaign, Illinois, 2002, Method Cg 2-83.
9. American Society for Testing and Materials, Standard Practice for the Bulk Sampling,
Handling, and Preparing of Edible Oils for Sensory Evaluation Laboratories, ASTM,
Philadelphia, Pennsylvania, 2000, Method E1346-90 (2000).
10. R. Przybylski, L. J. Malcolmson, N. A. M. Eskin, S. Durance-Tod, J. Mickle, and R. Carr,
Lebensm. Wiss.u. Technol., 26:205 (1993).
11. B. A. Rainey, J. Sensory Studies, 1:149 (1986).
12. K. Warner, in K. Warner and M. Eskin, eds., Methods to Assess Quality and Stability of
Oils and Fat-Containing Foods, AOCS Press, Champaign, Illinois, 1995.
13. American Society for Testing and Materials, STP 758 Guidelines for the Selection and
Training of Sensory Panel Members, ASTM, Philadelphia, Pennsylvania, 1981.
14. L. J. Malcolmson, M. Vaisey-Genser, R. Przybylski, D. Ryland, N. A. M. Eskin, and
L. Armstrong, J. Amer. Oil Chem. Soc., 73:1153 (1996).
15. EC 1991 Commission Regulation EC 2568/91, July 1991, on the Characteristic Methods
of Olive Oils and Their Analytical Methods. EC Journal L 8121.10.91 to the Regulation
EEC/2568/91.
16. IOOC. International Olive Oil Council. COI/T.2o/Document 15/Rev. 1 (Organoleptic
Assessment of Olive Oil. Resolution RES-3/75-IV/96, 20 November), Madrid, Spain,
1996.
17. R. Aparicio and M. T. Morales, J. Sci. Food Agric., 67:247 (1995).
18. L. L. Kok, W. R. Fehr, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 76:1175
(1999).
19. D. Rousseau and A. G. Marangoni, Food Research Int., 31:381 (1999).
20. C. Jacobsen, J. Adler-Nissen, and A. S. Meyer, J. Agric. Food Chem., 47:4917 (1999).
21. C. D. Evans, H. A. Moser, G. R. List, H. J. Dutton, and J. C. Cowan, J. Amer. Oil Chem.
Soc., 48:711 (1971).
22. E. N. Frankel, K. Warner, and K. J. Moulton, J. Amer. Oil Chem. Soc., 62:1354 (1985).
23. H-R. Liu and P. J. White, J. Amer. Oil Chem. Soc., 69:533 (1992).
REFERENCES 429
24. I. Petukhov, L. J. Malcolmson, R. Przybylski, and I. Armstrong, J. Amer. Oil Chem. Soc.,
76:889 (1999).
25. X-Q., Xu, V. H. Tran, M. V. Palmer, K. White, and P. Salisbury, Food Aust., 52:77 (2000).
26. P. N. Barlett, J. M. Elliot, and J. W. Gardner, Food Technol., 51:44 (2001).
27. N. Shen, S. Moizuddin, L. Wilson, S. Duvick, P. White, and L. Pollak, J. Amer. Oil Chem.
Soc., 78:937 (2001).
28. R. Aparicio, S. M. Rocha, I. Delgadillo, and M. T. Morales, J. Agric. Food Chem., 48:853
(2000).
29. F. Ullrich, and W. Grosch, Zeitschrift fur Lebensmittel-Untersuchung und Forschung,
184:277 (1987).
30. T. E. Acree, J. Barnard, and D. Cunningham, Food Chem., 14:273 (1984).
31. R. Miranda-Lopez, L. M. Libbey, B. T. Watson, and M. McDaniel, J. Food Sci., 57:985
(1992).
32. J. P. H. Linssen, J. L. G. M. Janssens, J. P. Roozen, and M. A. Posthumus, Food Chem.,
8:1 (1993).
33. J. Reiners and W. Grosch, J. Agric. Food Chem., 46:2754 (1998).
34. S. M. van Ruth, E. S. Shaker, and P. A. Morrissey, Food Chem., 75:177 (2001).
35. K. Hartvigsen, P. Lund, L. Hansen, and G. Holmer, J. Agric. Food Chem., 48:4858 (2000).
11
Antioxidants: Science,
Technology, and
Applications
P. K. J. P. D. Wanasundara1 and F. Shahidi2
1
Agriculture and Agri-Food Canada Saskatoon Research Center
Saskatoon, Saskatchewan, Canada
2
1. AN ANTIOXIDANTDEFINITION
In a biological system, an antioxidant can be defined as any substance that when

present at low concentrations compared to that of an oxidizable substrate would sig-
nificantly delay or prevent oxidation of that substrate (1). The oxidizable substrate
may be any molecule that is found in foods or biological materials, including car-
bohydrates, DNA, lipids, and proteins. Food is a multicomponent system composed
of a variety of biomolecules, and therefore, this definition describes well an antiox-
idant. However, regulatory bodies that overlook the food-supply categorize antiox-
idants under food additives and define them as substances used to preserve food by
retarding deterioration, rancidity, or discoloration due to oxidation (Code of
431
432 ANTIOXIDANTS: SCIENCE, TECHNOLOGY, AND APPLICATIONS
Federal Regulations, Food and Drug Administration). In foods, much of the work
on antioxidants has emphasized retardation of lipid oxidation, which eventually
triggers and transforms to the oxidation of other macromolecules such as proteins.
It is the intention of this chapter to summarize the available information on the
chemistry, technology, and regulatory aspects of compounds that can delay oxida-
tion of unsaturated fats and lipids in food.
2. HISTORY OF ANTIOXIDANTS AND THEIR USE
Antioxidants may occur as natural constituents of foods, and may intentionally be

added to products or formed during processing. Use of substances to enhance qual-
ity of food by means of delaying lipid oxidation has been in practice for centuries,
although it was not chemically defined or understood. The first recorded scientific
observation on oxidation inhibitors came from Berthollet in 1797 (2) and later from
Davy (3). Their theory was described as catalyst poisoning in oxidative reactors,
and this was well before the free radical theory of peroxidation had been proposed.
Duclaux (4) first demonstrated participation of atmospheric oxygen in oxidation of
free fatty acids. Later, it was found that oxidation of unsaturated acylglycerols can
generate rancid odors in fish oils (5).
The earliest reported work on the use of antioxidants to retard lipid oxidation
appeared in 1843, in which Deschamps showed that an ointment made of fresh
lard containing gum benzoin (contains vanillin) or populin (from polar buds, con-
tains saligenin and derivatives) did not become rancid as did the one with pure lard
(2). Interestingly, the first reports on antioxidants employed for food lipids were
about using natural sources; in 1852, Wright (6) reported that elm bark was effec-
tive in preserving butterfat and lard. Chevreul (7) showed that wood of oak, poplar,
and pine (in the order of decreasing efficacy) retarded the drying of linseed oil films
applied on them, and on all three, it took much longer time to dry than on glass.
Moureu and Dufraise (811) first reported the possibility of using synthetic chemi-
cals, especially phenolic compounds, to retard oxidative decomposition of food
lipids. Their work provided the basic information leading to theories of lipid oxida-
tion and antioxidants, which they referred to as inverse catalysis. Systematic
investigation of antioxidant activity based on the chemistry of radical chain perox-
idation of model chemicals was reported by Lowry and his collegues (12) and
Bolland and tenHave (13) of the British Rubber Producers Research Association.
Antioxidant synergism in food was first reported by Olcott and Mattill (14), and
this was significant in achieving oxidative stability in food by using a combination
of antioxidants found in the unsaponifiable fraction of oils. They described the anti-
oxidants as inhibitors and grouped them into acid type, inhibitols, and hydroqui-
none and phenolics. Bailey (15) and Scott (16) have provided the history and a
descriptive analysis of the development of antioxidants in their books, The Retar-
dation of Chemical Reactions and Antioxidants and Autoxidation, respectively.
Since the early 1960s, the understanding of autoxidation of unsaturated lipids and
antioxidative mechanisms have advanced significantly as a result of development of
SCOPE OF USING ANTIOXIDANTS IN FOOD 433
effective analytical tools. The last two decades have been very important to the anti-
oxidant research. Around the world a revival is seen in studying the natural antiox-
idants in foods and the potential health benefits of natural antioxidants in relation to
prevention and therapy of oxidative stress and related diseases. The emphasis has
largely been on their implications on vital biological reactions that have a direct
relationship to tissue injury and degenerative diseases. Enough scientific evidences
have already been accumulated in relation to these conditions with free radicals and
reactive oxygen species. Therefore, not only enhancing the shelf life stability of
foods has been examined, but also control of lipid oxidation by suppressing free
radical formation in foods to prevent their deleterious health effects has become
important. The quest for understanding the oxidation of lipids and its prevention
and control has continued since historical times and is still on.
3. SCOPE OF USING ANTIOXIDANTS IN FOOD
The function of an antioxidant is to retard the oxidation of an organic substance,

thus increasing the useful life or shelf life of that material. In fats and oils, anti-
oxidants delay the onset of oxidation or slow the rate of oxidizing reactions. Oxida-
tion of lipids chemically produces compounds with different odors and taste and
continues to affect other molecules in the food. The main purpose of using an anti-
oxidant as a food additive is to maintain the quality of that food and to extend its
shelf life rather than improving the quality of the food. Figure 1 illustrates how anti-
oxidants can affect the quality maintenance of food in terms of oxidative rancidity
Figure 1. Typical curves for oxidation of lipids (a) No antioxidant added; (b) and (c) represent
added or endogenous antioxidants. Antioxidant activity of (c) is higher than (b). IP1, IP2, and IP3
are induction period in hours or days.
development. Use of antioxidants reduces raw material wastage and nutrition loss
and widens the range of fats that can be used in specific products. Thus, antioxi-
dants are useful additives that allow food processors to use fats and oils economic-
ally in their product formulation.
4. OXIDATION OF FATS AND OILS AND MECHANISM

OF ANTIOXIDANTS
In fats and oils, the process of oxidation is similar to that oxidation of any other
unsaturated organic material and requires an initiation process, in order to generate
free radicals from the substrate. As antioxidants inhibit oxidation or autoxidation
process, the mechanism(s) involved need(s) to be discussed. Figure 1 explains
the relationship of antioxidant activity and oxidation of a lipid as examined by a
typical evaluation method.
Autoxidation is the oxidative deterioration of unsaturated fatty acids via an auto-
catalytic process consisting of a free radical chain mechanism. This chain includes
initiation, propagation, and termination reactions that could be cyclical once
started. The initiation process generates free radicals from the substrate. The a-
methylenic H atom is abstracted from the unsaturated lipid molecule to form a lipid
(alkyl) radical (R) (Scheme 1, Equation [1]). The lipid radical is highly reactive
and can react with atmospheric oxygen (3O2), a facile reaction resulting from the
diradical nature of the oxygen molecule, and it produces a peroxy radical (ROO)
(Scheme 1, Equation [2]). In the propagation reactions, the peroxy radical reacts
with another unsaturated lipid molecule to form a hydroperoxide and a new
unstable lipid radical (Scheme 1, Equation [3]). As a new free radical is generated
at each step, more oxygen is incorporated into the system. The newly propagated
lipid radical will then react with oxygen to produce another peroxy radical, result-
ing in a self-catalyzed, cyclical mechanism (Scheme 1, Equation [4]).
I
RH [1]
R + 3O2 [2]
ROO + RH [3]
ROOH [4]
RO+ RH [5]
Scheme 1. Possible reactions of the autoxidation process. R is an alkyl group of an

unsaturated lipid molecule. H is an a-methylenic hydrogen atom easily detachable because of
the activating influence of the neighboring double bond or bonds. RO is alkoxy radical,
ROO is peroxy radical, and I is an initiator.
OXIDATION OF FATS AND OILS AND MECHANISM OF ANTIOXIDANTS 435
Hydroperoxides are unstable and may degrade to radicals that accelerate propaga-
tion reactions. These are branching steps of lipid autoxidation process (Scheme 1,
Equations [5] and [6]). This chain reaction proceeds, and termination occurs only
when two free radicals combine to form a nonradical product. Autoxidation can
break down the substrate molecules as well as forming new molecules causing
gross changes in the chemical and physical properties of the oxidizing substrate
(1719). Degradation of hydroperoxides may generate new molecules that have
undesirable odors and flavors, associated with oxidative rancidity of unsaturated
lipids. Such sensory perceivable changes are noted when oxidation of unsaturated
lipids has been progressed to advanced stages. This is only a brief description of
autoxidation process.
A lipid that contains double bonds undergoes autoxidation induced by various
ways. It is now clear that metal-catalyzed decomposition of preformed hydroperox-
ides is the most likely cause for the initiation process. The direct oxidation of unsa-
turated lipids by triplet oxygen (3O2) is spin forbidden. This is because of the
opposite spin direction of ground state lipid of single multiplicity and oxygen of
triplet multiplicity, which does not match. When initiators are present, this spin bar-
rier between lipids and oxygen can readily be overcome and produce radicals by
different mechanisms. Ground state oxygen may be activated in the presence of
metal or metal complexes and can initiate oxidation either by formation of free radi-
cals or singlet oxygen. Exposure of lipids to light, metals, singlet oxygen and sen-
sitizers (chlorophyll, hemoproteins, and riboflavin), or preformed hydroperoxide
decomposition products causes generation of primary hydroperoxides. Photosensi-
tized oxidation or lipoxygenase-catalyzed oxidation also produces hydroperoxides.
Thermal oxidation is also autocatalytic and considered as metal-catalyzed
because it is very difficult to eliminate trace metals (from fats and oils or food)
that act as catalysts and may occur as proposed in Equation 4. Redox metals of
variable valency may also catalyze decomposition of hydroperoxides (Scheme 2,
Equations [6] and [7]). Direct photooxidation is caused by free radicals produced
by ultraviolet radiation that catalyzes the decomposition of hydroperoxides and per-
oxides. This oxidation proceeds as a free radical chain reaction. Although there
should be direct irradiation from ultraviolet light for the lipid substrate, which is
usually uncommon under normal practices, the presence of metals and metal com-
plexes of oxygen can become activated and generate free radicals or singlet oxygen.
ROOH + M2+ -
+ M3+ [6]
ROOH + M3+ H+ + M2+ [7]
Scheme 2. Possible reactions of generating hydroperoxides (Mn is the metal ion with
transitional valency).
Photosensitized oxidation is a direct reaction of light-activated, singlet oxygen with

unsaturated fatty acids, and subsequently hydroperoxides are formed. Photosensitized
oxidation happens because of the presence of molecules that can absorb visible or
near UV light to become electronically excited (sensitizers) (Equation 8). Pigments
initiating photosensitized oxidation in foods include chlorophylls, hemoproteins,
and riboflavin. The type I sensitizer serves as photochemically activated free radical
initiator, and type II sensitizers in the triplet state interact with oxygen by energy
transfer to form singlet oxygen (1O2) that reacts further with unsaturated lipid
(Equation 9). Under photosensitized oxidation conditions, the reaction of unsatu-
rated lipids with singlet oxygen (1O2) leads to rapid formation of hydroperoxides
(Equations 10) (17, 19, 20).
Sensitizerground +h Sensitizerexited [8]
Sensitizerexcited + 3O2 Sensitizerground + 1O2 [9]

1
O2 + RH ROO + H [10]
Scheme 3. Formation of hydroperoxides by photoxidation of a lipid with a sensitizer (hv is energy

in the form of UV light, sensitizers that are naturally present in photosensitive pigments, their
degradation products, or polycyclic aromatic hydrocarbons capable of transferring energy from
light to chemical molecules).
5. CLASSIFICATION OF ANTIOXIDANTS
Antioxidants may be broadly grouped according to their mechanism of action: pri-

mary or chain breaking antioxidants and secondary or preventive antioxidants.
According to this classification, some antioxidants exhibit more than one mechan-
ism of activity, therefore, referred to as multiple-function antioxidants. Another
commonly used classification categorizes antioxidants into primary, oxygen scaven-
ging, and secondary, enzymatic and chelating/sequestering antioxidants. However,
synergistic antioxidants are not included in this classification. During the past two
decades, several naturally occurring compounds have been added into the list of
antioxidants that are effective against oxidation of unsaturated fats and oils and
most of them fall into the multifunctional category. Classification of antioxidants
according to the mode of activity as primary and secondary is preferred in this dis-
cussion.
5.1. Primary Antioxidants

Primary antioxidants are also referred to as type 1 or chain-breaking antioxidants.
Because of the chemical nature of these molecules, they can act as free radical
acceptors/scavengers and delay or inhibit the initiation step or interrupt the
propagation step of autoxidation. Figure 2 illustrates possible events that primary
Figure 2. Possible interactions of primary and secondary antioxidants with lipid oxidation pathway in foods.
437
antioxidants may interfere along the lipid autoxidation pathway. Primary antioxi-
dants cannot inhibit photosensitized oxidation or scavenge singlet oxygen.
ROO + AH [11]
R [12]
ROO + A [13]
RO + AH [14]
RO + A [15]
A + A [16]
Scheme 4. Mechanism of primary antioxidant activity (AH is an antioxidant molecule).
The first kinetic study of antioxidant activity was conducted by Boland and tenHave
(13) who postulated the Equations 11 and 12. The primary antioxidants (AH) react
with lipid and peroxy radicals (ROO) and convert them to more stable, nonradical
products as shown in Scheme 4, Equations 13 and 14. These antioxidants are cap-
able of donating a hydrogen atom to lipid radicals and produce lipid derivatives and
antioxidant radicals (A) that are more stable and less readily available to partici-
pate in propagation reactions (Equation 12). Primary antioxidants have higher affi-
nities for peroxy radicals than lipids and react predominantly with peroxy radicals.
The following reasons have been listed for their high affinity. Propagation is the
slow step in lipid oxidation process; thus, peroxy radicals are found in compara-
tively larger quantities than other radicals. In addition, peroxy radicals have lower
energies than alkoxy radicals; therefore, they react more readily with the low-
energy hydrogen of primary antioxidants than unsaturated fatty acids. As the free
radical scavengers are found in low concentration, they do not compete effectively
with initiating radicals (e.g., hydroxyl radicals) (21, 22). Therefore, primary antiox-
idants inhibit lipid oxidation more effectively by competing with other compounds
for peroxy radicals, and they are able to scavenge peroxy- and alkoxy-free radicals
formed during propagation (Equation 3) and other reactions (Equations 4 and 5) in
autoxidation.
The antioxidant radical produced because of donation of a hydrogen atom has a
very low reactivity toward the unsaturated lipids or oxygen; therefore, the rate of
propagation is very slow. The antioxidant radicals are relatively stable so that they
do not initiate a chain or free radical propagating autoxidation reaction unless pre-
sent in very large quantities. These free radical interceptors react with peroxy radi-
cals (ROO) to stop chain propagation; thus, they inhibit the formation of peroxides
(Equation 13). Also, the reaction with alkoxy radicals (RO) decreases the decom-
position of hydroperoxides to harmful degradation products (Equation 14).
Most of the primary antioxidants that act as chain breakers or free radical
interceptors are mono- or polyhydroxy phenols with various ring substitutions.
CLASSIFICATION OF ANTIOXIDANTS 439
0
TABLE 1. Standard One-Electron Reduction Potential E o
at pH 7 for Selected Radical Couples (Adapted from Ref. 21).
0
Couple E o (mV) at pH 7.0
HO, H/H2O 2310

RO, H/ROH (alkoxy) 1600
ROO, H/ROOH (peroxyl) 1000
PUFA, H/PUFA-H (polyunsaturated fatty acid) 600
a-Tocopheroxyl, H/a-tocopherol 500
Ascorbate, H/ascorbate 282
Dehydroascorbic/ascorbate

174
The antioxidant effectiveness is influenced by the chemical properties of the com-

pound including hydrogen bond energies, resonance delocalization, and susceptibil-
ity to autoxidation. The ability of the primary antioxidant molecule to donate a
hydrogen atom to the free radical is the initial requirement. The ability of the
free radical interceptor (scavenger) to donate a hydrogen atom to a free radical
can be predicted from standard one-electron potentials (Table 1). According to
Buettner (21), each oxidizing species is capable of stealing an electron (or H
atom) from any reduced species listed below it. That means when the standard
one-electron reduction potential is concerned, the free radical scavengers that
have reduction potential below peroxy radicals are capable of donating an H
atom to peroxy radical and form a peroxide. The resulting antioxidant radical
should be of low energy, ensuring the lesser possibility of catalyzing the oxidation
of other molecules. The formed antioxidant radical is stabilized by delocalization of
the unpaired electron around the phenol ring to form a stable resonance hybrid (Fig-
ure 3) and as a result attained low-energy levels (18, 22, 23).
Antioxidant radicals are capable of participating in termination reactions with
peroxy (Equation 13), alkoxy (Equations 14 and 15), or antioxidant (Equation
16) radicals removing reactive free radicals from the system. In fats and oils con-
taining phenolic antioxidants, dimers of antioxidant molecules are usually found.
This is a good indication that antioxidant radicals readily undergo termination reac-
tions and form dimers as proposed in Equation 13. When considering all of these,
the primary antioxidants or free radical scavengers can inactivate at least two free
radicals, the first one during the interaction with peroxy radical and the second in
the termination reaction with another peroxy radical.
OH O O O O
ROO ROOH
Figure 3. Stable resonance hybrids of phenoxy radical of phenolic antioxidant [adapted from
(18)].
OH OH OH
C(CH3)3 (H3C)3C C(CH3)3
C(CH3)3
OCH3 OCH3 CH3
2-BHA 3-BHA BHT

Butylated hydroxyanisole (BHA) Butylated hydroxytoluene (BHT)
OH OH
C(CH3)3 HO OH R = C3H7 Propylgallate
R = C8H17 Octylgallate
R = C12H25 Dodecylgallate
OH COO-R
Tertiary-butylhydroquinone (TBHQ) Gallates
CH3
C2H5O
CH3
N CH3
H
6-Ethoxy-1,2-dihydro-2,2,4-trimethylquinoline (Ethoxyquin)
Figure 4. Chemical structures of synthetic phenolic antioxidants commonly used in fats and oils.
The compounds that exhibit primary antioxidant activity include polyhydroxy

phenolics as well as the hindered phenolics. There are several synthetic ring-
substituted phenolics as well as naturally occurring phenolic compounds that
may perform via the primary antioxidant mechanism, as discussed later in this
chapter. The common feature of all of these antioxidants is that they are mono-
or polyhydroxy phenols with various ring substitutes (Figure 4). Substitution
with an electron-donating group/s ortho and/or para to the hydroxyl group of phe-
nol increases the antioxidant activity of the compound by an inductive effect (e.g.,
2,6-di-tert-butyl-4-methylphenol or BHA). Thus, the presence of a second hydroxyl
group in the 2- (ortho) or the 4-position (para) of a phenol increases the antioxidant
activity (e.g., TBHQ). In the dihydroxybenzene derivatives, the semiquinoid radical
produced initially can be further oxidized to a quinone by reacting with another
lipid radical (Figure 5). This semiquinoid radical may disproportionate into a qui-
none and a hydroquinone molecule, and the process of this conversion contributes
to antioxidant activity as peroxy radical scavenging potential (18, 24). Table 2 sum-
marizes most commonly used primary antioxidants in fats and oils and lipid-
containing foods. Substitution with butyl or ethyl group/s para to the hydroxy
groups also enhances the antioxidant activity. Substitution of branched alkyl groups
at ortho positions enhance the ability of the molecule to form a stable resonance
structure that reduces the antioxidant radicals participation in propagation
reactions (18, 23).
OH
R ROOH
O
R
OH
ROO Semiquinoid radical
Dihydroxy quinone
derivative OH
ROO O
R
ROOH
OH
OH
R
O
R
OH
Hydroquinone derivative
O
Quinone derivative
Figure 5. Possible mechanism of antioxidant activity of dihydroxybenzene derivative.
To be most effective, primary antioxidants should be added during the induction

or initiation stage of the autoxidation reaction cascade. Antioxidants can scavenge
the formed free radicals, as the cyclical propagation steps have not occurred at this
stage. Addition of primary antioxidants to a lipid that already contains substantial
amounts of lipid peroxides may result in loss of antioxidant activity (25).
5.2. Secondary Antioxidants

Secondary antioxidants are also classified as preventive or class II antioxidants.
They offer their antioxidant activity through various mechanisms to slow the rate
of oxidation reactions. The main difference with primary antioxidants is that the
TABLE 2. Primary Antioxidants that are Commonly

Used in Foods.
Natural Synthetic
Carotenoids Butylated hydroxyanisole (BHA)

Flavonoids Butylated hydroxytoluene (BHT)
Phenolic acids Ethoxyquin
Tocopherols and tocotrienols Propyl gallate (PG)
Tertiary-butylhydroquinone (TBHQ)
TABLE 3. Compounds that Exhibit Secondary Antioxidant Activity.
Mode of Activity Compounds in use
Metal chelation Cirtic, Malic, Succinic and Tartaric acids

Ethylenediaminetetraacetic acid, Phosphates
Oxygen scavenging and reducing agents Ascorbic acid, Ascorbyl palmitate, Erythorbic acid,
Sodium erythorbate, Sulfites
Singlet oxygen quenching Carotenoids (b-Carotene, Lycopene and Lutein)
secondary antioxidants do not convert free radicals into stable molecules. They act
as chelators for prooxidant or catalyst metal ions, provide H to primary antioxi-
dants, decompose hydroperoxide to nonradical species, deactivate singlet oxygen,
absorb ultraviolet radiation, or act as oxygen scavengers. They often enhance the
antioxidant activity of primary antioxidants. Table 3 provides examples of some
of these compounds that exhibit secondary antioxidant activity.
5.2.1. Sequestering/Chelating Agents or Metal Deactivators Heavy metals

with two or more valency states with a suitable oxidation-reduction potential
between them (e.g., Co, Cu, Fe, Mn, etc.) shorten the induction period and increase
the maximum rate of oxidation of lipids. Trace amounts of these metal ions are pre-
sent in the lipid-containing foods coming from naturally present compounds or
included during processing operations. The effectiveness of copper as a catalyst
for hydroperoxide decomposition has been reported (2628). Transition metals
such as iron exhibit low solubility at pH values near neutrality (29). That means
in foods, transition metals may exist chelated to other compounds; many com-
pounds form complexes with these metals and change their catalytic activity. Che-
lation can increase the prooxidant activity of transition metals by making them
more nonpolar (increase solubility in lipids; 30), and some can increase oxidative
reactions by increasing metal solubility or altering redox potential (31). According
to Graf and Eaton (32), chelators may exert antioxidant activity by prevention of
metal redox cycling, occupation of all metal coordination sites, and formation of
insoluble metal complexes and steric hindrance of interactions between metals
and lipids or oxidation intermediates (e.g., peroxides). Chelation of these metal
ions or use of metal deactivators reduces the pro-oxidant activity by raising the
energy of activation for intiation reactions. The most effective form of chelating
agents as secondary antioxidants, which form s-bonds with metal ions because
they reduce the redox potential and stabilize the oxidized form of the metal ion.
Chelating agents such as heterocyclic bases that form p-complexes raise the
redox potential and may accelerate metal-catalyzed hydroperoxide decomposition
(Equations 7 and 8) and act as prooxidants.
Multiple carboxylic acid compounds such as citric acid, ethylenediaminetetraa-
cetic acid (EDTA), and phosphoric acid derivatives (polyphosphates and phytic
acid) are commonly used in extending the shelf life of lipid-containing foods
because of their metal chelating properties. Typically these chelators are water solu-
ble, but citric acid exhibits solubility in lipids, which allows it to inactivate metals
in the lipid phase (33). Chelators activity depends on pH and the presence of other
chelatable ions (e.g., Ca). Most food grade chelators are unaffected by food-proces-
sing operations and storage; however, polyphosphates may decrease their antioxi-
dant activity because of possible hydrolysis by endogeneous phosphatases in foods,
especially in raw meat (22).
Several proteins that exist in food (e.g., lactoferrin, ferritin, transferritin, heme
protein) possess strong binding sites for iron. Reducing agents (ascorbate, cysteine,
superoxide anion) to low pH causes release of iron from proteins and accelerates
lipid oxidation (34). Some amino acids and peptides found in muscle foods (e.g.,
carnosine) are capable of chelating metal ions and inhibit their prooxidant activity
(35, 36).
5.2.2. Oxygen Scavengers and Reducing Agents As oxygen is essential and

is one of the reactants in the autoxidation process, scavenging of oxygen molecular
species is one way of providing antioxidant activity. Ascorbic acid acts as a reduc-
ing agent and as an oxygen scavenger. The mechanism of antioxidant activity of
ascorbic acid is discussed elsewhere in this chapter.
Singlet oxygen is the excited state oxygen, and its inactivation is an effective
way of preventing initiation of lipid oxidation. Carotenoids are capable of inactivat-
ing photoactivated sensitizers by physically absorbing their energy to form the
excited state of the carotenoid. Later, the excited state carotenoid returns to ground
state by transferring energy to the surrounding solvent (37, 38). Other compounds
found in food, including amino acids, peptides, proteins, phenolics, urates, and
ascorbates also can quench singlet oxygen (20).
Compounds such as superoxide anion and peroxides do not directly interact with
lipids to initiate oxidation; they interact with metals or oxygen to form reactive spe-
cies. Superoxide anion is produced by the addition of an electron to the molecular
oxygen. It participates in oxidative reactions because it can maintain transition
metals in their active reduced state, can promote the release of metals that are
bound to proteins, and can form the conjugated acid, perhydroxyl radical depending
on pH, which is a catalyst of lipid oxidation (39). The enzyme superoxide dismu-
tase that is found in tissues catalyzes the conversion of superoxide anion to hydro-
gen peroxide.
Catalase is capable of catalyzing the conversion of hydrogen peroxides to water
and oxygen (40). Glucose oxidase coupled with catalase is well used commercially
to remove oxygen from foods, especially fruit juices, mayonnaise, and salad dres-
sings (19). Glutathione peroxidase that is found in many biological tissues also
helps to control both lipid and hydrogen peroxides (41, 42). These enzymic reac-
tions help to reduce various types of radicals that could be formed in lipid-contain-
ing biological systems.
5.3. Synergism and Synergists

Synergism is the cooperative effect of antioxidants or an antioxidant with other
compounds to produce enhanced activity than the sum of activities of the individual
Figure 6. Synergistic effect of antioxidants: (a) 0.32% (w/w) dipalmitoyl phosphatidylethanola-

mine, (b) 0.02% (w/w) propyl gallate, and (c) is (a)(b), evaluated in lard at 120 C, and the
induction period was used to compare antioxidant activity [modified and redrawn from (44)].
component when used separately (43). Figure 6 illustrates how synergistic effect is
expressed as antioxidant activity. Two types of synergism are observed, one invol-
ving primary antioxidants only and the other involving a combination of primary
antioxidants with metal chelators or peroxy scavengers.
In a combination of two or more free radical scavengers, rapid reaction with free
radicals occurs because of the differences in bond dissociation energies or steric
hindrance of free radical scavenger/ROO interactions (23). These differences result
in one scavenger being used faster than the other. Also, it is possible to regenerate
the primary antioxidant by transferring its radical to another scavenger. Ascorbic
acid together with a-tocopherol also shows a good synergism, which is explained
by the regeneration and recycling of the tocopheroxyl radical intermediate to the
parent phenol, a-tocopherol (44, 45).
In the combination(s) of free radical scavenger and metal chelator, the chelator
decreases the oxidation rates by inhibiting metal-catalyzed oxidation; thus, fewer
free radicals are generated in the system. Inactivation of antioxidants via termina-
tion reaction or participation in autoxidation occurs to a lesser extent in such situa-
tions. This makes the concentration of antioxidant, which is available to scavenge
free radicals, to be always greater at a given time than when no metal chelator is
present. Therefore, the combination of chelator and radical scavenger decreases
free radical generation and increases radical scavenging potential (22). Strong
synergistic activity has been observed in the mixtures of natural tocopherols and
citric acid. The synergistic effect of this mixture is caused by the chain-breaking
ability of tocopherols and metal chelation of citric acid (19).
In an antioxidant combination that contains compounds exhibiting different
mechanisms of action and physical properties, inhibition of oxidation occurs in
EVALUATION OF ANTIOXIDANT ACTIVITY 445
many different phases. This suggests that food antioxidants should be carefully
selected considering such factors as the type of oxidation catalyst, physical state
of lipid (bulk, emulsified), pH, temperature, and the ability to interact with other
components in the food.
6. EVALUATION OF ANTIOXIDANT ACTIVITY
Antioxidant activity of a given compound is assessed as resistance to oxidation of

lipids in the presence of that particular compound. Therefore, most of the
methods described and used to assess antioxidant activity follow oxidation and
the stages of oxidation of unsaturated lipid substrates. Many techniques have
been developed to determine the antioxidant efficacy of the compounds of interest,
but all of these have to be employed and interpreted carefully. Frankel (19, 46) has
listed following parameters that are fairly important in choosing methods to evalu-
ate antioxidants.
Substrate: Should be relevant to foods. Triacylglycerols and phospholipids, in

the bulk, emulsion, or liposome form represent the closest model in biological
systems including foods.
Conditions: Test under various conditions. Under different temperatures, with
metal catalysts, with surface exposure, etc., select these conditions to mimic
conditions in food.
Analysis: Measure relatively low levels of oxidation (below 1%) and include
measurement of initial or primary products of lipid oxidation (e.g., hydro-
peroxides, conjugated dienes) as well as secondary decomposition products
of lipid oxidation (e.g., carbonyls, volatiles, dialdehydes).
Concentrations: Compare antioxidants at the same mole concentration of active
compound, and an appropriate reference compound should also be used,
which may be a structurally related reference compound. With crude extracts
(e.g., natural antioxidants), compositional data are needed to compare
samples.
Calculations: Use the induction period, percentage inhibition or rates of hydro-
peroxide formation or decomposition, or IC50 value (concentration required
to achieve 50% inhibition) based quantification (this is discussed further in a
later section).
Obviously attention should also be paid to the system under examination. Thus,
bulk oil, water-in-oil, or oil-in water emulsions behave differently under similar
oxidation conditions.
An updated review by Antolovich et al. (47) discusses the methods of determin-
ing antioxidant activity extensively. The methods used in measuring antioxidant
activity may be categorized into three groups, which directly or indirectly measure
the rate or extent of the following:
1. Decay of substrate, probing compound, or oxygen consumption

2. Formation of oxidation products by the oxidizing substrate
3. Formation or decay of probing free radicals.
Methods that use approaches (1) and (2) measure antioxidant activity as an inhi-
bitory effect exerted by the test compound on the extent or rate of consumption of
reactants or the formation of oxidation products. The antioxidant activity (AA) of a
compound or a component mixture that is a function of many parameters of the
assay method employed may be defined using the following mathematical expres-
sions (47; Schemes 5 and 6):
AA f time or rate; temperature; substrate; concentration of antioxidant;

concentration of other substancesa ; partitioning behavior
a
e:g:; oxygen; peroxides; or other antioxidants or prooxidants
For a fixed set of assay conditions, AA could be defined independent of the test
method. Scheme 5 provides equations for a situation that measures time as the inde-
pendent variable.
AA = (tAH tCONTROL)/([AH])tCONTROL
where
tAH is time taken by the substrate to reach a predetermined level of oxidation based on the
test method
t CONTROL is the time for untreated substrate or control to reach the same level of oxidation
[AH] is the concentration of antioxidant in appropriate units
After rearranging,
AA = [(tAH / tCONTROL) 1] / [AH]
Scheme 5. Proposed expressions to calculate antioxidant activity.
According to the equations in Scheme 5, if:
tCONTROL tAH ; no antioxidant activity is exerted

tCONTROL < tAH ; an antioxidant activity is exhibited
tCONTROL > tAH ; a prooxidant activity is observed and AA has a negative value
A similar expression can be formulated for rate of oxidation.

Another expression that can be used is relative antioxidant activity (RAA;

Scheme 6).
RAAAH = AAAH/AAREFERENCE
where
AAAH is antioxidant activities of test compound
AAREFERENCE is the antioxidant activity of the reference antioxidant at the same molar
concentration
Rearranging gives,
AAAH = (RAAAH ) AAREFERENCE
Scheme 6. Proposed expressions to calculate relative antioixidant activity.
RAA represents the activity equivalence of the test compound relative to the
reference antioxidant, which is suitable for activity comparison.
Methods of category (3) tract the capacity of the test compound to capture radi-
cals or to inhibit radical formation rather than monitoring the actual oxidation pro-
duct formation or substrate oxidation. Several new methods are developed based on
this concept, and a variety of new parameters for expressing results are used. It is
expected that a high correlation exists between these two types of measurements. It
should be noted here that there are no standard units for reporting the antioxidant
activity because such activity (assay, capacity, efficiency, effectiveness, etc.) is
independent of the test procedure. Table 4 summarizes the methods available for
measuring antioxidant activity and how the results of such determinations are
expressed.
Another way of categorizing the methods of determining antioxidant activity
is (1) accelerated stability tests, and (2) free radical-based methods. Most of the
studies that are currently used tend to employ accelerated test systems and try to
relate them to real food systems.
6.1. Methods Based on Lipid/Substrate Oxidation (Stability Tests)

These methods are based on lipid (substrate) oxidation and specific to the analysis
of oxidation that occurs in food lipids. The tests employed strongly correlate to the
conditions that oils and fats are subjected to during processing, food preparation,
and storage. The substrate is a model compound that could be a pure triacylgly-
cerol, fatty acid methyl ester, or an actual edible oil/lipid. Favorable conditions
for substrate oxidation (e.g., high temperature) are provided to facilitate increased
rate of oxidation reactions in a controlled environment. The end point is determined
TABLE 4. Methods, Entities Tested, and Units to Express Results in Determining Antioxidant Activity.
Method/Test Reference/s Measurement Results and Units
Substrate oxidation or oxidation

product formation 9 9
Active oxygen method 49, 50 >
>
> Change of mass, peroxide value or >
> Induction period (h, d)
>
> >
>
>
= hydroperoxides, conjugated dienes, >
> Time to reach a set level of oxidation
>
Oven storage test 51, 52 2-thiobarbituric acid reactive substances, >
>
>
> during preinduction period (h, d)
>
> >
>
> anisidine value, formation of hexanal, >
> Rate of oxidation during preinduction
>
> >
>
Shelf storage test 48
>
; ethane, or pentane >
> period (mol kg1hr1, g L1d1)
>
>
>
> Concentration required to produce
>
>
>
= equivalent effect to reference antioxidant
during preinduction period (mol kg1,
>
>
>
> g L1)
>
>
>
> Concentration of a functional group after a
>
>
>
> set time period (mequiv. kg1)
>
>
>
> Concentration of an oxidation product after
>
>
>
> a set time period (mg kg1, ppm w/w)
>
>
>
> Scale reading after a set time period
>
>
;
(absorbance, conductivity, etc.)
Free radical capturing or suppression of

formation 9
DPPH quenching assay 53, 54 >
Ability to quench DPPH radical in solution >
>
>
>
> Percentage inhibition, EC50 (concentration of
>
>
Hydroxyl radical quenching assay 55 Ability to quench hydroxyl radicals >
> test compound required to decrease the
=
generated in a model system concentration of test free radical by 50%),
>
>
>
> TEC50 (time to decrease concentration of
>
>
Ability to quench superoxide radicals >
> test free radical by 50%).
Superoxide radical quenching assay 5658 >
>
generated in a model system ;
Electron paramagnetic resonance (EPR) 5961 Detects free radical involved in autoxidation Intensity or rate of change in EPR signal
spectrometry/spin trap tests and related process
Ferric Reducing Antioxidant Power 6264 Spectrophotometric measurement of Fe (II) Change of absorbance
complex formed due to reducing ability of
the test compounds
Oxygen Radical Absorption Capacity 6567 Based on Phycoerythrin assay Fluorencence intensity, mmol of Trolox
equivalents
Total Radical-trapping Antioxidant 68, 69 Measure oxygen consumption during controlled mmol peroxy radical deactivated L1
Parameter lipid oxidation induced by thermal decom-
position of 2,20 Azobis(2-aminopropane)
hydrochloride; AAPH
Trolox Equivalent Antioxidant Capacity 70, 71 Based on inhibition of production of mM L1 Trolox equivalents
2,20 Azinobis(3-ethylbenzthiazoline)-6-
sulfonic acid; ABTS radical cation and mM
concentration of a Trolox solution having
antioxidant capacity equivalent to 1.0-mM
solution of test substance
TABLE 5. Commonly Used Accelerated Stability Tests for Oils in Evaluating Antioxidants
(72).
Method/Test Conditions and characteristics
Ambient storage Atmospheric pressure and room temperature, too slow and
time consuming
Active oxygen method (AOM) Bubbling air in a closed environment, 98 C, do not represent
normal storage
Rancimat is the automated version, also OSI instrument
Light Atmospheric pressure and room temperature, rapid screening
test, photo-oxidation occurs
Metal catalysts Atmospheric pressure and room temperature, rapid screening
test, more decomposition occurs
Oxygen uptake Atmospheric pressure, 80100 C, do not represent normal
storage
Oxygen bomb 65115 psi, O2, 99 C, do not represent normal storage
Schaal oven Atmospheric pressure, 6070 C, generally correlates well
with actual storage
Weight gain Atmospheric pressure, 3080 C, not always very sensitive
by measuring chemical (e.g., primary or secondary oxidation products) or physical

changes (e.g., change of mass or energy) of the oxidizing substrate. Table 5
provides a summary of methods commonly used for stability testing of edible oils.
6.1.1. Shelf Storage Test The test material is stored under similar conditions as
in retail and is evaluated for the effectiveness of antioxidants in prolonging the pre-
mium quality of the product. Periodic evaluation of the lipid oxidation products
(primary or secondary) by chemical tests (e.g., peroxide value, conjugated diene
value, 2-thiobarbituric acid reactive substances, hexanal content) or sensory evalua-
tion will be used to find out the onset of oxidation. The main drawback of this kind
of evaluation is the time taken; therefore, rapid evaluation or accelerated methods
are often preferred (19, 51).
6.1.2. Active Oxygen Method (AOM) This is one of the widely used methods
for evaluating antioxidant activity. This test involves bubbling air through the
heated lipid sample to accelerate its oxidation. Periodic analysis of peroxide value
is carried out to determine the time required for the fat to oxidize under the condi-
tions provided by AOM. This method has also been referred to as the Swift stability
test. The fully automated version of this method is available as Rancimat apparatus
(Metrohm Ltd, Herisau, Switzerland) and is accepted as a standard method by ISO
(ISO6886) and American Oil Chemists Society (AOCS Cd 12b-92) (47 49, 73).
Similarly, the Oxidative Stability Instrument (OSI, Omnion, Inc., Rockland, MA)
uses a similar principle as AOM. This instrument is sensitive to the change of con-
ductivity of water, which receives the air passed through the oxidizing lipid. OSI
uses induction period or oxidative stability index as the measure of stability of
the antioxidant-containing lipid, and it is accepted by the AOCS (73) [AOCS

method Cd-126-92 (97)].
6.1.3. Oven Storage Test The lipid with or without antioxidants is allowed to
oxidize in an electrically heated convection oven (6070 C). The oil is periodically
assessed for change of its mass and its formation of primary oxidation products
(hydroperoxides; peroxide value, conjugated dienes; conjugated diene value) or
secondary products of oxidation (aldehydes; hexanal, dialdehydes, 2-thiobarbitiuric
acid reactive substances; TBARS) or off-odor formation (5052) . This method is
commonly referred to as the Schaal oven method and is widely used for bulk oil
substrates. Conditions provided in this process are suitable for a low degree of oxi-
dation; thus, the results correlate well with actual shelf stability of the antioxidant-
containing lipids.
6.1.4. Multiphase Systems Antioxidant activity depends very much on the

lipid substrate used for evaluation and the hydrophilic/lipophilic nature of the anti-
oxidative compound. Solubility and partition properties of the compound in the
medium affect the activity of antioxidants in the bulk lipid systems. As most foods
cannot be related to bulk oil systems (e.g., meat, fish, eggs, mayonnaise, salad dres-
sings, etc.), evaluation of antioxidants in multiphase systems is more relevant to
their physical and chemical nature. Because of the very same reasons, several stu-
dies have found that compounds exhibiting strong activity against oxidation of
lipids in bulk systems are often inefficient in colloidal and emulsion systems.
Three systems are generally used for such evaluations: emulsions (water-in-oil,
oil-in-water), liposomes (uni- or multilamellar vesicles formed with aqueous phase
and phospholipids), and micelles (emulsions formed with free fatty acids and aqu-
eous phase). A lengthy discussion about how to use these multiphase systems in
evaluating antioxidant activity is provided by Frankel (19).
Porter (74) has proposed a theory for distinguishing the effectiveness and beha-
vior of antioxidants in bulk oils and emulsions and membranes. The polar para-
doxical behavior (75, 76) describes the anomalous effect of antioxidants on lipids
when they are in different physical systems. Work carried out by Frankel et al. (77)
used the interfacial partitioning phenomenon to explain the reciprocal effect of anti-
oxidants in bulk oil versus multiphase/colloidal systems. This phenomenon recog-
nizes the discrete phases of the oxidative stability and antioxidative mechanism;
however, exact details are not yet fully understood. The oxidative stability of
food lipids varies according to their colloidal location because of the exposure of
the lipid to the antioxidant/proxidant is different in such an environment. The par-
titioning of the antioxidant between the aqueous and nonaqueous phases depends
on their solvent properties. Also, the partitioning of antioxidants into the nonaqu-
eous phase can exert an important effect on activity by protecting the lipid oxidiz-
able substrate. Figure 7 explains probable interfacial distribution of antioxidative
compounds based on their hydrophobicity and hydrophilicity in monophasic and
bipahsic systems. Surfactants can improve the solubility of lipophilic antioxidants
in the interface and exert a significant effect on activity (77, 78). Thus, it is important
Figure 7. A schematic of probable distribution of hydrophilic and hydrophobic antioxidants in

bulk oil (oilair interface) and oil-in-water emulsion interface [adapted from (46)].
to use several methods to measure different products of oxidation under various

conditions, including multiphase systems, especially in evaluating natural com-
pounds as antioxidants in foods or biological systems (46).
6.2. Methods Based on Radical Scavenging Ability

Several methods have been described and used based on the fact that antioxidants
are radical scavengers in aqueous and lipid phases. The radicals employed in these
methods do not necessarily originate from lipid oxidation. In general, two
approaches are used in these methods. One involves the generation of a free radical
species and direct measurement of its inhibition caused by the test compound; such
methods do not require an actual substrate to be oxidized. The other approach uses
assay systems that involve oxidation of a substrate that is coupled with generated
free radicals, which is actually an indirect measurement. The ability of a compound
to inactivate radicals that are generated in a model system is extrapolated to its
potential as an antioxidant in lipid and lipid-containing foods. These methods are
widely and effectively used as screening and comparison tests in search for natu-
rally occurring antioxidants. They are simple and easy to use but should be care-
fully interpreted. An extensive discussion of the methods of antioxidant activity
CH2OH CH2OH
H C OH H C OH
O O
a (DPPH) + O (DPPH) : H + O
H H
HO OH O OH
CH2OH CH2OH
H C OH H C OH
O O
(DPPH) + O (DPPH) : H + O
H H
O OH O O
b (DPPH) + HO OH (DPPH) : H + HO O
(DPPH) + HO O (DPPH) : H + O O
Figure 8. Proposed reaction between a,a-diphenyl-b-picrylhydroxyl (DPPH) radical and

Oxidizable groups. (a) oxidation of conjugated group of ascorbic acid to the dehydro form and
(b) oxidation of hydroquinone [adapted from (53)].
determination that are based on free radical scavenging is found in the review by
Antolovich et al. (47).
6.2.1. DPPH Free Radical Blois (53) showed that a,a-diphenyl-b-picrylhydra-

zyl radical (DPPH) can be used for determining antioxidant activity of ascorbic
acid, tocopherol, and quinones (Figure 8a and b). DPPH in ethanol shows a strong
absorption band at 517 nm (independent of pH from 5.0 to 6.5), and the solution
appears to be deep violet in color. As the DPPH radical is scavenged by the donated
hydrogen from the antioxidant, the absorbance is diminished according to the
stoichiometry.
As DPPH radical is paramagnetic and can become a stable diamagnetic mole-
cule by accepting an electron or hydrogen radical, it can exhibit a change of its
spin resonance using electron paramagnetic resonance (EPR). If the compound in
question is able to scavenge DPPH radicals, the EPR signal of DPPH is attenuated
(Figure 9a), and this can be quantified by integration into an appropriate calculation
program. Use of DPPH radical in combination with monitoring EPR signal has
been commonly used to assess natural antioxidants (54, 7981). Although DPPH
is a comparatively stable free radical at room temperature, it is not water soluble
and the reaction mechanism between the antioxidant and DPPH radical depends on
the structural conformation of the antioxidant (82).
6.2.2. Oxygen Radicals The oxygen radical absorbing capacity (ORAC) meth-
od (66, 67) is developed based on the ability of antioxidant compounds to scavenge
Figure 9. Electron paramagnetic resonance of (a) a,a-diphenyl-b-picrylhydroxyl (DPPH) radical

and (b) hydroxyl radical as spin adduct of DMPO-OH. (A test solution composed of 50 ml of 1:10
(v/v) diluted methanolic extract of Elusine coracana was added as a radical quencher, and
splitting constant of N and H was 14.791 G.) [From (79), with permission.]
oxygen (e.g., peroxy) radicals. The peroxy free radical generated using 2,20 -
azobis(2-amidinopropane) dihydrochloride (AAPH; as the generator) in a buffered
system is targeted to damage b-phycoerythrin (b-PE, a phycobilliprotein containing
a red photoreceptor pigment) molecule. The fluorescent signal of b-PE is recorded
and interpreted as ORAC (as micromole Trolox equivalents per weight of material).
An automated system of ORAC coupled with chromatographic systems is available
for measuring total antioxidant capacity of natural products (67).
COMMONLY USED ANTIOXIDANTS IN FOODS 455
6.2.3. Hydroxyl Radicals Hydroxyl radical generated from Fenton reaction

(Equation 17) in a buffered system can be used to evaluate hydroxyl radical scaven-
ging ability of an antioxidant. Shi et al. (59) have shown that EPR may be used to
assess the hydroxyl radical scavenging ability of compounds when an appropriate
spin trapping agent (i.e., 5,5-dimethyl-1-pyrroline-N-oxide; DMPO) is used (see
Figure 9b).
Halliwell et al. (55) have described a model that uses hydroxyl radicals gener-
ated from Fenton reaction to degrade 2-deoxy-D-ribose. The decomposed products
of deoxyribose are 2-thiobarbituric acid-reactive substances (TBARS). If the anti-
oxidant present in the system scavenges hydroxyl radicals generated, deoxyribose is
protected and the amount of TBARS produced is less.
Fe2+ + H2O2 Fe3+ + OH + OH [17]
Scheme 7. Hydroxyl radical generation via Fenton reaction.
6.2.4. Superoxide Radical The scavenging ability of antioxidants for superox-

ide radical anions generated in a model reaction is also employed to assess antiox-
idant activity. Superoxide radical anions can be generated in vitro by enzymatic
(Equation 18) or nonenzymatic reactions using a xanthinexanthine oxidase sys-
tem, which requires oxygen for the reaction to generate superoxide radicals. Quan-
tification of superoxide radicals can be achieved using chemiluminescence (58)
with an appropriate agent (e.g., lucigenin) or by using the oxygen consumption
kinetics (83). The nonenzymic system uses reduction of tetrazolium salt induced
by superoxide radicals generated in a phenzine methosulfate/NADH mixture (56,
57). Meyer et al. (84) showed that scavenging of superoxide anion radical is not
necessarily effective in preventing lipid oxidation by phenolic compounds in natur-
al extracts. Also, no equilibrium can be achieved when superoxide radicals are gen-
erated continuously during the test, which is seen as a shortcoming when using
O2 -scavenging ability to assess antioxidant activity.

Xanthine oxidase
Xanthine + H2O + 2O2 2O2
- + Uric acid + 2H+ [18]
Scheme 8. Generation of superoxide radical by an enzymatic reaction.
7. COMMONLY USED ANTIOXIDANTS IN FOODS
This discussion is carried out based on the origin of the antioxidative compound:
synthetic (manufactured chemical molecules) and natural (originated from food
related material), which is widely used by the food industry. Although compounds
TABLE 6. Physical Properties of Synthetic Antioxidants Used in Foods (18, 88, 89).
Gallates
Property/Characteristic BHA BHT Dodecyl Propyl TBHQ
Appearance Waxy solid White crystals White crystals White crystals White-tan crystals
Carry through properties Very good FairGood FairGood Poor Good
Boiling point ( C) 264270 265 Decompose above 148 300
Melting point ( C) 5052 6970 146148 146148 126128
Solubility (%, w/w) in
Corn oil 30 40 0 0 510
Glycerol 1 0 25 <1
Lard 3040 50 1 510
Methyl linoleate very soluble very soluble 1 1 >10
Propylene glycol 50 0 6.5 4 30
Water 0 0 <1 <1 <1
Synergism BHT & gallates BHA BHA BHA
such as a-tocopherol and D-ascorbic acid are synthesized, they are considered as
naturally existing compounds; thus, they are considered as natural and are dis-
cussed under natural antioxidants.
7.1. Synthetic Antioxidants

Synthetic antioxidants are manmade and are used to stabilize fats, oils, and lipid-
containing foods and are mostly phenolic-based. Many compounds are active as
antioxidants, but only a few are incorporated into food because of strict safety reg-
ulations. These phenolic derivatives usually contain more than one hydroxyl or
methoxy group. Ethoxyquin is the only heterocyclic, N-containing compound
that is allowed for use in animal feeds.
Synthetic phenolic antioxidants are p-substituted, whereas the natural phenolic
compounds are mostly o-substituted. The p-substituted substances are preferred
because of their lower toxicity. The m-substituted compounds are inactive. Syn-
thetic phenolic antioxidants are always substituted with alkyl groups to improve
their solubility in fats and oils and to reduce their toxicity (24, 85, 86). The primary
mechanism of activity of these antioxidants is similar to those of primary antioxi-
dants. An antioxidant molecule reacts with a peroxy radical produced by the oxi-
dizing lipid, thus forming a hydroperoxide molecule and an antioxidant free radical.
A similar path of reaction may occur with the alkoxy free radicals formed during
the decomposition of hydroperoxides. The antioxidant free radical so formed may
be deactivated by a lipid peroxy or an alkoxy radical or with another antioxidant
radical. Dimers and even trimers of antioxidant molecules are formed because of
the reaction of antioxidant radicals, and these may have a modest antioxidant activ-
ity of their own. With the help of synergists such as ascorbic acid, some of these
original antioxidant molecules may be regenerated. Quinones are formed from phe-
nolic antioxidants by reaction with peroxy radicals. When antioxidants are present
in excess, the reaction of antioxidant free radicals with oxygen may become impor-
tant; even their reaction with polyunsaturated fatty acids has some impact on the
course of oxidation. Therefore, at high concentrations, phenolic antioxidants may
act as pro-oxidants.
In most countries, use of synthetic antioxidants is regulated and the safety of the
compounds involved has been tested based on long-term toxicity studies. The abil-
ity of an antioxidant to withstand thermal treatment (e.g., frying or baking) and to
retain sufficient stabilizing activity for the food (fried or baked) is termed as carry
through property. Table 6 provides a summary of physical properties of commonly
used synthetic antioxidants. Several researchers have studied the effectiveness of
these compounds in suppressing lipid oxidation in fats and oil, and Tables 7
and 8 provide comparative effects of synthetic antioxidants (82).
7.1.1. Butylated Hydroxyanisole (BHA) This monophenolic compound exists

as a mixture of two isomers (Figure 4), 3-tertiary-butyl-4-hydroxyanisole (90%)
and 2-tertiary-butyl-4-hydroxyanisole (10%). The 3-isomer shows a higher antiox-
idant activity than the 2-isomer. BHA is commercially available as a white, waxy
TABLE 7. Effect of Different Antioxidants on Oxidation of Soybean Oil

[Adapted from (90)].
Antioxidant Activity (as time in hours taken to reach peroxide value of 70)
Antioxidant
(at 0.02% level) 45 Ca 98 Cb
Control (no added antioxidant) 168 5

Ascorbyl acid 288 43
Ascorbyl palmitate 456 13
BHA 216 9
BHT 240 11
PG 360 14
TBHQ 544 28
a
Oxidation at 45 C as a thin layer of oil.
b
Oxidation at 98 C at AOM conditions.
flakes that is lipid soluble. BHA exhibits good antioxidant activity in animal fats as
compared to vegetable oils. It has good carry through properties but is volatile at
frying temperatures. When BHA is included into packaging materials it easily
migrates to the containing food and delays lipid oxidation (18, 74, 91).
7.1.2. Butylated Hydroxytoluene (BHT) BHT is also a monohydroxyphenol

(Figure 4) and is widely used in foods. This fat-soluble antioxidant is available
as a white crystalline compound. BHA is less stable than BHT at high temperatures
and has lower carry-through properties. BHA and BHT act synergistically, and sev-
eral commercial antioxidant formulations contain both of these antioxidants. BHT
is effectively used in oxidation retardation of animal fats. It is postulated that BHA
TABLE 8. Effect of Antioxidants and Metal Inactivators on the Oxidation of Soybean Oil
[Adapted from (87)].
Antioxidant Activity
Antioxidant or Combination (hours based on peroxide value of soy bean oil)
Controla (contain 1500 ppm tocopherol) 26

C-treatedb (contain 45 ppm tocopherol) 17
C-treated ascorbic acidc 43
C-treated BHAc 18
C-treated BHA 0.01% citric acidc 56
C-treated BHTc 23
C-treated BHT 0.01% citric acidc 53
C-treated citric acidc 50
C-treated propyl gallatec 26
C-treated a-tocopherolc 21
C-treated a-tocopherol 0.01% citric acidc 53
a
Naturally present mixed tocopherol.
b
Carbon black treated oil to remove natural tocopherol partially.
c
Antioxidant concentration is 0.57 mmol/kg.
interacts with peroxy radicals to produce a BHA phenoxy radical. This BHA phe-
noxy radical may abstract a hydrogen atom from the hydroxyl group of BHT. BHA
is regenerated by the H radical provided by BHT. The BHT radicals so formed can
react with a peroxy radical and act as a chain terminator (92, 53).
7.1.3. Ethoxyquin Ethoxyquin, 6-ethoxy-1,2-dihydro-2,2,4-trimethylquinoline

(Figure 4), is used as an antioxidant in animal feeds primarily to protect carotenoid
oxidation. It may also be used in fish products, fish oil, poultry fats, potatoes,
apples, and pears during storage and especially to protect pigment oxidation in
ground chili and paprika (93). Ethoxyquin may act as a free radical terminater.
Dimerization of the radical may inactivate the antioxidant (94).
7.1.4. Gallates Esters of gallic acid (Figure 4), namely, n-propyl, n-octyl, and
n-dodecyl gallates, are approved antioxidants for food use. Propyl gallate (PG),
the most commonly used gallate, is slightly water soluble and is available as a white
crystalline powder. PG is not suitable for use in frying oils because it is volatile at
the high temperatures of frying (18, 92). Octyl and dodecyl gallates are more lipid
soluble and heat stable and have better carry-through properties. Gallates can che-
late metal ions effectively, thus retarding lipid oxidation catalyzed by metal ions.
However, this may negatively effect the esthetic appeal of the food, because of
the dark color of the metal-gallate complexes. Therefore, gallate formulations are
always available with a metal chelator such as citric acid to prevent any discolora-
tion in the incorporated food. Gallates show synergistic activity with both primary
and some of the secondary antioxidants (94). Propyl gallate works well with BHA
and BHT because of a synergistic action, however, its use together with TBHQ is
not permitted in the United States (25).
7.1.5. Tertiary-Butylhydroquinone (TBHQ) This is a diphenolic antioxidant

(Figure 4) and is widely used in a variety of fats and oils. TBHQ has excellent
carry-through properties and is a very effective antioxidant for use in frying oils.
It is available as a beige color powder that is used alone or in combination with
BHA or BHT. TBHQ can be used in a variety of lipid-containing foods and fats
and oils. Chelating agents such as monoacyglycerols and citrates enhance the activ-
ity of TBHQ, mainly in vegetable oils and shortenings.
TBHQ reacts with peroxy radicals to form a semiquinone resonance hybrid. The
semiquinone radical intermediate may undergo different reactions to form more
stable products; they can react with one another to form dimers, dismutate, and
regenerate as semiquinones; and they can react with another peroxy radical
(Figure 5; 18, 24). A possible mutagenic effect of TBHQ has been the subject of
extensive studies, and few countries in the world including Japan and the European
Union countries do not yet approve its use in foods. Since 1999, TBHQ is included
in the class IV preservative list in Canada, with a maximum usage level of 0.02%
(Canada Food and Drug Act).
7.2. Natural Antioxidants

Use of plant parts (bark, leaves, seeds, etc.) and their extracts to preserve food from
developing a rancid taste is a practice that has continued since prehistoric time.
There is evidence that even for the industrial materials, plant-based components
were used as antidrying agents to prevent oxidation and polymerization of polyun-
saturated fatty acid-rich plant oils (2, 5, 48). During the past two decades, intensive
research has been carried out on naturally occurring antioxidative compounds from
different sources. The main drive behind this search was to reduce the use of syn-
thetic compounds as food additives because of their potential negative health effects
and as a result of consumer demand.
Plant-based components have increasingly been advocated as safe and natural
antioxidants considering their existence in regular foods that are consumed. Much
of the interest on naturally occurring antioxidants is developed because of the trend
to minimize or avoid the use of synthetic food additives. Continuous effort in
searching for naturally occurring antioxidative compounds during the past 20 years
has helped to develop efficient models for activity screening, structure function
relationship assessment, categorizing sources of antioxidant groups, developing
methods of isolating purified antioxidative compounds from natural sources, and
developing branded foods (e.g., claims for marketing purpose). Again, one has to
keep in mind that the safe it is natural is based on the fact that these secondary
metabolites are present in small concentrations in regular foods. If these are to be
added to foods that are largely consumed, they should undergo all safety clearances.
There are many naturally occuring compounds that act as antioxidants in fats and
lipid-containing foods. Among these, only a few are currently approved and
employed in foods. Groups of compounds that are found naturally and exhibit
strong antioxidant activity are discussed here.
7.2.1. Ascorbic Acid and Ascorbic Acid Esters and Salts Vitamin C or
ascorbic acid is widespread in nature but sparingly associated with fats of oils
because of its hydrophilic nature (95). Ascorbic acid in the free form, salts of
sodium and calcium, and esters of stearic and palmitic are commonly used as anti-
oxidants in foods. Erythorbic acid is the D-isomer of naturally present L-ascorbic
acid (Figure 10) and is often used as an antioxidant in dried fruits and a cure
O
CH2OH CH2OH CH2OC(CH2)14CH3
H C OH HO C H HO C H
O O O
O O O
H H H
HO OH HO OH HO OH
L-Ascorbic acid D-Isoascorbic acid Ascorbyl palmitate
(Vitamine C) (Erythrobic acid)
Figure 10. Chemical structures of L-ascorbic acid, erythorbic acid, and ascorbyl palmitate.
Figure 11. Regeneration path of tocopherol by ascorbic acid and during participation in radical
scavenging in biological systems [adapted from (96)].
accelerator in cured meat. Unlike ascorbic acid, erythorbic acid is not a natural con-
stituent of foods and has minimal vitamin C activity. Similar to ascorbic acid,
erythorbic acid is highly water-soluble but remains insoluble in oils (97).
In foods, water-soluble ascorbic acid acts as a secondary antioxidant and parti-
cipates in various antioxidative and related functions. Ascorbic acid is capable of
quenching various forms of oxygen (singlet oxygen, hydroxyl radicals, and super-
oxide). When ascorbic acid acts as a hydrogen donor, ascorbyl radical so produced
may reduce or terminate radical reactions; hydroperoxides may then be converted
into stable products. Ascorbic acid can reduce primary antioxidant radicals and thus
act as a synergist. A very good example is donating a hydrogen atom to tocopheryl
radical and then regeneration of tocopherol (Figure 11), which is commonly
observed in the biological systems. In addition to that, ascorbic acid can shift the
redox potential of food systems to the reducing range and can act synergistically
with chelators and regenerate primary antioxidants other than tocopherols (98).
In vivo ascorbic acid acts as a primary antioxidant and in tissues it is essential for
the prevention of oxidative cellular damage by hydrogen peroxide (99). In a solu-
tion, ascorbic acid readily oxidizes to dehydroascorbic acid. This formation occurs
through one- or two-electron transfer that is due to its enediol structure. It is
a reductone and has a very high affinity for oxygen. The 2- and 3-positions of
ascorbic acid are unsubstituted. Oxidation happens via the intermediate semi-
dehydroascorbic acid or monodehydroascorbic acid or ascorbate free radical. The
semidehydroascorbic acid is either reduced to give ascorbic acid again or oxidized
to give dehydroascorbic acid. In nature, these compounds complete a redox system
(Figure 12). The redox cycle is completed in living tissues by enzymatic reduction
of dehydroascorbic acid to ascorbic acid. Seib (101) has reviewed the oxidation and
other reactions of ascorbic acid.
Ascorbyl palmitate and ascorbyl stearate are synthetic derivatives of ascorbic
acid. Ascorbic palmitate is soluble in lipid-containing foods because of its relatively
good hydrophobicity (88). Ascorbyl palmitate is hydrolyzed by the digestive sys-
tem to provide nutritionally available ascorbic acid and palmitic acid, but health
claims cannot be made for its vitamin C contribution.
As an antioxidant, ascorbic acid is very attractive as it carries GRAS (generally
recognized as safe) status with no usage limits; it is a natural or nature-identical
2 (L-ascorbate)
a e
2 (semidehydroascorbic acid)
e +e
b
c
Dehydroascorbic acid 2 (L-ascorbic acid)
+2e
CH2OH
H C OH O
O e O
a O HO
+e
H O
OH O
O OH
Semidehydro ascorbic acid
(ascorbate free radical or
monodehydro ascorbic acid)
b 2 (Semidehydro ascorbic acid) dimer
+ H2O
CH2OH
H C OH
O O
O O
HO HO +
H
O OH HO OH
OH
Dehydroascorbic acid L-Ascorbic acid
+2e
c Dehydroascorbic acid L-Ascorbic acid
NADPH 2 NADP (in vivo)

2H2SO3 2H2SO4 (in vitro)
Figure 12. Ascorbic acid-dehydroascorbic acid redox system (a) oxidation of ascorbate to
semidehydroascorbic acid, (b) disproportionation of semidehydroascorbic acid, and (c) reduction
of dehydroascorbic acid [From (100), with permission].
product and is highly recognized as an antioxidant among the nutrient category.

Ascorbic acid is also used as an acidulant and flavorant. Heat treatment makes
ascorbic acid unstable, and it participates in nonenzymic browning reactions and
degrades through reductone formation. Ascorbic acid and its salts (Na- and Ca-
ascorbate) are water soluble and are not applicable as antioxidants in oils and
fats. These salts are used extensively for stabilizing beverages that contain oxidiz-
able substrates. In fats, ascorbyl palmitate is used mostly because of its superior
solubility. Ascorbyl palmitate also has GRAS status, and there are no restrictions
for its usage level.
7.2.2. Carotenoids Carotenoids are ubiquitously found lipid-soluble-colored

compounds, mainly from green plants, fruits, and vegetables. The two classes of
carotenoids, carotenes and xanthophylls, are composed of 40-carbon isoprenoid
or tetraterpenes with varying structural characteristics. Carotenes are polyene
hydrocarbons and vary in their degree of unsaturation (e.g., b-carotene, lycopene;
Figure 13). Xanthophylls are derived from carotenes by hydroxylation and epoxi-
dation, thus containing oxygen (e.g., astaxanthene, canthaxanthin). Some carote-
noids exhibit biological activity of vitamin A and hence are categorized as
provitamin A. b-Carotene is the most abundantly found provitamin A. Many fats
and oils, especially those from plant sources, contain b-carotene, and it contributes
to the deep intense orange red color of many oils (94, 102).
Carotenoids can act as primary antioxidants by trapping free radicals or as sec-
ondary antioxidants by quenching singlet oxygen. In foods, carotenoids usually act
as a secondary antioxidant; however, at low oxygen partial pressure (<150 mm Hg,
in the absence of singlet oxygen), carotenoids may trap free radicals and act as a
chain-breaking antioxidant (95, 104). At high oxygen concentrations, the antioxi-
dant activity of b-carotene is diminished. Increased oxygen concentration leads to
the formation of carotenoid peroxy radicals because the conjugated double bonds of
carotenoid molecules are very susceptible to the attack by peroxide radicals. This
reaction favors autoxidation of b-carotene over inactivation of lipid peroxy radicals
carotene, , carotene
Lycopene, , carotene
OH
HO
Lutein, , carotene-3,3-diol
Figure 13. Chemical structures of some carotenoids.

(103). At low oxygen concentrations, the lifetime of the carotenoid radical is long
enough to permit its reaction with another peroxy radical and to form nonradical
species. The unsaturated structure of b-carotene allows the molecule to delocalize
electrons in the radical and to produce a resonance-stabilized product with the per-
oxy radical. This carotene radical participates in termination reactions and converts
peroxy radicals to less damaging products. Lieber (104) has provided details of
antioxidative reactions of carotenoids. The combination of carotenoids and toco-
pherols results in synergistic action (105, 106). The stability of carotenoids is
affected by oxygen, heat, pH, light, and metals; therefore, care should be taken
in their handling as antioxidants.
Singlet oxygen, which is unstable, preferentially transfers energy to b-carotene
to produce triplet state b-carotene. This occurs through an exchange electron trans-
fer mechanism. Triplet state b-carotene releases energy in the form of heat, and the
carotenoid is returned to its normal energy state. This mechanism allows the caro-
tenoid molecule to be an effective quencher of numerous molecules of singlet oxy-
gen (102). The ability of carotenoids to quench singlet oxygen is related to the
number of carbon double bonds in their chemical structures. Carotenoids with
nine or more conjugated double bonds are very effective antioxidants. Because
of the presence of additional functional groups in the hydrocarbon structure,
xanthophylls cannot perform as effectively as antioxidants (94).
7.2.3. Tocopherols and Tocotrienols Tocopherols and tocotrienols are the

natural antioxidative compounds found widely in different tissue, even if it is in
trace amounts. Tocopherols are found abundantly in vegetable oil-derived foods.
These monophenolic compounds possess varying antioxidant activities. Tocopherols
and tocotrienols comprise the group of chromanol homologs that exert vitamin E
activity in the diet. These different homologs vary in the extent of methylation of
the chromane ring. The a-, b-, g-, and d-tocopherols contain a saturated phytyl (tri-
methyltridecyl) side chain (Figure 14). The corresponding tocotrienols have unsa-
turated phytyl chain at the 3-, 7-, and 11-positions. Only RRR isomers are found
naturally. Synthetic a-tocopherol (all rac-a-tocopherol) is a combination of eight
sterioisomers that are found in equal amounts in the mixture. Biologically, RRR-
a-tocopherol is the most active vitamin E homolog. The antioxidant activity of
a-, b-, g-, and d-tocopherols and tocotrienols decreases in the order of a >
b > g > d in vivo (107, 108) and a > b > g > a in bulk oils and fats (108).
Vitamin E activity of tocopherols decreases in the order of a > b > g > d (107).
Antioxidant activity of tocopherols is mainly by scavenging peroxy radicals
thus interrupting chain propagation (95), which is based on the tocopherol
tocopherylquinone redox system. The active configuration is the phenolic group
in the benzene ring, located at the para position to the oxygen atom bound next
to the dihydropyrone cycle. Alpha-tocopherol donates a hydrogen atom to a peroxy
radical resulting in an a-tocopheryl semiquinone radical (Figure 15). This radical
may further donate another hydrogen to produce methyltocopherylquinone or react
with another tocopheryl semiquinone radical to produce an a-tocopherol dimer. The
methyltocopherylquinone is unstable and will yield a-tocopherylquinone (100).
R1
HO
H3C H3C CH3
R2 O CH3
CH3
R3
Tocopherol
R1
HO
CH3 CH3 CH3
R2 O CH3
CH3
R3
Tocotrienol
Tocopherol or Tocotrienol R1 R2 R3
CH3 CH3 CH3
CH3 H CH3
H CH3 CH3
H H CH3
Figure 14. Chemical structures of tocopherols and tocotrienols.
The a-tocopheryl dimer continues to possess antioxidant activity. Also, two toco-
pheryl semiquinone radicals can form one tocopheryl quinone molecule and regen-
erate one tocopherol molecule. The decomposition products of tocopherols (during
thermal oxidation) can slowly oxidize and release tocopherol that can act as an anti-
oxidant (100).
Commercially, tocopherol is available as a pure all-rac-a-tocopherol, mixed
tocopherols having various contents of a-, b-, g-, or d-tocopherols (diluted in vege-
table oil) and synergistic mixtures containing tocopherols, ascorbyl palmitate or
other antioxidants, and synergists such as lecithin, citric acid, and carriers. Extrac-
tion of tocopherols from natural sources and chemical synthesis of tocopherols are
well described by Schuler (100).
Tocopherols are considered as natural antioxidants for lipid-containing foods
and marketed as all natural. They are permitted in food application according
to GMP regulations (21CFR 182.3890). Natural tocopherols are limited to 0.03%
(300 ppm) in animal fats (9 CFR 318.7). As most vegetable oils naturally contain
tocopherols, the addition of this antioxidant may pose prooxidant effects.
7.2.4. Other Phenolic Antioxidative Compounds from Plants Higher plants

are rich in a myriad of phenolic compounds in their secondary metabolite pool.
Among these, phenolic acids and polyphenolic derivatives are found to be the
most important series of hydrophilichydrophobic antioxidative compounds natu-
rally present. In foods, these polyphenolic compounds act as radical scavengers
HO
O C16H33
-Tocopherol
ROO
ROOH
O C16H33
-Tocopheryl semiquinone radical

O
ROO
O C16H33
ROOH
O C16H33
O
HO
CH2
O C16H33 -Tocopherol dimer
CH2
Methyl tocopherylquinone
O C16H33
e
OH
C16H33
O
CH3
-Tocopherylquinone
(Stable)
Figure 15. Possible mechanism of participation of a-tocopherol in free radical scavenging.
or metal chelators, and some may play a multifunctional role. Numerous plants and
their parts have been identified as sources of phenolic acids, flavonoids, and related
compounds. Antioxidant activity of phenolic compounds from various plant sources
has been reported in several peer-reviewed papers, reviews, and books. Interested
readers are referred to reviews by Shahidi et al. (18, 109112) about the sources of
natural phenolics possessing antioxidative activity that are obtained from plants and
have potential applications in oils and fats (as discused below).
7.2.4.1. Antioxidants from Cereals, Oilseeds, and Related Sources Seeds rich
in oils are also abundant sources of various types of antioxidative compounds.
Among these carotenoids, phenolic acids, and their derivatives, flavonoids, phytic
acid, lignans, and tocopherols are predominantly found depending on the plant gen-
era and species. Reviews by Wanasundara et al. (110) and Shukla et al. (111) dis-
cuss antioxidants of oilseeds and their products in detail.
7.2.4.1.1. PHENOLIC ACIDS AND THEIR DERIVATIVES Phenolic acids are found in plants
and have the basic chemical structure of C6-C1 (benzoic acids) and C6-C3 (cinnamic
acids) (Figure 16). A range of substituted benzoic acid or cinnamic acid derivatives
comprise these two major families of phenolic acids that are found in plants. Both
of these families occur as free, conjugated, or esterified form, and sometimes as
depsides such as chlorogenic acid (3-caffeoyl-quinic acid). Phenolic acids serve
as free radical acceptors and chain breakers. The presence of a phenolic ring in
the molecular structure and side chains facilitates the radical accepting ability of
phenolic acids. According to Chimi et al. (113) and Pokorny (114), monohydroxy
phenolic acids are less efficient as antioxidants than polyhydroxy phenolic acids.
The presence of the CH CH-COOH group in cinnamic acids ensures a greater
antioxidative activity than a
COOH group as in benzoic acids. The participation
of the CC bond is important in stabilizing the antioxidant radical by resonance.
The antioxidant activity of phenolic acids and their esters depends on the num-
ber of hydroxyl groups in the molecule, and this would be strengthened by steric
COOH R1 R2 R3
H H H Benzoic
H OH H p-Hydroxybenzoic
OH OH OH Gallic
R3 R1
OH OH OH Procatechuic
R2 OCH3 OH OCH3 Syringic
OCH3 OH H Vanillic
Benzoic acids
COOH
R1 R2 R3
OH OH H Caffeic
H H H Cinnamic
H OH H p-Coumaric
R3 R1 OCH3 OH H Ferulic
R2 OCH3 OH OCH3 Sinapic
Cinnamic acids
Figure 16. Chemical structures of antioxidative phenolic acids.
hindrance. Hydroxylated cinnamic acids are more effective antioxidants than their
benzoic acid counterparts. When the acid group is esterified with a bulky group
such as a sugar, the antioxidant potency of the molecule is further enhanced
(115). Introduction of a second hydroxyl group in the ortho or para position
increases the antioxidant activity of hydroxylated phenolic acids. Therefore, acids
with ortho diphenolic groups (caffeic and protocatechuic acids) are more efficient
antioxidants than their respective monophenolic acids ( p-hydrobenzoic and p-
coumaric acids). Gallic acid that has three hydroxyl groups is more active than pro-
teocatechuic acid, but more than three hydroxyl groups in the structure does not
appear to improve the antioxidant efficiency in oil systems (114). The aromatic
ring substituted with two or three phenolic groups in the ortho position are particu-
larly important; some hydroxyl groups may be methoxylated. Substitution of one or
two methoxy groups at the ortho position relative to the hydroxyl group markedly
increases the antioxidant activity of phenolic acids. Therefore, sinapic acid is a
more efficient antioxidant than ferulic acid, which is more efficient than p-coumaric
acid. For the same reason, syringic acid is more active than vanillic acid and p-
hydroxybenzoic acid (115). Ortho substitution of the phenolic acid with electron
donor alkyl or methoxy groups increases the stability of aryloxy radical and thus
the antioxidant activity. Methoxy substitution strengthens the antioxidant activity
than the addition of a hydroxyl group to the molecule (113, 114).
7.2.4.1.2. LIGNANS Lignans are compounds with great chemical diversity and
found in all parts of the plants. They are dimers of phenyl proponoid (C6-C3) units
linked by the central carbons of their side chains (116). Among these, bisepoxy lig-
nans and cyclolignans that occur in oilseeds (sesame and flax) exhibit strong anti-
oxidative activity in aqueous and lipid media. Lignans of sesame (Sesamum
indicum L.) seed include sesamin, sesamolin, sesaminol, and sesamol, which act
as endogenous antioxidants for the oils (117119). Sesamolin may undergo chemi-
cal changes during thermal treatment and under processing conditions (e.g., bleach-
ing) and forms sesaminol and sesamolinol (120122). High oxidative stability of
sesame oil obtained from roasted seed may be largely attributed to the presence
of lignan compounds (123125).
Lignans of flaxseed exist as secoisolariciresinol diglucoside (SDG; 126). SDG is
a potent antioxidant in biological systems because of its tendency to associate in the
aqueous phase. Much of the work on antioxidant activity of SDG is related to its
radical mediated disease prevention (127, 128). Lignans of both sesame (121, 122)
and flax (129, 130) have shown hydrogen-donating ability and scavenging activity
for various free radicals.
7.2.4.1.3. STEROLS Phytosterols are mostly associated with unrefined vegetable

oils and exist as derivatives of phenolic acids (e.g., ferulic acid). Several studies
are available on antioxidant activity of sterols and their derivatives from sources
such as corn fiber, oats, and rice. These compounds can be obtained from the unsa-
ponifiable fraction that is removed during vegetable oil refining. Triterpene alcohols
and hydrocarbons (131), or sterols (Figure 20) from oats (132, 133), rice (134, 135),
and corn fiber (135, 136), were able to exert the antioxidative effect on frying oils
and as a result displayed antipolymerizing effects. It has been suggested that dona-
tion of a hydrogen atom from the allylic methyl group in the side chain of sterols
followed by isomerization to a relatively stable tertiary allylic free radical may
be the mechanism for sterol antioxidant activity (132). The ethylidine group
(CH3 CH ) in the side chain of the sterol molecule seems essential for performing
the antipolymerizing effect on the frying oils at high temperatures (137, 138). The
g-oryzanol of rice bran performed good antioxidant activity in a linoleic acid model
system at 37 C as opposed to frying temperatures. g-Oryzanol of rice is composed
of at least ten compounds that are mainly ferulic acid derivatives of triterpene alco-
hols, stigmasterol, campesterol, sitosterol, and cycloartinol (Figure 20; 137). Plant
sterol-based compounds are available as physiological antioxidants to prevent cer-
tain disease conditions. Commercial preparations of sterolbased natural antioxi-
dants for high-temperature food applications are not abundantly available yet.
7.2.4.2. Antioxidants from Labiatae Herbs Among phenolic diterpenes, carno-

sic acid and carnosine are the very active antioxidants that are commercially avail-
able as natural antioxidants for lipid-containing foods. Chemically, carnosic acid
has a structure consisting of three six-membered rings, including a dihydric poly-
phenolic ring and a free carboxylic acid, and carnosol is a derivative of carnosic
acid containing a lactone ring (Figure 17). These active antioxidants are found espe-
cially in plants of the Labiatae family (oregano, rosemary, sage, and thyme). Com-
mercial extracts of these plants are produced by organic solvent extraction of the
plant parts and subsequent deodorization and bleaching. Commercially these anti-
oxidants are available as powder, paste, or liquid and formulations in propylene gly-
col, medium-chain triacylglycerols, or vegetable oils. Schuler (100) has discussed
preparation of rosemary antioxidants and their effectiveness in oils compared with
other antioxidants.
Extensive studies (139, 140) on rosemary extracts containing carnosol, carnosic
acid, and rosmarinic acid have shown that the activities of these natural antioxidants
are system-dependent and that their effectiveness in different food systems is diffi-
cult to predict. In bulk vegetable oils (corn, soybean and peanut) and fish oils, car-
nosol and carnosic acid are effective antioxidants. It has been hypothesized that this
OH
OH OH COOH
HO HO
O O OH
HOOC C
O
O
H H
H 3C CH3 H3C CH3
Carnosic acid Carnosol Rosemarinic acid

Figure 17. Chemical structures of carnosine, canosic acid, and rosemarinic acid.
behavior of rosemary antioxidants is attributed to the partitioning differences in the

biphasic system. These hydrophilic antioxidants are oriented in the oilair interface
and protect the bulk oil phase from oxidation. However, in oil-in-water emulsions,
they are less effective in the oxidation of the oilwater interface, where most of the
oxidation reactions take place (see illustration in Figure 7). It has been hypothe-
sized that the hydrophilic antioxidants (e.g., rosmarininc acid, gallic acid, cate-
chins, propyl gallate) partitioned more (>90%) into the aqueous phase, thus
allowing a considerably low concentration of antioxidants in the oil phase, causing
less antioxidative protection (141). Carnosic acid and carnosol are oxidized during
oxidation at 60 C and higher temperatures; however, antioxidant activities are
maintained; apparently the oxidation products are active antioxidants at high tem-
peratures. These compounds have good carry-through properties and protect frying
oils and fried foods.
7.2.4.3. Flavonoids from Green Tea Among the natural flavonoids studied as
antioxidants for lipid-containing foods, polyphenolic catechins from green tea
(Camellia sinensis L) have been extensively scrutinized (142, 143). Extracts of
immature leaves of the plant (green tea) are rich in flavan-3-ols and their gallic
acid derivatives, namely, ()-catechin, ()-epicatechin, ()-gallocatechin, ()-
epicatechin gallate, ()-epigallocatechin, and epigallocatechin gallate (Figure 18).
Flavonoids are a heterogeneous group of phenolic compounds having a benzo-g-
pyrone structure (Figure 18) in the molecule and occur ubiquitously in plants.
Approximately 90% of flavonoids occur in plants in the glycosidic form (143, 144).
Antioxidant activity of flavonoids is bimodal, and they are very effective in counter-
acting lipid oxidation; however, this is very much dependent on the chemical and
physical properties of the system. Flavonoids function as primary antioxidants in
systems when metal catalyzed oxidation is not present. Because of their lower
redox potentials (230 < E0 < 750 mV), flavonoids (145) are thermodynamically
able to reduce highly oxidizing free radicals with redox potential in the range of
23101000 mV, such as alkoxy, hydroxyl, peroxy, and superoxide (See Table 1,
21) radicals by hydrogen donation. Flavonids can form resonance-stabilized radi-
cals while scavenging oxidative free radicals (18). For a molecule that has 30 , 40 -
dihydroxylation donation of one H atom to a free radical may produce a flavonoid
aryloxy radical. This flavonoid aryloxy radical may react with a second radical and
acquire a stable quinone structure (Figure 19a; 145, 146). At the same time, flavo-
noids are good metal chelators that can be used for inhibition of metal-catalyzed
oxidation initiation. Metal chelation ability of flavonoids is caused by the ortho-
diphenol structure in rings A and B (3-hydroxy-4-keto group or the 5-hydroxy-4-
keto) and ketol structure in ring C (Figure 19b). An ortho quinol group at the B ring
has also demonstrated metal ion chelating activity (18, 147).
The position and the degree of hydroxylation (especially at the A and B rings)
are of primary importance in determining the antioxidant activity of flavonoids.
Dihydroxylation at ortho position of the B ring contributes to antioxidant activity;
however, para and meta hydroxylation of the B ring do not occur naturally. All fla-
vonoids with 30 ,40 -dihydroxy configuration possess antioxidant activity (18, 147).
Hydroxylation/substitution
3 3 5 7 3 4 5
2 4 Flavone
8 B
9 O 5 H OH OH H H H Chrysin
7
2 6 H OH OH H OH H Apigenin
A C OH OH OH OH H Luteolin
6 3 H
5
10 H(OH) Flavonol
O OH OH OH OH OH H Quercetin
OH OH OH OH OH OH Myricetin
Flavone (Flavonol) OH OH OH H OH H Kaempferol
O-rutinose OH OH OH OH H Rutin
3 3 5 7 3 4 5
2 4 OH OH OH H OH H Pelargonidin
8 B
9 O 5 OH OH OH OH OH H Cyanidin
7
2 6 OH OH OH OCH3 OH H Peonidin
A C OH OH OH OH OH Delphinidin
6 OH
5
10 OH(O-glyc) OH OH OH OCH3 OH OH Petunidin
OH OH OH OCH3 OH OCH3 Malvidin
Anthocyanidin (Anthocyanin)
3 Absolute
2 4 3 5 7 3 4 5 configuration
8 B
9 O 5 OH OH OH OH OH H 2R:2S (+)-Catechin
7
2 6 OH OH OH OH OH H 2R:3R ()-Epicatechin
A C 3
6 OH OH OH OH OH OH 2R:3S (+)-Gallocatechin
5
10 OH OH OH OH OH OH OH 2R:3R ()-Epigallocatechin
-Gallic OH OH OH OH H 2R:3R ()-Epicatechin
Flavan-3-ols acid gallate
-Gallic OH OH OH OH OH 2R:3R ()-Epigallocatechin
acid gallate
3
3 5 7 3 4 5
2 4
8 B Flavonone
7
9 O 5
H OH OH H OCH3 H Hesperetin
2 6
A C H OH OH H OH H Naringenin
6 3
5
10 H(OH)
Flavononol
O OH OH OH OH OH H Taxifolin
Flavonone (Flavanonol)
Figure 18. Chemical structures of flavonoids according to their hydroxylation.
Other important features include carbonyl group at position 4 and a free hydroxyl
group at position 3 or 5 (91). Among aglycones, the presence of a free 3-hydroxyl
group in the C ring is a requirement for maximal radical scavenging activity of fla-
vonoids (148, 149). When a disaccharide is glycosylated to a flavonoid (e.g., rutin),
(a) OH O
OH OH
R RH
O O
ROO ROOH
O O
3,4-Dihydroxyflavonoid Aroxyl radical
ROO R
ROOH RH
O O
O O
O O
O O
Stable quinone
(b) HO Mn+
OH
OH
HO O
Mn+
Mn+
Figure 19. (a) Free radical scavenging by flavonoids. (b) Binding sites of flavonoids for metal
ions.
the substituent at position 3 becomes a poorer leaving group; thus, the molecule
becomes less oxidizable and exhibits a lower antioxidant activity in free fatty
acid systems than monosaccharide glycosides (e.g., quercetin; 149). Many of the
flavonoids and related substances display a significant antioxidant behavior in
lipid-aqueous and lipidlipid food systems (147). Most of these compounds have
very low solubility in the lipid phase, and it is a serious disadvantage if the aqueous
phase is present to a in considerable extent in the food.
The antioxidant activity of these catechins and the derivatives showed a marked
difference depending on the substrate used for evaluation. In bulk corn oil that was
oxidized at 50 C epigallocatechin, epigallocatechin gallate and epicatechin gallate
exhibited better antioxidant activity than epicatechin or catechin. These catechins
have been very effective in retarding oxidation of polyunsturated fatty acids-rich
5Avenasterol (a) Sitosterol (a)
O
H3CO
O
HO
Campesteryl ferulate (b)
O
H3CO
O
HO
Cycloarteryl ferulate (b)
H3CO
O
HO
Stigmasteryl ferulate (b)
Figure 20. Chemical structure of antioxidative sterols identified from (a) oats, and (b) rice and
corn fiber.
vegetable, and marine oils (150152). In the oil-in-water emulsions, all catechins
tested were pro-oxidants; however, in the liposomes comprising lecithin, epigallo-
catechin gallate was the best antioxidant, followed by epicatechin, epigallocatechin,
epicatechin gallate, and catechin (78, 152). When tea catechins were added to noo-
dles and to the frying oils, they were able to improve the oxidative stability of the
fried product and the oil used for frying (153). In addition to that, tea polyphenols
exhibited protecting ability against b-carotene oxidation; i.e., tea catechins were
able to exert an antidiscoloring effect on beverages containing b-carotene that
were UV-light irradiated (153).
The antioxidant activity of individual tea polyphenols in different model assays
showed a proportional relationship to the number of hydrogen radical donors of
catechins. A synergistic effect was observed between tea catechins and caffeine,
ascorbic, citric, malic, and tartaric acids and tocopherols (153). Formation of oxi-
dation products of ()-catechin during the antioxidative process has been observed
in oxidation model studies. According to the proposed mechanism, ()-catechin
can scavenge four radicals per molecule (154, 155). Yamamoto et al. (156) have
summarized the chemistry and application aspects of green tea, especially in rela-
tion to using their catechins.
8. ESTIMATION AND ANALYSIS OF ANTIOXIDANTS IN FOODS
Qualitative and quantitative detection of antioxidants and potential antioxidative

compounds is of utmost importance to researchers and industry and regulatory
agencies. Numerous methods based on colorimetry, spectrometry, fluorometry, vol-
tammetry, polarography, thin-layer chromatography, paper chromatography, gel
permeation chromatography, gas chromatography, and high-performance liquid
chromatography have been described for both natural and sysnthetic compounds
for their antioxidant activity in foods (157). Almost all of these procedures require
considerable sample preparation and estimation of individual antioxidants. A con-
siderable number of procedures have been developed and tested in collaborative
studies for determination of commonly used antioxidants, which are strictly regu-
lated for their use. Kochhar and Rossell (157) have provided an elaborative discus-
sion about the methods used to determine BHA, BHT, gallates, tocopherols, and
TBHQ. Official methods have been developed to determine phenolic antioxidants
under the category of food additives (AOAC method 983.15; 158).
9. TECHNOLOGICAL CONSIDERATIONS
IN USING ANTIOXIDANTS
The type of food to which antioxidants may be added is variable and ranges from
baked goods, biscuits, chewing gum, dry snacks, fruit drinks, mayonnaise, meat pro-
ducts, nuts, and oils and fats, among others. For food applications, the antioxidants
must be effective at low concentrations (below 0.02%, w/w) because at high concen-
trations, they may act as pro-oxidants. The antioxidants should also be nontoxic.
REGULATORY STATUS AND SAFETY ISSUES 475
Usually the antioxidant is directly added to the food as a concentrate in lipid/oil,

dissolved in a food grade solvent, or in an emulsified form that may be sprayed onto
the food product. Antioxidants must be thoroughly blended with the lipid to obtain
their maximum potency. To be effective, antioxidant(s) should partition between
oilair interfaces in bulk oil systems or between oilwater interfaces in the emul-
sion systems. Antioxidants must be added, as soon as possible, to the fresh product
as they cannot reverse any oxidation reactions that have already occurred (157).
Metal deactivators such as citric acid are added to vegetable oils after heating,
during the cooling stage of deodorization step of vegetable oils (19). An effective
antioxidant should be stable under processing conditions, especially at high tem-
peratures, and possess a good carry through property.
Antioxidants that are added to fats and oils are usually in the form of liquid for-
mulations. The major considerations of devising antioxidant formulations include
the following:
1. Antioxidants with different degrees of potency are formulated in an anti-

oxidant combination/mixture.
2. Better control and accuracy in applying of antioxidants should be achieved
with the mixture or formulation.
3. Ability to use synergistic effect of antioxidants to enhance their activity.
4. Complete distribution or solvation of antioxidants in fats and oils.
5. Prevent or minimize discoloration associated with specific antioxidants.
6. Ease of handling.
Most synthetic antioxidants are formulated with polypropylene glycol, glyceryl

monooleate, mono- and diacylglycerols, or vegetable oils as carriers to enhance
their solubility or dispersibility in foods. Several synergistic mixtures are available
commercially, especially citric acid with synthetic antioxidants. Commercial
preparations of natural antioxidants are predominantly tocopherol- and ascorbic
acid ester-based. Few formulations are available with rosemary, sage, and tea
chatechin-based antioxidative ingredients. In these formulations, vegetable oils
and starches are used as carriers and citric acid is also included. Elliot (159) has
extensively discussed the technological considerations when using ascorbic acid,
b-carotene, and tocopherols as antioxidants in lipid-containing foods. These antiox-
idants should be handled differently than the synthetic antioxidants because of their
reactive nature.
10. REGULATORY STATUS AND SAFETY ISSUES

OF SYNTHETIC AND NATURAL ANTIOXIDANTS
All synthetic antioxidants are generally categorized under direct food additives.
They are subjected to careful scrutiny and complex toxicological studies for
approval. However, the usage and approval of an antioxidant may differ from
one country to another. Many countries have adopted regulations similar to the Uni-
ted States regarding the usage of these antioxidants; however, significant differ-
ences exist among different countries on their type, application, and usage levels.
Table 9 provides a summary of regulations governing the use of synthetic antiox-
idants in Canada and the United.
In Canada, the use of antioxidants is regulated under the Food and Drug Act
(Heath Canada), and in United States, it is regulated by the Federal Drug Admin-
istration and the U.S. Department of Agriculture. When it comes to the European
Economic Community, directives regulate the use of antioxidants; however, indivi-
dual member countries still have the control of usage levels. In Japan, the Food
Sanitation Law specifies the use of antioxidants (48).
Antioxidative compounds naturally present in food are not covered under present
regulations; obviously, it is not a controlled substance as it is part of the raw mate-
rial of food processing. However, if an antioxidative compound isolated from a
natural source is to be added to food, the compound should comply with the appro-
priate regulations and safety clearances.
11. SAFETY CONSIDERATIONS OF ANTIOXIDANTS

USED IN FOODS
The safety of food additives is always a controversial discussion because of their

possible toxic effects during long-term intake. It has become clear that antioxidants
may share a number of toxic properties at high doses. However, it is logical to con-
sider using antioxidants at low levels to reduce the deleterious effects of consuming
lipid oxidation products that may be produced if no antioxidants are used. The
intake amount of each antioxidant is different with the food and the dietary habits.
The use of antioxidants in different countries is limited by specific regulations,
established on the basis of their safety for use and technological need. Many coun-
tries follow the recommendations of the Joint FAO/WHO Expert Committee on
Food Additives and Contaminants (JECFA) on the safe use of food additives.
The safety evaluations produced by international bodies such as the Joint FAO/
WHO expert committee on food additives (JECFA) and the European Commission
(EC) scientific committee for food additive (SCF) are used to establish acceptable
daily intake (ADI). ADI is defined as the average amount of the substance that can
be consumed daily for a lifetime without health hazards and expressed on the basis
of bodyweight. In determining ADI, a range of toxicity tests is carried out. From
these tests, the effect that is most sensitive is studied to ascertain the maximum dose
at which that effect is no longer observed (no-effect level). A reduction or a safety
factor is used to refine the no-effect level, considering the possible difference in
sensitivity between species (animals and human) and individuals. To ensure there
is an adequate margin of safety for consuming groups, an arbitrary safety factor of
100 is normally used (159). A factor lower than 100 is used if ADI is based on
human toxicity study data. For some compounds, ADI is not specified. Accord-
ing to the Joint FAO/WHO Expert committee recommendations, this is specified
TABLE 9. Regulations Governing the Use of Synthetic Antioxidants in Canada and the United States.
Canadaa United Statesb
Item Number and Permitted in

Compound or upon Maximum level of use Citation and Permitted in Maximum level of use
Ascorbic acid Class IV, A.1 21 CFR 182.3013 GRAS with GMP
In fats and oils, monoglycerides GMP
and diglyceride, shortenings,
Unstandardized foods
Ascorbyl palmitate Class IV, A.2 21 CFR 182.3149 GRAS with GMP
In fats and oils, lard, GMP
monoglycerides and
diglyceride, Shortenings
Unstandardized foods except GMP
meat and meat byproducts, fish,
poultry meat and its byproducts
Margarine Not to exceed 0.02% of the fat
content, alone or in combination
with ascorbyl stearate
Ascorbic stearate Class IV, A.3
and diglyceride, Shortenings
with ascorbyl palmitate
BHA Class IV, B.1 21 CFR 172.110 Alone or in combination
with BHT
In fats and oils, shortenings, Not to exceed 0.02%, alone or in Dehydrated potato shreds 50 ppm
margarine combination with BHT, PG or
TBHQ
Dried breakfast cereals, dehy- Not to exceed 0.005%, alone or in Active dry yeast 1,000 ppm
drated potato products combination with BHT or PG
TABLE 9. (Continued)
Chewing gum Not to exceed 0.02%, alone or in Beverages and desserts 2 ppm
combination with BHT or PG prepared from dry
mixes
Essential oils, citrus oil flavors, dry Not to exceed 0.125%, alone or in Chewing gum base 1,000 ppm
flavours combination with BHT or PG
Citrus oils Not to exceed 0.5%, alone or in Dry breakfast cereals 50 ppm
combination with BHT or PG
Partially defatted pork or beef fatty Not to exceed 0.065%, alone or in Dry diced glazed fruit 32 ppm
tissues combination with BHT
Vitamin A liquids for foods 5 mg/1,000,000 international units Dry mixes for beverages 90 ppm
and desserts
Dry beverage mixes, dry dessert 0.009% Edible fats and oils 200 ppm
and confection mixes excluding butterfat and
margarine
Active dry yeast 0.1% Emulsion stabilizers for 200 ppm
shortenings
Unstandardized foods except Not to exceed 0.02% of total fat Essential oils 1,000 ppm
preparations of meat and meat content of food, alone or in
by products, fish, poultry meat combination with BHT or PG
and its by products
Dry vitamin D preparations for 10 mg/1,000,000 international Margarine 200 ppm
food units
Margarine Not to exceed 0.01% of the fat Potato flakes 50 ppm
with BHT or PG or both
Dried cooked poultry meat 0.015% of the fat content alone or Potato granules 10 ppm
in combination with PG or citric
acid or both
Sweet potato flakes 50 ppm
BHT Class IV, B.2 21 CFR 172.115 Alone or in combination
with BHA
Fats and oils, lard, shortening Not to exceed 0.02%, alone or in Dehydrated potato 50 ppm
combination with BHA, PG or shreds
TBHQ
Dried breakfast cereals, dehy- Not to exceed 0.005%, alone or in Dry breakfast cereals 50 ppm
drated potato products combination with BHA or PG
Chewing gum Not to exceed 0.02%, alone or in Chewing gum base 1,000 ppm
combination with BHA or PG
Essential oils, Citrus oil flavors, Not to exceed 0.125%, alone or in Edible fats and oils 200 ppm
Dry flavours combination with BHA or PG excluding butter fat and
margarine
Citrus oils Not to exceed 0.5%, alone or in Emulsion stabilizers for 200 ppm
combination with BHA or PG shortenings
Essential oils 1,000 ppm
Partially defatted pork or beef Not to exceed 0.065%, alone or in Margarine 200 ppm
fatty tissues combination with BHA
Vitamin A liquids for foods 5 mg/1,000,000 international Potato flakes 50 ppm
units
Parboiled rice 0.0035% Potato granules 10 ppm
Unstandardized foods except Not to exceed 0.02% of total fat Sweet potato flakes 50 ppm
preparations of meat and meat content of food, alone or in
byproducts, fish, poultry meat combination with BHA or PG
and its byproducts
Dry vitamin D preparations for 10 mg/1,000,000 international
food units
content, alone or in combina-
tion with BHA or PG or both
Citric acid Class IV, C.3

and diglyceride, shortenings,
preparations of meat and meat
byproducts, fish, poultry meat
and its byproducts
Dried cooked poultry meat Not to exceed 0.015% of fat
tion of BHA or PG or both
Monoacylglycerol Class IV, M.1 21 CFR 172.832 Not to exceed 200 ppm of
citrate total fat
and diglyceride, shortenings
and its byproducts
Margarine Not to exceed 0.01% of fat con-
tent, alone or in combination of
mono i sopropyl citrate or
stearyl citrate or both
Mono isopropyl Class IV, M.2 21 CFR 582.6511 GRAS with GMP
citrate
In fats and oils, monoglycerides GMP Use as a sequestrant
and diglyceride, shortenings
and its byproducts
Margarine Not to exceed 0.01% of fat con-
tent, alone or in combination of
mono glyceride citrate or
stearyl citrate or both
Propyl gallate Class IV, P.1 Not to exceed 0.02%, alone or in 21 CFR 184.1660 Not to exceed 0.02% of
In fats and oils, lard, shorten- combination with BHA, BHT or fat content including
ings TBHQ essential (volatile) oil
content
Chewing gum base 100 ppm
Dried breakfast cereals, dehy- Not to exceed 0.005%, alone or in Edible oils and fats 100 ppm
drated potato products combination with BHA or BHT excluding butterfat and
margarine
Chewing gum Not to exceed 0.02%, alone or in
combination with BHA or BHT
Margarine 100 ppm
Essential oils, dry Not to exceed 0.125%, alone or in
flavors combination with BHA or BHT
Citrus oils Not to exceed 0.5%, alone or in
combination with BHA or BHT
Unstandardized foods except Not to exceed 0.02% of the fat
preparations of meat and meat content alone or in combination
byproducts, fish, poultry meat with BHA or BHT
and its byproducts
tion with BHA or BHT or both
Dried cooked poultry meat 0.015% of the fat content alone or
in combination with BHA or
citric acid or both
TBHQ Class IV, T.1A Not to exceed 0.02% alone or in 21 CFR 172.185 Not to exceed 0.02% of
In fats and oils, lard, combination with BHA or BHT fat content including
shortenings or PG essential (volatile) oil,
Alone or in combination
with BHA and/or BHT
Tocopherols Class IV, T.2 GMP 21 CFR 182.3890 GRAS with GMP
(a-,concentrate In fats and oils, lard, monogly-
or mixed) cerides and diglyceride,
shortenings
by products, fish, poultry meat
and its by products
a
From Canada Food and Drugs Act.
b
From U.S. Code of Federal Regulations.
REFERENCES 483
TABLE 10. Acceptable Daily Intake (ADI) Levels of Antioxidants Commonly Used as
Food Additives.
Antioxidant ADI Reference
Ascorbic acid and Derivates No ADI specified for salts of Na, K, Ca 162
01.25 mg/kg bw for ascorbyl palmitate or stearate or
sum of both if used together 163
BHA 00.5 mg/kg bw (JECFA) 164
00.5 mg/kg bw temperory (SFA)
BHT 00.125 mg/kg bw (JECFA) 161
00.05 mg/kg bw (SFA) 165
TBHQ 00.2 mg/kg bw temperory allowed 161
Tocopherols 0.152.0 mg/kg bw for dl-a-tocopherol and
d-a-tocopherol concentrate 161
on the basis of the available data (biochemical, chemical, toxicological and other),
the total daily intake of the substance, from its use at levels necessary to achieve the
desired effect and from its acceptable background in food, and does not, in the opi-
nion of the committee, represent a hazard to health. For that reason and for the rea-
sons stated in the individual evaluations, the establishment of an ADI expressed in
numerical form is not deemed necessary (160, 161). For some of the compounds,
because of the insufficient information available, an ADI level is not specified or
No ADI allocated. Table 10 provides information on ADI available for food anti-
oxidants in use. Barlow (160) has provided a very descriptive examination on the
toxicological studies on antioxidants used as food additives.
Toxicological risks may develop when the daily doses of a compound rise above
a certain threshhold limit; therefore, toxicity is a matter of dose as well. Natural
antioxidants, especially carotenoids, phenolic acids, flavonoids, and sterols, may
also exert in vivo pro-oxidative activity. Rietjens et al. (166) have provided an ela-
borate discussion on the pro-oxidative chemistry and toxicity of well-known natural
antioxidants, including ascorbic acid, tocopherols, carotenoids, and flavonoids.
REFERENCES
1. B. Halliwell, M. A. Murcia, S. Chirico, and O. I. Aruoma, Crit. Rev. Food Sci. Nutr., 35, 7
(1995).
2. C. Moureu and C. Dufraisse, Chem. Rev., 3, 113 (1926).
3. H. Davy, Ann. Chim. Phys., 4, 347 (1817).
4. M. E. Duclaux, Compt. rend., 102, 1077 (1886).
5. M. Tsujimoto, J. Coll. Engl. Imp. Univ. Tokyo, 4181 (1908).
6. C. W. Wright, Am. J. Pharm., 24, 180 (1852).
7. E. Chevreal, Ann. Chim. Phys., 47, 209 (1856).
8. C. Moureu and C. Dufraisse, Compt. rend., 174, 258 (1922).
9. C. Moureu and C. Dufraisse, Compt. rend., 175, 127 (1922).

10. C. Moureu and C. Dufraisse, J. Chem. Soc., 127, 1 (1925).
11. C. Moureu and C. Dufraisse, Chem. Ind., 47, 819 (1928).
12. C. D. Lowry, G. Egloff, J. C. Morrel, and G. C. Dryer, Ind. Eng. Chem., 25, 804 (1933).
13. J. L. Bolland and P. tenHave, Discuss. Farad. Soc., 2, 252 (1947).
14. H. S. Olcott and H. A. Mattill, J. Am. Chem. Soc., 58, 1627 and 2208 (1936).
15. K. C. Bailey, The Retardation of Chemical Reactions, Edward Arnold Co., London,
1937, pp. 111, 179.
16. G. Scott, Atmospheric Oxidation and Antioxidants, Elsevier, London, 1965, pp. 1 and 15.
17. M. W. Ranney, Antioxidants Recent Developments, Noyes Data Corp., Park Ridge,
New Jersey, 1979, p. 1.
18. F. Shahidi and P. K. J. P. D. Wanasundara, Crit. Rev. Food Sci. Nutr., 32, 67 (1992).
19. E. N. Frankel, Lipid Oxidation, The Oily Press, West Ferry, Dundee, 1998, pp. 13, 79,
115, 129, and 167.
20. D. G. Bradley and D. B. Min, Crit. Rev. Food Sci. Nutr., 31, 211 (1992).
21. O. R. Buettner, Arch. Biochem. Biophys., 300, 535 (1993).
22. E. A. Decker, in C. C. Akoh and D. B. Min, eds., Food Lipids: Chemistry, Nutrition and
Biotechnology, 2nd ed., Marcel Dekker, Inc., New York, 2002, p. 517.
23. W. W. Nawar, in O. Fennema, ed., Food Chemistry, 3rd ed., Marcel Dekker, Inc.,
New York, 1996, p. 225.
24. M. H. Gordon, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier Applied Science,
New York, 1990, p. 1.
25. D. F. Buck, Cereal Foods World, 29, 301 (1984).
26. N. Uri, in H. M. Sinclair, ed., Proceedings of International Conference in Biochemistry:
Problems of Lipids, Academic Press, New York, 1958, p. 30.
27. J. K. Kochi, Free Radicals, 1, 591 (1993).
28. J. Pokorny, in H.W.-S. Chan, ed., Autooxidation of Unsaturated Lipids, Academic Press,
London, 1987, p. 141.
29. H. B. Dunford, Free Rad. Biol. Med., 3, 405 (1987).
30. C. Ruben and K. Larsson, J. Dispersion Sci. Technol., 6, 213 (1985).
31. J. R. Mahoney and E. Graf, J. Food Sci., 51, 1293 (1986).
32. E. Graf and J. W. Eaton, Free Rad. Biol. Med., 8, 61 (1990).
33. R. S. Lindsay, in O. Fennema, ed., Food Chemistry, 3rd ed., Marcel Dekker, Inc.,
New York, 1996, p. 767.
34. B. Halliwell and J. M. C. Gutteridge, Arch. Biochem. Biophys., 246, 501 (1986).
35. C. E. Brown, J. Theor. Biol., 88, 245 (1981).
36. E. A. Decker, V. Ivanov, B. Z. Zhu, and B. Frei, J. Agric. Food Chem., 49, 511 (2001).
37. P. Palozza and N. I. Krinsky, Methods Enzymol., 213, 403 (1992).
38. W. Stahl and H. Sies, Ann. N. Y. Acad. Sci., 691, 10 (1992).
39. J. Kanner, J. B. German, and J. E. Kinsella, Crit. Rev. Food Sci. Nutr., 25, 317
(1987).
40. A. Claiborne, in C. R. A. Greenwald, ed., CRC Hand Book of Methods of Oxygen Radical
Research, Vol. I, CRC Press, Boca Raton, Florida, 1985, p. 283.
REFERENCES 485
41. W. A. Gunzler and L. Flohe, in R. A. Greenwald, ed., CRC Hand Book: Methods of
Oxygen Radical Research, Vol 1. CRC Press, Boca Raton, Florida, 1985, p. 285.
42. M. Maioriono, C. Gregolin, and F. Ursini, Methods Enzymol., 186, 448 (1990).
43. W. W. Nawar, in O. Fennema ed., Food Chemistry, 2nd ed., Marcel Dekker, Inc.,
New York, 1986, p. 139.
44. S. Z. Dziedzic, J. L. Robinson, and B. J. F. Hudson, J. Agric. Food Chem., 34, 1027
(1986).
45. S. Englard and S. Seifter, Ann. Rev. Nutr., 6, 365 (1986).
46. E. N. Frankel and A. S. Meyer, J. Sci. Food Agric., 80, 1925 (2000).
47. M. Antolovich, P. D. Prenzler, E. Patsalides, S. McDonald, and K. Robards, Analyst,
127, 183 (2002).
48. D. E. Pratt, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, Edible Oil and Fat
Products, Products and Application Technology, 5th ed., Wiley, New York, 1995, p. 523.
49. S. Z. Dziedzic and B. J. F. Hudson, J. Am. Oil Chem. Soc., 61, 1042 (1984).
50. F. Shahidi and U. N. Wanasundara, in C. C. Akoh and D. B. Min, eds., Food Lipids:
Chemistry, Nutrition and Biotechnology, 2nd ed., Marcel Dekker, Inc., New York, 2002,
p. 465.
51. H. S. Olcott and E. Einest. J. Am. Oil Chem. Soc., 35, 161 (1958).
52. P. Ke and R. G. Ackman, J. Am. Oil Chem. Soc., 53, 636 (1976).
53. M. S. Blois, Nature, 181, 1199 (1958).
54. T. Yoshida, K. Mori, T. Hatano, T. Okumura, I. Uchura, K. Komagoe, Y. Fujita, and
T. Okuda, Chem. Pharm. Bull., 37, 1919 (1989).
55. B. Halliwell, J. M. C. Gutteridge, and O. I. Aruoma, Anal. Biol. Chem., 63, 215
(1987).
56. V. Ponti, M. U. Dianzani, K. Cheeseman, and T. F. Slater, Chem. Biol. Interact., 23, 281
(1978).
57. C. Yuting, Z. Rongling, J. Znongjan, and J. Yong, Free Rad. Biol. Med., 199, 19 (1990).
58. N. Cotelle, J.-L. Bernier, J.-P. Catteau, J. Pommery, J.-C. Wallet, and E. M. Gaydou, Free
Rad. Biol. Med., 20, 15 (1996).
59. X. Shi, N. S. Dalal, and A. C. Jain, Food Chem. Toxicol., 29, 1 (1991).
60. A. E. Holley and K. H. Cheeseman, Brit. Med. Bull., 49, 494 (1993).
61. M. K. Sharma and G. R. Buettner, Free Rad. Biol. Med., 14, 649 (1993).
62. I. F. F. Benzie and J. J. Strain, Anal. Biochem., 239, 70 (1996).
63. I. F. F. Benzie and J. J. Strain, Methods Enzymol., 299, 15 (1999).
64. P. T. Gardner, T. A. C. White, D. B. McPhail, and G. G. Duthie, Food Chem., 68, 471
(2000).
65. G. Cao, H. M. Alessio, and R. G. Cutler, Free Rad. Biol. Med. 14, 303 (1993).
66. G. Cao, C. P. Verdon, A. H. B. Wu, H. Wang, and R. L. Prior, Clin. Chem., 413, 1738
(1995).
67. C. Guo, G. Cao, E. Sofic, and R. L. Prior, J. Agric. Food Chem., 45, 1787 (1997).
68. D. D. M. Wayner, G. W. Burton, K. U. Ingold, and S. Locke, FEBS Lett., 187, 33 (1985).
69. M. Valkonen and T. Kuusi, J. Lipid Res., 38, 823 (1997).
70. C. A. Rice-Evans and N. J. Miller, Methods Enzymol., 234, 279 (1994).
71. C. A. Rice-Evans, N. J. Miller, P. G. Bolwell, P. M. Bramley, and J. B. Pridham, Free

Rad. Res., 22, 375 (1995).
72. E. N. Frankel, Trends Food Sci. Technol., 4, 220 (1993).
73. The Official Methods and Recommended Practices of the American Oil Chemists Society,
5th ed., AOCS Press, Champaign, Illinois, 2000.
74. W. L. Porter, in M. G. Simic and M. Karel, eds., Autooxidation in Food and Biological
Systems, Plenum Press, New York, 1980, p. 295.
75. W. L. Porter, E. D. Black, and A. M. Drolet, J. Agric. Food Chem., 37, 615 (1989).
76. W. L. Porter, Toxicol. Ind. Health, 9, 93 (1993).
77. E. N. Frankel, S. W. Huang, J. Kanner, and J. B. German. J. Agric. Food Chem., 42,
1054 (1994).
78. S. W. Huang and E. N. Frankel, J. Agric. Food Chem., 45, 3033 (1997).
79. G. Siripriya, K. Chandrasekharan, V. S. Murty, and T. S. Chandra, Food Chem., 57, 539
(1996).
80. L. A. Santiago, M. Hiramatsu, and A. Mori, J. Nutr. Sci. Vitaminol., 38, 297 (1992).
81. M. Wettasinghe and F. Shahidi, Food Chem., 70, 17 (2000).
82. V. Bondet, W. Brand-William, and C. Berset, Lebensm-wiss.u.-Technol., 30, 609 (1997).
83. N. Darmon, Y. Fernandez, C. Cambon-Gros, and S. Mitiavila, Food Addit. Contam., 7,
560 (1990).
84. A. S. Meyer, D. S. Rrbk, and J. Adler-Nissen, Food Chem., 51, 171 (1994).
85. J. Pokorny, in J. Pospsil and P. P. Klemchuk, eds., Oxidation Inhibition of Organic
Materials, Vol. I, CRC Press, Boca Raton, Florida, 1989, p. 349.
86. J. Pokorny, in M. S. Rahman, ed., Handbook of Food Preservation, Marcel Dekker, Inc.,
New York, 1999, p. 309.
87. E. N. Frankel, P. M. Cooney, H. A. Moser, J. C. Cowen, and C. D. Ewan, Fette Seifen
Anstrichm., 10, 1036 (1959).
88. P. P. Coppen, in J. C. Allen and R. J. Hamilton, eds., Rancidity in Foods, Applied
Science, London, 1983, p. 76.
89. J. D. Dziezak, Food Technol., 9, 94 (1986).
90. W. M. Cort, Adv. Chem. Ser., 200, 533 (1983).
91. S. Z. Dziedzic and B. J. F. Hudson, Food Chem., 12, 205 (1983).
92. H. Y. D. Belitz and W. Grosch, Food Chemistry, Springer Verlag, New York, 1987, p. 128.
93. P. Vinas, C. K. Hernandez Cordoba, and C. Sanchez-Pedrenos, Food Chem., 42, 421
(1991).
94. D. W. Reische, D. A. Lillard, and R. R. Eitenmiller in C. C. Akoh and D. B. Min, eds.,
Food Lipids: Chemistry, Nutrition and Biotechnology, 2nd ed., Marcel Dekker, Inc.,
New York, 2002, p. 489.
95. E. Niki, Chem. Phys. Lipids, 44, 227 (1987).
96. G. Scott, Antioxidants in Science, Technology, Medicine and Nutrition, Albion Publish-
ing, West Sussex, Chichester, England, 1997, p. 202.
97. J. J. Pokorny, Trends Food Sci. Technol., 9, 223 (1991).
98. D. L. Madhavi, R. S. Singhal, and P. R. Kulkarani, in D. L. Madhavi, S. S. Deshpande,
and D. K. Salunkhe, eds., Food Antioxidants: Technological, Toxicological and Health
Perspectives, Marcel Dekker, New York, 1996, p. 159.
REFERENCES 487
99. C. H. Foyer, in R. G. Alscher and J. L. Hess, eds., Antioxidants in Higher Plants, CRC
Press, Boca Raton, Florida, 1993, p. 31.
100. P. Schuler, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier Applied Science,
New York, 1990, p. 99.
101. P. A. Seib, Int. J. Vit. Nutr. Res. Suppl., 27, 259 (1985).
102. S. S. Desphande, U. S. Desphande, and D. K. Salunkhe, in D. L. Madhavi, S. S.
Deshpande, and D. K. Salunkhe, eds., Food Antioxidants: Technological, Toxicological
and Health Perspectives, Marcel Dekker, New York, 1996, p. 361.
103. G. W. Burton and K. U. Ingold, Science, 224, 569 (1984).
104. D. C. Leiber, Ann. N. Y. Acd. Sci., 691, 20 (1992).
105. P. Palozza and N. I. Krinsky, Arch. Biochem. Biophys., 297, 184 (1992).
106. K. M. Haila, S. M. Lievenon, and M. I. Heinonen, J. Agric. Food Chem., 44, 2096 (1996).
107. C. J. Dillard, V. C. Gavino, and A. L. Tappel, J. Nutr., 113, 226 (1983).
108. F. V. Timmerman, Fat Sci. Technol., 92, 201 (1990).
109. F. Shahidi and M. Naczk, Food Phenolics: Sources, Chemistry, Effects and Applications,
CRC Press, Boca Raton, Florida, 1994, pp. 9, 53, 75, 109.
110. P. K. J. P. D. Wanasundara, F. Shahidi, and V. K. S. Shukla, Food Rev. Int., 13, 225 (1997).
111. V. K. S. Shukla, P. K. J. P. D. Wanasundara, and F. Shahidi, in F. Shahidi, ed., Natural
Antioxidants: Chemistry, Health Effects and Applications, AOCS Press, Champaign,
Illinois, 1997, p. 97.
112. F. Shahidi and M. Naczk, Phenolics in Food and Nutraceuticals, CRC Press, Boca Raton,
Florida, 2003.
113. H. Chimi, A. Sadik, B. Le Tutor, and M. Rahmani, R. Fr. Des Crops Grass., 35, 339
(1988).
114. J. Pokorny, in H. W. S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic Press,
London, 1988, p. 141.
115. M.-E. Cuvelier, H. Richard, and C. Berset, BioSci. Biotech. Biochem., 56, 324 (1992).
116. R. D. Howarth, Anal Rep. Prog. Chem., 33, 266 (1936).
117. T. Osawa, M. Nagata, M. Namiki, and Y. Fukuda, Agric. Biol. Chem., 49, 3351 (1985).
118. Y. Fukuda, T. Osawa, M. Namiki, and T. Ozaki, Agric. Biol. Chem., 49, 301 (1985).
119. H. Katsuzaki, S. Kawakishi, and T. Osawa, Heterocycles, 36, 933 (1993).
120. H. Katsuzaki, T. Osawa, and S. Kawakishi, in Food Phytochemicals for Cancer
Prevention (II), American Chemical Society Press, Champaign, Illinois, 1994, p. 275.
121. Y. Fukuda, M. Nagata, T. Osawa, and M. Namiki, J. Am. Oil Chem. Soc., 63, 1027 (1986).
122. Y. Fukuda, M. Isobe, M. Nagata, T. Osawa, and M. Namiki, Heterocycles, 24, 923 (1986).
123. G. G. Yen and S. L. Shyu, Food Chem., 31, 215 (1989).
124. H. Yoshida, J. Shigezaki, S. Takagi, and G. Kajimoto, J. Sci. Food Agric., 6, 407 (1995).
125. F. Shahidi, R. Amarowicz, H. A. Abou-Gharbia, and A. Y. Shehata, J. Am. Oil Chem.
Soc., 74, 147 (1997).
126. J. E. Brekke and H. J. Klosterman, Proc. North Dakota Agric. Sci., 10, 18 (1956).
127. L. U. Thompson and W. E. Ward, in G. S. Gilvani and J. J. B. Anderson, eds.,
Phytoestrogens and Health, American Oil Chemists Society Press, Champaign, Illinois,
2002, p. 405.
128. H. Hyvarinen, J.-M. Pihlava, V. Hietanieni, E.-L. Ryhanen, and H. Kohonen, Innov. Food
Technol., 15, 64 and 68 (2002).
129. K. Prasad, Mol. Cell. Biochem., 168, 117 (1997).
130. D. D. Kitts, Y. V. Yuan, A. N. Wijewickreme, and L. U. Thompson, Mol. Cell. Biochem.,
202, 91 (1999).
131. D. Boskou and H. Katsikas, Acta Alimentaria, 8, 317 (1979).
132. M. H. Gordon and P. Magos, Food Chem., 10, 141 (1983).
133. P. J. White and L. S. Amstrong, J. Am. Oil Chem. Soc., 63, 525 (1986).
134. Z. Xu and S. Godber, J. Am. Oil Chem. Soc., 78, 645 (2001).
135. T. Wang, K. B. Hicks, and R. Moreau, J. Am. Oil Chem. Soc., 79, 1201 (2002).
136. T. Ohta, S. Yamasaki, Y. Egashira, and H. Sanada, J. Agric. Food Chem., 42, 653 (1994).
137. Z. Xu and S. Godber, J. Agric. Food Chem., 47, 2724 (1999).
138. L. L. Tiang and P. J. White, J. Am. Oil Chem. Soc., 7, 1087 (1994).
139. E. N. Frankel, S.-W. Huang, R. Aeschbach, and E. Prior, J. Agric. Food Chem., 44, 131
(1996).
140. E. N. Frankel, S.-W. Huang, E. Prior, and R. Aeschbach, J. Sci. Food Agric., 72, 20
(1996).
141. S.-W. Huang, E. N. Frankel, J. B. German, and R. Aeschbach, J. Agric. Food Chem., 45,
1991 (1997).
142. D. A. Balentine, S. A. Wiseman, and L. C. Bouwens, Crit. Rev. Food Sci. Nutr., 37, 693
(1997).
143. S. A. Wiseman, D. A. Balentine, and B. Frei, Crit. Rev. Food Sci. Nutr., 37, 705 (1997).
144. J. Kuhnau, World Rev. Nutr. Diet., 24, 117 (1976).
145. S. V. Jovanovic, S. Steenken, M. Tosic, B. Marjanovic, and M. G. Simic, J. Am. Chem.
Soc., 116, 4846 (1994).
146. P.-G. Pietta, J. Nat. Prod., 63, 1035 (2000).
147. D. E. Pratt and B. J. F. Hudson, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier
Applied Science, New York, 1990, p. 171.
148. W. Bors, W. Hiller, C. Michel, and M. Saran, Methods Enzymol., 186, 343 (1990).
149. A. Hopia and M. Heinonen, J. Am. Oil Chem. Soc., 76, 139 (1999).
150. U. N. Wanasundara and F. Shahidi, Food Chem., 50, 393 (1994).
151. U. N. Wanasundara and F. Shahidi, J. Am. Oil Chem. Soc., 73, 1183 (1996).
153. E. N. Frankel, S. W. Huang, and R. Aeschbach, J. Am. Oil Chem. Soc., 74, 1309 (1997).
154. M. Koketsu, in T. Yamamoto, L. R. Juneja, D.-C. Chu, and M. Kim, eds., Chemistry and
Application of Green Tea, CRC Press, Boca Raton, Florida, 1997, p. 37.
155. Y. Hirose, H. Yamamoto, and M. Nakayama, Agric. Biol. Chem., 54, 567 (1990).
156. T. Yamamoto, L. R. Juneja, D.-C. Chu, and M. Kim, Chemistry and Application of Green
Tea, CRC Press, Boca Raton, Florida, 1997.
157. S. P. Kochhar and J. B. Rossell, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier
Applied Science, New York, 1990, p. 19.
158. W. Horowitz, ed., AOAC Official Methods, Association of Official Analytical Chemists
17th ed., AOAC International, Arlington, VA, 2002, p. 1604.
REFERENCES 489
159. J. G. Elliot, Food Technol., 199, 46, (1999).

160. S. Barlow, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier Applied Science,
New York, 1990, p. 253.
161. Joint FAO/WHO Expert Committee on Food Additives, Technical Report Series, World
Health Organization, 751, 18 (1987).
165. R. Haigh, Food Chem. Toxicol., 24, 1031 (1986).
166. I. M. C. M. Rietjens, M. G. Boersma, L. de Haan, B. Spenklink, H. M. Awad, N. H. P.
Cnubben, J. J. van Zanden, H. van der Woude, G. M. Alink, and J. H. Koema, Environ,
Toxicol. Pharmacol., 11, 321 (2002).
12
Antioxidants:
Regulatory Status
1. INTRODUCTION
Oxidation of unsaturated lipids is a major cause of food quality deterioration by

giving rise to the development of off-flavor compounds and loss of nutritional value
of food products (1). Although it has been known for a long time that lipid oxida-
tion can be induced by catalytic systems such as light, temperature, enzymes,
metals, and metalloproteins; the mechanism of oxidation reactions remained uncer-
tain until the 1940s when free radicals and reactive oxygen species were found to be
involved in oxidation processes by the pioneering work of Farmer et al. (2), Bolland
and Gee (3), and Bateman et al. (46). Furthermore, antioxidants were found to
protect lipids against oxidation either by quenching free radicals or scavenging
oxygen, among others (6). Antioxidants are substances that, when present in foods
at low concentrations compared with that of an oxidizable substrate, markedly
delay or prevent the oxidation of the substrate (7). Antioxidants that fit in this defi-
nition include free radical scavengers, inactivators of peroxides, and other reactive
oxygen species (ROS), chelators of metals, and quenchers of secondary lipid
oxidation products that produce rancid odors (8). Antioxidants have also been
491
492 ANTIOXIDANTS: REGULATORY STATUS
used in the health-related area because of their ability to protect the body against
damage caused by ROS as well as reactive nitrogen species (RNS) and those of
reactive chlorine species (RCS) (9).
Antioxidants can be broadly classified by their mechanism of action as primary
antioxidants, which break the chain reaction of oxidation by hydrogen donation and
generation of more stable radicals, and secondary antioxidants, which slow the oxi-
dation rate by several mechanisms, including chelation of metals, regeneration of
primary antioxidants, decomposition of hydroperoxides, and scavenging of oxygen,
among others. These substances may occur naturally in foods, such as tocopherols
and ascorbic acid; however, natural antioxidants are often, at least partially, lost
during processing or storage, thus exogenous antioxidants are intentionally added
to products or their precursors participate in the formation of antioxidants during
processing. Although there are many of compounds that have been proposed to
inhibit oxidative deterioration processes, only a few can be used in food products
(10). Antioxidants for use in food processing must be inexpensive, nontoxic, effec-
tive at low concentrations (0.0010.02%), capable of surviving processing (carry-
through), stable in the finished products, and devoid of undesirable color, flavor, and
odor effects. In general, the selection of antioxidants depends on products, compati-
bility, and regulatory guidelines (11). In this chapter, the properties and applications
of antioxidants in foods as well as their regulatory status are discussed.
2. SYNTHETIC ANTIOXIDANTS
Although the use of antioxidants dates back to ancient times when herbs and spices
were used in food preservation, modern antioxidant technology is only about
60 years old. Since free radicals were found to be responsible for lipid oxidation,
hundreds of natural and synthetic compounds have been evaluated for their efficacy
as radical scavengers or for their other inhibitory effects. Among them, only four
synthetic antioxidants are widely used in foods; namely, butylated hydroxyanisole
(BHA), butylated hydroxytoluene (BHT), propyl gallate (PG), and tert-butylhydro-
quinone (TBHQ) (7, 12). Scientists are attempting to develop novel synthetic anti-
oxidants aimed at retarding the effects of free-radical-induced damage in various
food products as well as in the human body cells (13). Synthetic antioxidants
used in the food industry can be added as direct additives or indirectly through
diffusion from packaging material (6).
All antioxidants have points of strengths and weaknesses. Therefore, certain
points, such as thermal stability, effective concentration, and synergism, should
be taken into consideration when selecting antioxidants for use in particular foods.
Regulatory status is another factor that cannot be ignored, especially for some anti-
oxidants that have been reported to show potential adverse health effects. Synthetic
antioxidants have been tested for safety and approval for use in food at low concen-
trations on the basis of complex toxicity studies (10). Allowable limits for use of
antioxidants vary greatly from country to country, and depend on the food product
under consideration (11).
SYNTHETIC ANTIOXIDANTS 493
OH OH
C(CH3)3
C(CH3)3
OCH3 OCH3
2isomer 3isomer
Figure 1. Chemical structures of BHA molecules.
2.1. BHA (Butylated Hydroxyanisole) and BHT (Butylated

Hydroxytoluene)
Phenolic and polyphenolic compounds are the most active dietary antioxidants (14).
The structural variation of phenolic antioxidants directly influences their physical
properties, resulting in differences in their antioxidant activity. BHA and BHT are
examples of phenols, in which the aromatic ring contains alkyl groups (hindered
phenols), which are extremely effective as antioxidants (11).
Chemically, BHA is a mixture of two isomers (2-tertiary-butyl-4-hydroxyani-
sole and 3-tertiary-butyl-4-hydroxyanisole) (Figure 1). The 3-isomer is generally
considered to be a better antioxidant, and accounts for 90% of the commercial
BHA (12). BHA is a white, waxy solid that is sold in the form of flake or tablet.
It is a highly fat-soluble monophenolic antioxidant that is extensively used in bulk
oils as well as oil-in-water emulsions (1012). It is effective in animal fats and rela-
tively ineffective in vegetable oils. Demonstrating considerable effectiveness in
controlling the oxidation of short-chain fatty acids, BHA is frequently used for the
preservation of coconut and palm kernel oil in cereal and confectionery products
(10). BHA is good in baking because of its stability to heat and its mild alkaline
conditions, although its application in frying is limited due to its volatility (11).
However, it can be added to packaging materials to provide protection to food pro-
ducts inside the package through volatilization (12). BHA is particularly useful in
protecting the odor and flavor of essential oils (10). Furthermore, BHA has been
reported to possess antimicrobial activity (1517) and is known to act synergisti-
cally with other antioxidants such as BHT.
BHT (3,5-di-tert-butyl-4-hydroxytoluene) (Figure 2) is a white crystalline solid
with properties similar to BHA (12). It is appropriate for thermal treatment but not
as stable as BHA (11). Being able to regenerate BHA, BHT is commonly used in
OH
(CH3)3C C(CH3)3
CH3
Figure 2. Chemical structure of BHT.

combination with BHA to provide greater antioxidant activity (18). BHT does not
have an optimum concentration; usually, BHA/BHT mixtures are added to foods at
levels of up to 0.02% (10, 11). Both BHA and BHT have a slight phenolic odor, and
may impart undesirable odor in foods when used at high temperature for an
extended period of time (10, 11).
Although synthetic antioxidants have widely been used in the food industry,
there are some arguments about their safety (19). The use of BHA and BHT in
foods has been decreased due to their potential action as promoters of carcinogen-
esis (20). In addition to the carcinogenicity of BHA in the forestomach of rodents,
BHA and BHT have been reported to be cytotoxic (2123). Furthermore, a sugges-
tion has been made that BHT be withdrawn from use in all foods because of its
possible adverse effects on the kidney and liver as well as lung tissues of rat
(24, 25). However, some scientists have noted that the metabolism of BHT in rat
and man are too widely different to allow a proper hazard assessment of BHT in
humans (26). It is generally considered that permitted food antioxidants, such as
BHA and BHT, have a considerable safety margin; for instance, the dose for
enhancement of carcinogenesis is at least 1500-fold greater than that in human
exposure (27, 28). Meanwhile, BHA and BHT have been reported by some
researchers to pose no cancer hazard to humans and, on the contrary, have health
benefits related to their anticarcinogenic and antimutagenic properties as well as
inhibition of cholesteral oxidation (2932).
Despite positive and negative reports of these synthetic antioxidants on human
health, their use is subject to regulation, in the United States, under the Food and
Drug Administration (FDA) and the U.S. Department of Agriculture (USDA); in
Canada, the Food and Drug Regulations (National Health and Welfare); in Europe,
the European Economic Community (EEC); and in Japan, the Food Sanitation Law.
Many other countries have adopted regulations similar to those used in the United
States, with significant differences existing both in the antioxidants approved and in
their application and level of usage (10, 11). According to the existing food additive
regulations published by the FDA, BHA and BHT are lawful for use individually or
in combination at a maximum level of 0.02%, or 200 ppm, based on the lipid con-
tent of food products, as specified by the Code of Federal Regulations (CFR) (6, 7, 12).
Although BHA and BHT are effective at low concentrations, they become pro-
oxidant at high levels in foods (11, 33). As specified in 21CFR, 172.100, and
172.115, limitations for BHA and BHT, alone or in combination for specific pro-
ducts, are as follows: 10 ppm in potato granules; 50 ppm in dehydrated potato
shreds, dry breakfast cereals, potato flakes, and sweet potato flakes; and 200 ppm
in emulsion stabilizers for shortenings (11). BHA and BHT are not allowed in fish
products (5). The summery of regulations, applications, and properties of BHA and
BHT are shown in Tables 1 and 2.
The daily dietary intakes of BHA and BHT have been estimated in many coun-
tries. The daily intakes of BHA and BHT in Japan in 1998 were 0.119 and
0.109 mg/d/person, which reflect 0.5% and 0.7% of the acceptable daily intake
(ADI), respectively (35). The estimates of theoretical maximum daily intake
(TMDI) of BHA and BHT in Brazil published in 2001 were in the range of
TABLE 1. Properties, Applications, and Regulations of BHA.
NAME: Butylated hydroxyanisole (BHA)

CATEGORY: Antioxidant
FOOD USE: Bakery products/Meat products/Spices/Cereals/Dehydrated
mashed potatoes/Beverage mixes/Dessert mixes/Nuts/Vita-
mins/Yeast/Vegetable oils/Animal fats/Processed cheeses/
Margarine/Essential oils/Chewing gum base
SYNONYMS: Mixture of two isomers: 3-tertiary butyl-4-hydroxyanisole and
2-tertiary butyl-4-hydroxyanisole/(1,1-dimethylethyl)-4 meth-
oxyphenol/E320/ Antracine 12/Embanox/Nipantiox/Sustane
BHA/ Sustane 1-F/Tenox 4B/ Tenox 5B
FORMULA: (CH3)3CC6H3OCH3OH
MOLECULAR MASS: 180.25
PROPERTIES AND
APPEARANCE: White waxy flakes or tablets
MELTING RANGE IN C: 4855
FLASH POINT IN C: 130
PURITY %: Not less than 98.5 of 2-isomer and not less than 85 of 3-isomer
SOLUBILITY % AT VARIOUS TEMPERATURE/pH COMBINATIONS:
in water: at 20 C Insoluble
in vegetable oil: at 25 C 30% cottonseed oil
40% coconut, corn, peanut oils
50% soybean oil
in ethanol solution: 100% at 25 C >25%
in propylene glycol: at 20 C 70%
FUNCTION IN FOODS: Antioxidant preservative by terminating free radicals formed
during autoxidation of unsaturated lipids. It also possesses
antimicrobial activity as a phenolic compound.
ALTERNATIVES: BHT; PG; TBHQ
SYNERGISTS: BHT; propyl gallate; methionine; lecithin; thiodipropionic acid;
citric acid; phosphoric acid
FOOD SAFETY ISSUES: This antioxidant has not been subjected to great criticism of
safety. However, suspected for tumor formation in animals with
forestomach.
LEGISLATION: USA: Maximum usage level approved for general use; FDA
0.02% and USDA 0.01% of weight of fat.
Special applications include:
Chewing gum base: 0.01% by weight of chewing gum base
Active dry yeast or dry material
Emulsion stabilizer: 0.02% by weight of emulsion, shortenings,
stabilizer
Potato flakes, sweet potato flakes: 0.005% by weight, dry
breakfast cereal, of food material, packaging material
Potato granules: 0.001% by weight of potato granules
Dry mixes for beverages: 0.009% of material and desserts
Beverages and desserts, prepared from dry mixes: 0.0002%
Dry diced glazed fruits: 0.0032%
Flavor substances: 0.5% of essential oil content
U.K. and EUROPE: approved
CANADA: approved
AUSTRALIA/PACIFIC RIM and JAPAN: approved
Adapted from (34).

TABLE 2. Properties, Applications, and Regulations of BHT.
NAME: Butylated hydroxytoluene (BHT)

FOOD USE: Breakfast cereals/Baked goods/Potato chips/Vegetable
oils/Snack foods/Butter/Margarine/Frozen seafoods/
Chewing gum base
SYNONYMS: 2,6-bis (1,1-dimethylethyl)-4-methylphenol/2,6-di-tert-
butyl-p-cresol/2,6-di-tert-butyl-4-methylphenol/E321/
Antracine 8/Ionol CP/Dalpac/Impruvol/Vianol/Tenox
BHT/Tenox 8/Sustane BHT
FORMULA: [(CH3)3C]2C6H2CH3OH
PROPERTIES AND APPEARANCE: White granular crystals with slight odor
PURITY %: Not less than 99
in water: at 20 C Insoluble
in vegetable oil: at 25 C 30% cottonseed, coconut, corn,
peanut and soybean oils
in ethanol solution: 100% at 25 C 25%
in propylene glycol: at 20 C Insoluble
FUNCTION IN FOODS: Antioxidant preservative; prevents oxidative rancidity
development in oil-containing foods by terminating free
radicals formed during autoxidation of unsaturated lipids.
It possesses antimicrobial activity as a phenolic com-
pound.
ALTERNATIVES: BHA; PG; TBHQ
SYNERGISTS: BHA
FOOD SAFETY ISSUES: This antioxidant has not been subjected to great criticism
over safety.
LEGISLATION: USA: Maximum usage level approved for general use;
FDA 0.02% and USDA 0.01% of weight of fat.
Enriched rice: 0.0033%
Nonalcoholic beverages, frozen raw breaded shrimp,
mixed nuts and margarine: 0.02% based on oil content
Dry sausage: 0.003%
Fresh pork sausage, brown-and-serve sausage, pre-
grilled beef patties, pizza toppings, meatballs, dried
meats: 0.01%
Rendered animal fat or combination with vegetable fat,
poultry fat or various poultry procucts: 0.01%
Dry breakfast cereals: 0.005%
Emulsion stabilized for shortening: 0.02%
Potato granules: 0.001%
Potato flakes, sweet potato flakes, dehydrated potato
shreds: 0.005%
U.K. and EUROPE: approved
CANADA: approved
Adapted from (34).

0.090.15 and 0.050.10 mg/kg body weight, respectively, indicating that it is un-
likely to exceed the ADI (0.5 and 0.3 mg/kg body weight) (36). In the Netherlands,
the mean intake of BHA and BHT in 2000 was 105 and 351 mg/day (37). In Italy,
the likelihood of exceeding the ADI for BHA was very low. However, the TMDI of
BHT was above the ADI. The three food categories, pastry, cake, biscuits,
chewing gums, and vegetable oils and margarine, were the major sources of
BHT and contributed 74% of the TMDI (38).
2.2. TBHQ (tert-Butylhydroquinone) and Gallates

TBHQ (Figure 3) is a beige powder or is a white-to-tan crystal that is used fre-
quently in frying applications with highly unsaturated vegetable oils. Its solubility
in different solvents declines in the order of alcohol > fats > water. As a diphenolic
antioxidant, TBHQ is more effective in vegetable oils than BHA and BHT. It is
stable to heat and is regarded as the most effective antioxidant in preventing the
oxidation of frying oils and an alternative or supplement to oil hydrogenation
for increasing oxidative stability (10, 11). TBHQ shows excellent synergism
with other antioxidants such as citric acid. A ternary mixture containing TBHQ,
monoacylglycerol citrate (MGC), and ascorbyl palmitate (AP) exhibited the highest
thermal stability and provides optimum protection for oil during high-temperature
processing (39). TBHQ mixed with BHA and BHT can increase the smoke point of
fats and oils (40).
Three esters of gallic acid are approved for use in foods, namely propyl gallate
(PG), octyl gallate, and dodecyl gallate (Figure 4). PG is a white crystalline powder
that is slightly soluble in both water and fat, whereas the higher octyl and dedecyl
gallate are practically insoluble in water but dissolve easily in fats and oils (10).
OH
C(CH3)3
OH
Figure 3. Chemical structure of TBHQ.
OH
HO OH
COOR
R = C3H7 Propyl Gallate

= C8H17 Octyl Gallate
= C12H25 Dodecyl Gallate
Figure 4. Chemical structures of different alkyl gallates.

Therefore, PG is widely used in foods where lipid-soluble antioxidants such as

BHA, BHT, and TBHQ are not suitable. PG is inappropriate for frying due to its
poor stability at high temperatures. It decomposes at its melting point of 148 C (11,
12). Gallates can form undesirable, dark-colored complexes with iron and copper;
thus, they are sold as a mixture with metal chelators such as EDTA. Gallates also
act synergistically with other antioxidants (11, 12).
As synthetic antioxidants, the safety of TBHQ and gallates has been questioned
and, similar to BHA and BHT, both drawbacks and benefits of TBHQ and gallates
have been reported. According to the literature, TBHQ exhibited a nontypical mode
of cell death and proved cytotoxic toward human monocytic leukaemia cells (41). It
caused apoptosis and significantly promoted DNA damage (4244). PG suppressed
humoral immunity (45). The coadministration of TBHQ or PG with sodium nitrite
promoted forestomach carcinogenesis (46). Meanwhile, TBHQ and PG have been
reported to be beneficial to human health. TBHQ was reported to be an effective
inhibitor of cholesterol oxidation (32, 47), and PG provided inhibition of foodborne
pathogens (48). Anticarcinogenic and antimutagenic activities of TBHQ and PG
have also been reported (29, 49).
In accordance with regulations concerning the use of antioxidants in foods,
TBHQ is permitted for food use by the FDA and the USDA at less than 0.02%
and 0.01%, respectively. At levels higher than 0.02%, TBHQ may exert a pro-
oxidant effect. The CFR specifies the maximum addition of TBHQ as 0.02%
(200 ppm) (6, 12). However, the combination of TBHQ and PG is illegal (12). In
Japan and European coutries, addition of TBHQ in foods is not allowed (7, 11),
although its use in Canada is quite recent and dates back to 1999. PG is the only
gallate permitted in foods in the United States and Canada, but the use of higher
alkyl gallates is approved in several European countries (10). Gallates have opti-
mum concentrations for antioxidant activity and may act as pro-oxidants when
used at high levels (10). Tables 3 and 4 show detailed information on the regulatory
status of TBHQ and PG in various applications, respectively. With respect to daily
intake, according to investigations in Brazil and Italy, the estimate of TMDI for
TBHQ and gallates was very low and unlikely to exceed the ADI (36, 38).
2.3. Erythorbic Acid and Ascorbyl Palmitate

Erythorbic acid (or D-ascorbic acid) (Figure 5) is a white or slightly yellow crystal-
line powder that is often used in fruits and cured meats to enhance curing action and
to stabilize the color of food products (34). Unlike its isomer L-ascorbic acid,
erythorbic acid is not a natural constituent of foods and has minimal Vitamin C
activity (11).
Ascorbyl palmitate (Figure 6), a synthetic derivative of ascorbic acid, is a white
powder with a soapy taste and citrus-like odor (34). It has better lipid-solubility
compared with that of ascorbic acid and its salts, and is often used in combination
with a-tocopherol in lipid-containing foods (11). Ascorbyl palmitate prevents oxi-
dative rancidity by quenching singlet oxygen, among other modes of action (34).
TABLE 3. Properties, Applications, and Regulations of TBHQ.
NAME: Tert-butylhydroquinone (TBHQ)

FOOD USE: Dry cereals/Edible fats/Margarine/Pizza toppings/
Potato chips/Poultry/Dried meats/Sausages/Beef pat-
ties/Vegetable oils
SYNONYMS: 2-(1,1-dimethylethyl)-1,4-benzenediol/mono-t-butyl
hydroquinone/ Sustane TBHQ/Tenox TBHQ
FORMULA: (CH3)3CC6H3(OH)2
MOLECULAR MASS: 295
PROPERTIES AND APPEARANCE: White to tan color solid crystals, having a characteristic
odor
MELTING RANGE IN C: 126.5128.5
PURITY %: 99
SOLUBILITY % AT VARIOUS TEMPERATURE /pH COMBINATIONS:
in water: at 20 C <1% at 100 C 5%
in vegetable oil: at 20 C 10% in corn, cottonseed, and soybean oils
in ethanol solution: 100% 25%
in propylene glycol: at 20 C 30%
FUNCTION IN FOODS: Prevents oxidative rancidity developmentin foods by
terminating free radicals formation
ALTERNATIVES: BHA; BHT
SYNERGISTS: BHA; citric acid
FOOD SAFETY ISSUES: Has shown mutagenicity in vivo; therefore, some
countries consider that TBHQ does not meet current
standards of toxicity testing.
LEGISLATION: USA: Not allowed to use in combination with PG. For
general usage, FDA-0.02%, USDA-0.01%, based on
lipid content of food.
Special food use:
Nonalcoholic beverages
Margarine, mixed nuts: 0.02% alone or in combination
based on lipid content
Dried meats
Fresh pork or beef sausages
Pre-grilled beef patties
Pizza toppings
Meatballs: 0.01% based on weight of finished product
Rendered animal fats
EUROPE, U.K., NORWAY, DENMARK, SWEDEN,
SWITZERLAND: not allowed for food use
CANADA: allowed for food use
AUSTRALIA/PACIFIC RIM:
AUSTRALIA, NEW ZEALAND: allowed for food use
JAPAN: not allowed for food use
Adapted from (34).

TABLE 4. Properties, Applications, and Regulations of PG.
NAME: Propyl gallate (PG)

FOOD USE: Chewing gum base/Nonalcoholic beverages/Margarine/
Mixed nuts/Fresh or dry sausages/Pre-grilled beef
patties/Rendered animal fat/Pizza toppings and
meatbal
SYNONYMS: n-propyl-3,4,5-trihydroxybenzoate/3,4,5-trihydroxyben-
zoic acid/Gallic acid, propyl ester/E310/Nipa 49/
Nipagallin P/Tenox PG/Sustane PG
FORMULA: (HO)3C6H2COOCH2CH2CH3
PROPERTIES AND APPEARANCE: White crystalline powder with slight odor
PURITY %: Not less than 98 and not more than 102.5 on the dried
basis
in water: at 20 C <1%
in vegetable oil: at 20 C 1% in cottonseed oil
2% in soybean oil
insoluble in corn oil
in ethanol solution: 100% at 25 C >60%
FUNCTION IN FOODS: Prevents oxidation rancidity development in lipid-con-
taining foods by terminating free radicals formation
during autoxidation of unsaturated lipids.
ALTERNATIVES: BHA; BHT; TBHQ; octyl gallate; dodecyl gallate
SYNERGISTS: BHA; BHT
FOOD SAFETY ISSUES: Not subjected to great criticism over safety.
LEGISLATION: USA: not allowed to use in combination with TBHQ.
For general usage, FDA- 0.02% and USDA 0.01%
alone or in combination with BHT or BHA by weight of
lipid proportion of food.
Chewing gum base: 0.1%
Nonalcoholic beverages: 0.1%
Margarine: 0.1%
Mixed nuts: 0.02% based on oil content
French beef or pork sausages
Brown-and-serve sausages
Pre-grilled beef patties
Pizza toppings and meatballs: for all these 0.01% based
on weight of finished product
Rendered animal fats or combination of such fats with
vegetable fat: 0.01% based on lipid content
U.K.: approved
EUROPE: listed
CANADA: approved
Adapted from (34).

CH2OH
HO C H
O
O
H
HO OH
Figure 5. Chemical structure of erythorbic acid.
CH2OOC(CH2)14CH3
H C OH
O
O
H
HO OH
Figure 6. Chemical structure of ascorbyl palmitate.
More recently, replacement of the palmitate moiety with oleate has been proposed
to increase solubility as well as antioxidant activity in oils (50).
Both erythorbic acid and ascorbyl palmitate have Generally Recognized as
Safe (GRAS) status with the FDA (11). No restrictions on their usage levels are
imposed except for the maximum addition of 0.02% for ascorbyl palmitate in mar-
garine (34).
Table 5 shows the maximum levels permitted by the FDA for the four major syn-
thetic antioxidants (BHA, BHT, PG, and TBHQ) in specific applications (51). The
regulatory status for these antioxidants in the USA, Canada, and Europe is given in
Table 6; Table 7 summarizes their status in other countries for which a listing could
be found. In addition to the major synthetic antioxidants discussed above (BHA,
BHT, TBHQ, gallates, erythorbic acid, and ascorbyl palmitate), several other
TABLE 5. Maximum Levels Permitted for Antioxidants in Specific Applications.
Food Type Maximum Permitted Levels (ppm)

BHA BHT PG TBHQ
Active dry yeast 1000

Beverages from dry mixes 2
Dehydrated potato shreds 50 50
Dried meant 100 100 100 100
Dry breakfast cereals 50 50
Dry diced fruits 32
Dry mixed for beverages and desserts 50
Dry sausage 30 30 30 30
Emulsion stabilizers for shortenings 200 200
Fresh sausage 100 100 100 100
Potato flakes 50 50
Poultry products 100 100 100 100
Adapted from (51).

TABLE 6. Regulatory Responsibility for Major Antioxidants.
U.S. FDA USDA Canada (NHW) Europe (EEC)
BHA 21 CFR 182.3169 9 CFR 318.7 Table XI, Part IV, B.1, 320 E320
BHT 21 CFR 182.3173 9 CFR 318.7 Table XI, Part IV, B.2, 321 E321
Gallates 21 CFR 184.1660 9 CFR 381.147 Table XI, Part IV, P.1, 324 E310-312
TBHQ 21 CFR 172.185 9 CFR 361.147 Table XI, Part IV, T.1A, 325 Not approved
Tocopherols 21 CFR 182.3890 9 CFR 318.147 Table XI, Part IV, T.2, 325 E306-309
Adapted from (10).
TABLE 7. Regulatory Approval Status of Major Antioxidants in Different Countries.
Antioxidants

Country BHA BHT Gallates TBHQ
Afghanistan
Argentina
Australia
Austria
Bahrain
Barbados
Belgium
Brazil
Chile
China
Columbia
Cyprus
Denmark
Ecuador
Finland
France
Germany
Gibraltar
Greece
Hong Kong
Hungary
Indonesia
Iran
Ireland
Israel
Italy
Jamaica
Japan
Kenya
Korea, South
Luxembourg
Malaysia
Malta
Mauritius
Mexico
Morocco
NATURAL ANTIOXIDANTS 503
Netherlands
New Zealand
Nigeria
Norway
Pakistan
Panama
Papua New Guinea
Peru
Philippines
Portugal
Saudi Arabia
Singapore
South Africa
Spain
Sweden
Switzerland
Taiwan
Thailand
Trinidad/Tobago
Turkey
United Kingdom
Uruguay
Venezuela
Zimbabwe
Adapted from (10).
synthetic antioxidants have been used less frequently in the food and feed industry.
These include ethoxyquin, trihydroxybutyrophenone (THBP), and some secondary
antioxidants such as thiodipropionic acid and dilauryl thiodipropionate (10, 11).
Novel synthetic antioxidants have been created in order to obtain stronger antioxi-
dant activity than that of traditional ones (52). However, the general consumer
rejection of synthetic food additives has led to a decrease in their use and an
increased interest in their replacement with natural ingredients.
3. NATURAL ANTIOXIDANTS
Concerns about the safety of synthetic antioxidants have given rise to a large body
of research on natural sources of antioxidants. Natural antioxidants allow food pro-
cessors to produce stable products with clean labels of all-natural ingredients, as
described by Reiche (11). They bring less rigorous burden-of-safety proof than that
required for synthetic products (11). In addition to their antioxidant activity, some
natural antioxidants, such as vitamins, minerals, and enzymes, are also regarded as
nutrients due to their bioactivity. However, natural antioxidants may possess several
drawbacks, including high usage levels, low antioxidant efficiency, undesirable
flavor or odor, and possible loss during processing (11). The safety of natural
antioxidants cannot be taken for granted because of their potential mutagenicity,

carcinogenicity, teratogenicity, or other pathogenic activities (10). A case in point
is nordihydroguariaretic acid (NDGA), which was removed from the GRAS list and
is no longer of practical use (11, 19). NDGA is a natural constituent of creosote
bush.
Ascorbic acid and tocopherols are the most important commercial natural anti-
oxidants. In addition, many naturally occurring phenolic antioxidants have been
identified in plant sources and vegetable extracts that may lend themselves for
use in a variety of food applications (53). Recent research has focused on isolation
and identification of effective antioxidants of natural origin (11).
3.1. Tocopherols and Tocotrienols

Tocopherols and tocotrienols, collectively known as tocols, are monophenolic and
lipophilic compounds that are widely distributed in plant tissues (7). The main com-
mercial source of natural tocopherols is the soybean oil. Tocotrienols, less common
than tocopherols, are present in palm oil, rice bran oil, as well as cereals and
legumes (11). Tocopherols and tocotrienols are classified into a-, b-, g-, and d-,
depending on their chemical structures (Figure 7). In general, tocotrienols have a
stronger antioxidant effect on lipid oxidation than tocopherols. The antioxidant
activity of tocopherols is dependent on temperature and is in the order of d- >
g- > b- > a-tocopherol (7). Tocopherols (mixed natural concentrate) are a golden
R1
HO
R2 O
CH3
R3
R1 R2 R3
CH3 CH3 CH3 -tocopherol
CH3 H CH3 -tocopherol
H CH3 CH3 -tocopherol
H H CH3 -tocopherol
R1
HO
R2 O
CH3
R3
R1 R2 R3
CH3 CH3 CH3 -tocotrienol
CH3 H CH3 -tocotrienol
H CH3 CH3 -tocotrienol
H H CH3 -tocotrienol
Figure 7. Chemical structure of tocopherols and tocotrienols.

brown colored, slightly viscous liquid with a characteristic odor. However, synthetic
tocopherol (mixed a-, g-, and d-) is a yellow to brownish viscous oily and odorless
liquid (34). Tocols are soluble in vegetable oils but insoluble in water. They func-
tion as a free radical terminator in autoxidation reactions, and they are often used in
food products deficient in natural antioxidants, such as animal fats, waxes, and but-
terfat, among others (34, 54). Tocopherols act synergistically with ascorbic acid,
citric acid, and phospholipids.
As natural antioxidants, tocopherols have GRAS status, and they are regarded as
safe food additives. However, as noted earlier for other antioxidants, excessive
addition of tocopherols may lead to pro-oxidant effects (6, 11). Furthermore, the
hemorrhagic toxicity of a large dose of a-tocopherol has been reported (55). There-
fore, use of tocopherols as antioxidants in foods is subject to regulations. The FDA
(21CFR 182.3890) and the USDA (9CFR 318.147), Canada NHW (Table XI, Part
IV, T.2, 325), and the EEC (E306-309) govern the regulations of tocopherols in
foods in the United States, Canada, and European countries, respectively (10). In
the United States, natural tocopherols are limited to 0.03%, i.e., 300 ppm in animal
fats, and 0.02% in combination with BHA, BHT, and PG (9CFR 318.7). In the UK
and some other European countries, their maximum addition is not to exceed
500 ppm (10, 34). With respect to ADI of a-tocopherol, it has been reported that
an intake of 1000 mg/day is without risk, and 3200 mg/day is without any consis-
tent risk (56). Actually, in the United States, the majority of men and women fail to
meet the current recommendation for Vitamin E intake, according to a recent report
on American diets (57).
3.2. Ascorbic Acid and Ascorbate Salts

L-ascorbic acid (Vitamin C) and its salts (sodium ascorbate and calcium ascorbate)
(Figure 8) are widespread in plant tissues or are produced synthetically in large
quantities (11). Ascorbic acid is a white or slightly yellow crystalline powder
that is extensively used to stabilize beverages, fruits, and vegetables. Its application
in fats and oils, however, is limited because of the insolubility in lipids. It acts as an
antioxidant with multiple functions, including quenching various forms of oxygen,
reduction of free radicals, and regeneration of primary antioxidants (34). The effect
of ascorbic acid on lipid stability in foods is mainly due to synergistic interactions
with other antioxidant compounds (58). It shows excellent synergism with a-toco-
pherol, citric acid, BHA, BHT, and metal chelators. Ascorbic acid strongly inhibits
the depletion of a-tocopherol by regenerating it (6, 59).
CH2OH
H C OH
O
O
H
HO OH
Figure 8. Chemical structure of ascorbic acid.
In addition to antioxidant activity, ascorbic acid also functions as Vitamin C, a

flavorant, an acidulant, a color fixing, and a reducing agent in food products (11,
34). Moreover, it can diminish the generation of odor-active compounds in emul-
sions (60). However, the natural ascorbic acid in foods can be easily destroyed dur-
ing processing as a result of susceptibility to heat, light, pH, oxygen, acrid smoke,
and water activity; thus, it is often added to foods exogenously (61, 62).
Ascorbic acid and ascorbate salts have GRAS status with no usage limits.
According to the literature, Vitamin C is safe at supplementation levels of up to
600 mg/day, and higher levels of up to 2000 mg/day are without risk (56). The
ascorbic acid and its salts carry GRAS status and with minimal associated organo-
leptic problems; thus, they are safe, stable, and good antioxidant candidates for use
in foods (63). As natural or natural-identical products, they are highly recognized as
antioxidant nutrients by consumers (11).
3.3. Carotenoids
Carotenoids are yellow, orange, and red lipid-soluble pigments that occur widely in
plants, fruits, and vegetables. They are 40-carbon isoprenoids with varying struc-
tures (Figure 9), and can be classified as carotenes and xanthophylls (11). Certain
carotenoids are also referred to as pro-vitamins such as b-carotene, a-carotene, and
b-cryptoxanthin. Carotenoids are antioxidant nutrients that act mainly as secondary
antioxidants in foods by quenching singlet oxygen. They may also prevent oxida-
tion by trapping free radicals in the absence of singlet oxygen (11). Carotenoids are
a good synergist with tocopherols. b-Carotene, lutein, lycopene, and isozeaxanthin
are typical carotenoids that effectively retard oxidation in foods. Astaxanthin has
antioxidant activity that is ten times greater than that of b-caroten, lutein, zeax-
anthin, and canthaxanthin, and is often used in fish products (64).
b-Carotene is a purple hexagonal prism or a red leaflet that is often used in fruit
juices, cheese, dairy products, fats, and oils (34). It has poor solubility in most
common solvents, and is highly reactive and unstable to heat, light, pH, oxygen,
and the presence of metals, resulting in limited applications as a food antioxidant
(11). In a high-oxygen concentration, b-carotene may exhibit a pro-oxidant, rather
than an antioxidant effect in food products (61). Carotenoids are natural constitu-
ents of foods and have GRAS status. No limitation on their addition level has been
stipulated.
In addition to the three major classes of natural antioxidants (tocols, ascorbic
acids, and carotenoids), several other natural substances have been identified that
show antioxidant activity through different mechanisms; these include phospholi-
pids, flavonoids, protein hydrolyzates, organic acids, sterols, Maillard reaction pro-
ducts, and enzymes. These are naturally occurring constituents of food and act as
endogenous antioxidants that help prevent oxidation reactions. They are also
regarded as a potential for replacement of synthetic antioxidants. A great deal of
research has been conducted on evaluation of their antioxidant activity and methods
of extraction. A variety of natural products can serve as sources of natural antiox-
idants, among which fruits and vegetables, spices and herbs, oilseeds, and animal
-Carotene
-Carotene
Lycopene
OH
HO Lutein
OH
HO Zeaxanthin
O
Canthaxanthin
O
O
OH
HO Astaxanthin
O
Figure 9. Chemical structures of carotenoids.
and microbial products have been considered (64, 65). Grapes, berry fruits, and
citrus are rich sources of antioxidants (65). Among vegetables, garlic, broccoli,
mushroom, and pulses have been shown to possess antioxidant effects (6567); spi-
nach powder has been reported to be capable of improving lipid stability in deep-
fat-fried products (68). Aside from fruits and vegetables, several studies have con-
firmed that many spice and herb extracts show strong antioxidant activity, such as
rosemary, sage, oregano, cinnamon, thyme, green tea, and evening primrose extract
(65, 6973). Flaxseed, sunflower, soybean, cottonseed, rapeseed, and sesame seed
typify the sources of antioxidants from oilseeds (64, 74, 75). More recently, Shahidi
et al. (76) have reported the antioxidant activity of de-fatted Niger seed extract.
Animal products can serve as good sources of natural antioxidants, such as protein
hydrolyzates (peptides and amino acids), carotenoids, chitosan, and enzymes (64,
77, 78). Furthermore, microbial fermentation is becoming a promising method for
producing natural antioxidants (64).
These antioxidants are all-natural ingredients of foods and have GRAS status.
However, because of the ability of some natural antioxidants to exhibit pro-oxidant
activity, caution should be exercised when adding them to food systems (79).
Furthermore, the safety of natural compounds with antioxidant activity should be
established.
4. CONCLUSIONS
In foods that may undergo oxidation, antioxidants, endogenous or exogenous, func-

tion as an inhibitor to oxidation reactions through various mechanisms. Neverthe-
less, natural antioxidants are deficient in some foods and can easily deteriorate
during processing or in storage, necessitating the use of synthetic antioxidants.
However, most synthetic antioxidants are effective at low concentrations, and the
addition of higher levels may lead to a pro-oxidant effect. Additionally, large doses
of synthetic antioxidants have been reported to impart safety problems. Therefore,
caution must be taken when selecting and adding antioxidants in food systems.
TABLE 8. Antioxidants Conventionally Permitted in Foods.
ascorbic acid, sodium, calcium salts glycine

ascorbyl palmitate and stearate gum guaiac
anoxomer lecithin
butylated hydroxyanisole (BHA) ionox-100
butylated hydroxytoluene (BHT) polyphosphates
tert-butylhydroquinone (TBHQ) propyl, octyl, and dodecyl gallates
citric acid, stearyl, and isopropyl esters tartaric acid
erythorbic acid and sodium salt trihydroxybutyrophenone
ethoxyquin tocopherols
ethylenediaminetraacetic acid (EDTA) thiodipropionic acid, dilauryl and distearyl esters
and calcium disodium salt
Adapted from (64).

REFERENCES 509
TABLE 9. ADI of Some Antioxidants Permitted in Foods.
Antioxidant ADI (mg/kg body weight)
propyl gallate 02.5

BHA 00.5
BHT 00.125
TBHQ 00.2
tocopherols 0.152.0
gum guaiac 02.5
ethoxyquin 00.06
phosphates 070.0
EDTA 2.5
tartaric acid 030
citric acid not limited
lecithin not limited
ascorbic acid not limited
sulfites (as sulfur dioxide) 00.7
ascorbyl palmitate or ascorbyl stearate
(or the sum of both) 01.25
Adapted from (64).
Meanwhile, the safety of natural antioxidants should not be taken for granted as
antioxidants from natural sources are attracting more and more attention. Adher-
ence to regulatory guidelines remains a necessity.
The most common antioxidants permitted for use in foods in most countries are
shown in Table 8. Table 9 presents the ADI of some antioxidants allocated by the
Joint FAO/WHO Expert Committee on Food Additives (JECFA). The food produ-
cer has full responsibility for the choice of suitable antioxidants according to the
corresponding guidelines governed by regulatory laws of the individual country
or the international bodies that declare their safety (64).
REFERENCES
1. K. R. Cadwallader and C. L. Howard, in C. Mussinan and M. Morello, eds., Flavor

Analysis: Developments in Isolation and Characterization, ACS Symposium Series 705,
American Chemical Society, Washington D.C., 1998, pp. 343358.
2. E. H. Farmer, G. F. Bloomfield, A. Sundralingam, and D. A. Sutton, Trans. Faraday Soc.,
38, 348356 (1942).
3. J. L. Bolland and G. Gee, Trans. Faraday Soc., 42, 236252 (1946).
4. L. Bateman, H. Hughes, and A. L. Morris, Disc. Faraday Soc., 14, 190199 (1952).
5. W. W. Nawar, in O. R. Fennema, ed., Food Chemistry, Marcel Dekker, Inc., New York,
1996, p. 255320.
6. M. van Aardt, S. E. Duncan, T. E. Long, S. F. OKeefe, J. E. Marcy, and S. R. Sims,
J. Agric. Food Chem., 52, 587591 (2004).
7. F. Shahidi, Nahrung., 44, 158163 (2000).

8. E. A. Decker, in C. C. Akoh and D. B. Min, eds., Food Lipids: Chemistry, Nutrition and
Biotechnology, Marcel Dekker, Inc., New York, 1998, p. 397.
9. F. Shahidi, in F. Shahidi, ed., Natural Antioxidants: Chemistry, Health Effects and
Applications, AOCS Press, Champaign, Illinois, 1997, pp. 111.
10. D. E. Pratt, in Y. H. Hui, ed., Baileys Industrial Oil & Fat Products, vol. 3, John Wiley &
Sons, Inc., New York, 1996, pp. 524545.
11. D. W. Reische, D. A. Lillard, and R. R. Eitenmiller, in C. C. Akoh and D. B. Min, eds.,
Food Lipids: Chemistry, Nutrition and Biotechnology, Marcel Dekker, Inc., New York,
1998, pp. 423448.
12. D. F. Buck and M. K. Edwards, Food Tech. Int., 2933 (1997).
13. H. H. Hussain, G. Babic, T. Durst, J. S. Wright, M. Flueraru, A. Chichirau, and L. L.
Chepelev, J. Org. Chem., 68, 70237032 (2003).
14. F. Shahidi and M. Naczk, in Food Phenolics: Sources, Chemistry Effects and Applications,
Technomic Publishing Co., Lancaster, Pennsylvania, 1995.
15. S. M. Razavi-Rohani and M. W. Griffiths, J. Food Safety, 16, 5974 (1996).
16. P. Giridhar and S. M. Reddy, J. Food Sci. Tech. India, 38, 397399 (2001).
17. B. J. O. Efiuvwevwere and E. Efi, J. Sci. Food Agric., 79, 17831787 (1999).
18. K. Omura, J. Amer. Oil Chem. Soc., 72, 15651570 (1995).
19. F. Shahidi and P. K. Wanasundara, Crit. Rev. Food Sci. Nutr., 32, 67103 (1992).
20. R. S. Farag, G. S. El-Baroty, and A. Basuny, Int. J. Food Sci. Nutr., 54, 159174 (2003).
21. H. Verhagen, P. A. E. L. Schilderman, and J. C. S. Kleinjans, Chemico-Biological
Interations, 80, 109134 (1991).
22. T. A. Sarafian, S. Kouyoumjian, D. Tashkin, and M. D. Roth, Toxicol. Lett., 133, 171179
(2002).
23. M. Saito, H. Sakagami, and S. Fujisawa, Anticancer Res., 23, 46934701 (2003).
24. C. M. Siman and U. J. Eriksson, Toxicol. Lett., 87, 103108 (1996).
25. R. S. Lanigan and T. A. Yamarik, Int. J. Toxicol., 21, 1994 (2002).
26. H. Verhagen, H. H. G. Beckers, P. A. W. V. Comuth, L. M. Maas, F. Hoor, P. T. Henderson,
and J. C. S. Kleinjans, Food Chem. Toxicol., 27, 765772 (1989).
27. M. H. Lessof, J. R. Soc. Med., 85, 513515 (1992).
28. J. Whysner, C. X. Wang, E. Zang, M. J. Iatropoulos, and G. M. Williams, Food Chem.
Toxicol., 32, 215222 (1994).
29. M. Hirose, S. Takhashi, K. Ogawa, M Futakuchi, T. Shirai, M. Shibutani, C. Uneyama,
K. Toyoda, and H. Iwata, Cancer Lett., 143, 173178 (1999).
30. D. Slamenova, E. Horvathova, S. Robichova, L. Hrusovska, A. Gabelova, K. Keibl,
J. Jakubikova, and J. Sedlak, Environ. Mol. Mutagen., 41, 2836 (2003).
31. G. M. Williams, M. J. Iatropoulos, and J. Whysner, Food Chem. Toxicol., 37, 10271038
(1999).
32. A. Valenzuela, J. Sanhueza, and S. Nieto, Biol. Res., 36, 291302 (2003).
33. B. Halliwell, M. A. Murcia, S. Chirico, and O. I. Aruoma, Crit. Rev. Food Sci. Nutr., 35,
720 (1995).
34. F. Shahidi, in J. Smish and L. Hong-Shum, eds., Food Additive Databook, Blackwell
Science, Ltd., Oxford, U.K., 2003, p. 75.
REFERENCES 511
35. H. Ishiwata, M. Nishijima, and Y. Fukasawa, J. Food Hygienic Soc. Japan, 44, 132143
(2003).
36. G. C. Maziero-C-Baunwart and M. C. F. Toledo, Food Addit. Contam., 18, 363373
(2001).
37. A. A. Botterweck, H. Verhagen, R. A. Coldhohm, J. Kleijans, and P. A. Van-den-Brandt,
Food Chem. Toxicol., 38, 599605 (2000).
38. C. Leclercq, D. Arcella, and A. Turrini, Food Chem. Toxicol., 38, 10751084 (2000).
39. S. S. M. Allan and H. M. A. Mohamed, J. Food Lipids, 9, 277293 (2002).
40. G. C. Yen, C. H. Shao, C. J. Chen, and P. D. Duh, Lebensmittel. Wissenschaft. und.
Technologie, 30, 648652 (1997).
41. T. Okubo, Y. Yokoyama, K. Kano, and I. Kano, Food Chem. Toxicol., 41, 679688 (2003).
42. A. N. Kong, P. Yu, W. Lei, S. Mandlekar, T. H. Tan, and D. S. Ucker, Restor. Neurol.
Neurosci., 12, 6370 (1998).
43. O. A. Badary, R. A. Taha, A. M. Gamal-el-Din, and M. H. Abdel-Wahab, Drug Chem.
Toxicol., 26, 8798 (2003).
44. Y. Li, Aseacat, P. Kuppusamy, J. L. Zweier, J. D. Yager, and M. A. Trush, Mutat. Res., 518,
123133 (2002).
45. H. M. Kim, S. B. Han, W. I. Chang, B. H. Hyun, G. T. Oh, C. J. Ahn, and Y. N. Cha,
J. Toxicol. Sci., 21, 4145 (1996).
46. M. Miyauchi, H. Nakamura, F. Furukawa, H. Y. Son, A. Nishikawa, and M. Hirose,
Cancer Lett., 178, 1924 (2002).
47. A. Valenzuela, J. Sanhueza, and S. Nieto, J. Amer. Oil Chem. Soc., 79, 325328 (2002).
48. T. C. King, I. W. Cheng, and M. G. Johnson, Trends Food Sci. Technol., 9, 168175
(1998).
49. M. Hirose, T. Ito, S. Takahashi, M. Ozaki, T. Ogiso, Y. Nihro, T. Miki, and T. Shitai, Eur.
J. Cancer Prev., 7, 233241 (1998).
50. F. Viklund, J. Alander, and K. Hulk, J. Amer. Oil Chem. Soc., 80, 795799 (2003).
51. N. S. Hettiarchchy and U. Kalapathy, in G. L. Christen and J. S. Smith, eds., Food
Chemistry: Principles and Applications, Science Technology System, 2000, p. 266.
52. C. X. Zhang, H. Wu, and X. C. Weng, Food Chem., 84, 219222 (2004).
53. F. Shahidi, Lipid Tech., 12, 8084 (2000).
54. S. J. Byrd, Cereal Foods World, 46, 4953 (2001).
55. O. Takahashi, Food Chem. Toxicol., 33, 121128 (1995).
56. A. T. Diplock, J. L. Charleus, G. Crozier-Willi, F. J. Kok, C. Rice-Evans, M. Roberfroid,
W. Stahl, and J. Vina-Ribes, Br. J. Nutr., 80, 77112 (1998).
57. J. E. Maras, O. I. Bermudez, N. Qiao, D. J. Bakun, E. L. Boody-Alter, and K. Tucker, J. Am.
Diet. Assoc., 104, 567575 (2004).
58. M. Makinen, M. Kahkonen, and A. Hopia, Eur. J. Lipid Sci. Tech., 103, 683687 (2001).
59. K. Hiramoto, Y. Miura, G. Ohnuki, T. Kato, and K. Kikugawa, J. Oleo Sci., 51, 569576
(2002).
60. S. M. van-Ruth, J. P. Roozen, M. A. Posthumus, and F. J. H. M. Jansen, J. Amer. Oil Chem.
Soc., 76, 13751381 (1999).
61. J. Lee, N. Koo, and D. B. Min, Comprehensive Rev. Food Sci. Food Safety, 3, 2133 (2004).
62. S. T. Mayne, J. Nutr., 133, 933940 (2003).
63. J. G. Elliott, Food Technol., 53(2), 4648 (1999).

64. K. Mikova, in J. Pokorny, N. Yanishlieva, and M. Gordon, eds., Antioxidants in Food:
Practical Applications, Woodhead Publishing, Ltd., Cambridge, England, 2001, pp. 267
283.
65. C. Kaur and H. C. Kapoor, Int. J. Food Sci. Technol., 36, 703725 (2001).
66. F. Shahidi, U. D. Chavan, M. Naczk, and R. Amarowicz, J. Agric. Food Chem., 49, 926
933 (2001).
67. R. Amarowicz, M. Naczk, R. Zadernowski, and F. Shahidi, J. Food Lipids, 7, 195205
(2000).
68. J. Lee, S. Lee, H. Lee, K. Park, and E. Choe, J. Agric. Food Chem., 50, 56645669 (2002).
69. J. Mancini-Filho, A. Van-Koiij, D. A. Mancini, F. F. Cozzolino, and R. P. Torres, Boll.
Chim. Farm., 137, 443447 (1998).
70. F. Shahidi, Lipid Tech., 12, 8084 (2000).
71. F. Shahidi, R. Amarowicz, Yuehua-He, and M. Wettasinghe, J. Food Lipids, 4, 7586
(1997).
72. M. Wettasinghe and F. Shahidi, J. Agric. Food Chem., 47, 18011812 (1999).
73. F. Shahidi and D. M. Alexander, J. Food Lipids, 5, 125133 (1998).
74. R. Amarowicz, B. Raab, and F. Shahidi, J. Food Lipids, 10, 5162 (2003).
75. F. Shahidi, R. Amarowicz, H. A. Abou-Gharbia, and A. A. Y. Shehata, J. Amer. Oil Chem.
Soc., 74, 143148 (1997).
76. F. Shahidi, C. Desilva, and R. Amarowicz, J. Amer. Oil. Chem. Soc., 80, 443450 (2003).
77. F. Shahidi and R. Amorowicz, J. Amer. Oil. Chem. Soc., 73, 11971199 (1996).
78. F. Shahidi, J. Kamil, You-Jin-Jeon, and Se-Kwon-Kim, J. Food Lipids, 9, 5764 (2002).
79. H. T. Osborn and C. C. Akoh, J. Amer. Oil Chem. Soc., 80, 847852 (2003).
13
Toxicity and Safety
of Fats and Oils
David D. Kitts
University of British Columbia
Vancouver, British Columbia, Canada
1. INTRODUCTION
For the past half centrury, healthy eating concepts have evolved around
avoiding fat. This is demonstrated in part from the USDAs dietary guideline
handbook, which recommends that fats and oils be consumed sparingly. At
present, fat accounts for approximately 35% of the calories in a standard
North American diet. With many campaigns directed at low-fat diets to protect
consumer health and heart, a current situation exists where efforts to reduce
total fat intake have resulted in a shift away from fat foods to high carbo-
hydrate diets that, in turn, contain similar potential health-related concerns.
For example, an increased awareness to the hazards of possible elevations in
triacylglycerols attributed to a high carbohydrate intake has surfaced with
syndrome- X profiles and increased risk to heart disease. It is vitally important
to acknowledge the importance of fat in the diet as a principal source of
energy, essential fatty acids, and fat-soluble vitamins, in addition to expressing
513
514 TOXICITY AND SAFETY OF FATS AND OILS
a concern about potential adverse reactions attributed to specific components

in fats and oils, that may initiate or promote disease. Moreover, dietary fats
provide important organoleptic qualities and a strong satiety signal. Fat is a
primary component of cell membranes; for example, the brain alone represents
70% fat.
Safety of fats and oils has mostly centered on the very visible components of
the total crude lipid fraction of foodstuffs, such as saturated fat and cholesterol
and products of lipid hydrogenation and oxidation. Since the increased awareness
of health risks attributed to dietary fats that started in the 1950s by Ancel Keys
from University of Minnesota, who initially described a link between dietary
saturated fatty acid and cholesterol-induced increase in blood cholesterol,
much interest has been given to the source of dietary fat. Saturated and trans-fatty
acids and cholesterol accelerate atherogenesis, whereas monounsaturated and
polyunsaturated fatty acids (PUFA) are associated with reduced incidence of
coronary heart disease. With the greater awareness that consumption of n-3
PUFA exert a protective potential against heart disease by lowering concentra-
tions of very low-density lipoprotein, triacylglycerides and reduced blood
pressure, platelet aggregation, and thrombosis (1), we are now appreciating
that not all fats are equal in health implications. Moreover, relatively less visible
components of dietary fat intake may have an even greater role in preventing,
or alternatively initiating and propagating, chronic diseases. For example, the
introduction of plant sterol and stanol esters into the diet has been shown to
effectively lower serum cholesterol and low-density lipoproteins (2); reliable
biomarkers for coronary heart disease. Vitamin E is another example of an
important lipid soluble nutrient that has antioxidant activity and may, in concert
with Vitamin C, protect against LDL oxidation, vascular endothelial dysfunction,
and atherosclerosis (3). Alternatively, fat-soluble hydrocarbons with toxic
potential must also be realized. Examples of these include the heterocyclic
amines derived in cooked foods or the naturally present mycotoxins (e.g.,
aflatoxins) that represent important determinants of genotoxicity in humans
(4) and manmade organic pollutants (e.g., polycyclic aromatic hydrocarbons
(PAHs), polybrominated diphenyl ethers (PBDEs), and organochlorine pesticides
(OPs) that infilter the food chain and bioaccumulate to notable levels in foods
consumed by man (e.g., fish and seafoods (5).
The use of chemical aids and technologies to stabilize lipids also represents
a need to evaluate the balance between positive attributes that may reduce
the risk of exposure to dietary oxidized lipids, or alternatively, negative
consesquences, such as generation of trans-fatty acids derived from selec-
tive hydrogenation of vegetable oils. This chapter is intended to update
the information on topics of toxicity and safety of fats and oils described
earlier (6), as they relate to: (1) natural consitutents of fats and oils; (2)
derived products of oxidation and hydrogenation; (3) occurance of natural
and pollutant contaminants; and (4) additives used to preserve the stability,
functionality, and nutritional quality of many constituents present in fats and
oils.
ADVERSE EFFECTS OF FATS AND ASSOCIATED CONSTITUENTS 515
2. ADVERSE EFFECTS OF FATS AND ASSOCIATED

CONSTITUENTS
2.1. Total Fat Intake

The level of dietary fat intake can represent both an initiator and a promoter of many
adverse conditions that lead to a health risk (Figure 1). For example, a relationship
between dietary fat intake and oxidative status will influence gene expression for
drug-metabolizing enzymes, such as phase I and phase II enzymes, as well as glu-
cose-6-phosphate dehydrogenase (G6PDH). Feeding diets containing 20% soybean
oil to rats produces greater reductions in G6PHD and glutathione peroxidase activ-
ities compared with counterparts receiving 5% soybean oil (7, 8). The subsequent
availability of reducing equivalents and glutathione for antioxidant enzyme activity
are compromised by the higher fat intake. A completely separate example is seen
with exposure to polychlorinated biphenyls (PCB)/dioxins, where the most impor-
tant route for human exposure is food consumption (e.g., >90% of total exposure).
The concentration of these particular contaminants in herbage consumed by cows is
controlled by atmospheric deposition, and cows that consume large quantities of
herbage can produce milkfat that contains PCB and dioxin contaminants. There-
fore, exposure to these specific lipid-soluble xenobiotics in milkfat sources from
certain geographic regions will be related to the level of intake of this particular
fat source. Total intake of dioxin from dairy fat-containing products is 30% for
adults and 50% for children. Background levels of PCBs are actually estimated
from typical dietary intakes of food stuffs that include milk, eggs, meats, and
fish-lipid matrices that contain these environmental contaminants (9). Using this
trend of thought, high intakes of dietary lipids that contain lipid oxidation products
in various quantities influence oxidative status, and, therefore, high intakes of total
fat could increase the probability of exposure to exogenous- and endogenous-
derived lipid oxidation products. Lipid oxidation yields a complex mixture of
byproducts that include hydroxyl and dihydroxy fatty acids, hydroperoxides, vola-
tile aldehydes, and alkyl and olefinic radicals (10), which are absorbable through
the digestive tract (11) and incorporated into membrane phospholipids whereby
they alter membrane fluidity (12, 13). The effect of feeding high-fat diets contain-
ing thermally oxidized lipids can lead to peroxidized tissue proteins (14), an upre-
gulation of peroxisomal-proliferator-activating receptor alpha (15) and increased
number of intestinal aberrant crypts that are indicative of a precancerous cellular
event (16). Thus, in these examples, both the quanity and the quality of the dietary
lipid present a potential health hazard for the consumer.
2.2. Saturated Fatty Acids

Medium-chain fatty acids are saturated fatty acids because of the relatively shorter
hydrocarbon chain, which does not facilitate unsaturation. The safety of medium-
chain triacylglycerol (MCTs) in dietary oil has been debated, and associated effects
on cholesterol metabolism remain unclear. Although some studies have shown that
516
Figure 1. Schematic diagram shows the interaction between lipid constituents, products of lipid oxidation, and xenobiotics, which can initiate or promote
chronic disease states (6).
MCTs are essentially nontoxic, noncarcinogenic, and nonmutagenic for human con-
sumption with a safety level up to 1 g/kg (17), other studies have indicated that MCT
oil-containing diets can increase blood cholesterol levels (18). MCTs, on a percent
energy basis, have half the potency of palmitic acid (C16:0) in raising plasma
cholesterol (18). Palmitic acid (C16:0) can lead to increases in blood cholesterol
levels; however, when ingested in a diet that contains a recommended intake of
C18:2,n-6, the effect on both total and LDL cholesterol levels are minimized
(19). This has been shown with fat blends, such that hypercholesterolemia
was not observed in animals fed either butter or tallow fat sources that were blend-
ed with soybean oil in a low-cholesterol-containing diet (20, 21). In gerbils
and monkeys, the relative ratio between C14:0 to C18:2 n-6 fatty acids as well
as dietary cholesterol are important factors in modulating increases in serum
cholesterol levels (22, 23).
Intake of saturated fat sources has also been associated with insulin resistance,
leading to altered glucose metabolism, type II diabetes, and impaired glucose
tolerance (24). Comparatively, saturated fat has a more deleterious effect on
fat-induced insulin sensitivity than both mono- and polyunsaturated fat sources (24).
Higher intakes of saturated fat and trans-fat adversely affect glucose metabolism
and insulin resistance, whereas higher intakes of polyunsaturated fat and possibly
long-chain n-3 fatty acids are beneficial (25). Within the category of saturated
fats, dietary saturated, short-chain, and o6 fatty acids have been found to have the
most deleterious effects on insulin action associated with insulin sensitivity, as
opposed to medium- and long-chain fatty acids and o3 fatty acids (26). Intramuscular
triacylglycerol (MTG) and elevated plasma free fatty acid (FFA) levels also have
roles in insulin-mediated glucose uptake, reflecting a pivotal role of the high
saturated fatty acid content in the MTG (27). Changing dietary fat quality by
substituting saturated for monounsaturated fat can impair insulin sensitivity,
as saturated fat has a greater deleterious impact on insulin sensitivity (28). For
example, substituting a monounsaturated fatty acid diet (MUFA diet) for a saturated
fatty acid diet (SAFA diet) has been shown to be favorable for only those subjects
that had a lower-than-average total fat intake. This intervention improved insulin
sensitivity, but had no effect on insulin secretion. Notably, the addition of n-3 fatty
acids to MUFA and SAFA diets affected neither insulin secretion nor insulin
sensitivity (28).
2.3. Monounsaturated Fatty Acids (MUFA)

Dietary MUFA have been found to have several positive effects that include, in
addition to lowering human LDL-cholesterol plasma levels, positive effects on lipo-
protein oxidation, coagulation, and fibrinolysis (29). Low-fat, monounsaturate-
rich diets reduce the susceptibility of low-density lipoproteins to peroxidation
ex vivo (30). In hypercholesterolemic subjects, diets containing MUFA resulted
in favorable alterations in the fatty acid composition and oxidative profile of LDL
in hypercholesterolemic subjects that were characterized as an increase in lipid
peroxide lag time and a decrease in lipid peroxide formation (30). In animal studies,
rats that were fed long-chain MUFA diets showed only a small, significant increase
in peroxisomal b-oxidation, and a slight decrease in mitochondrial oxidation (31).
Feeding low-fat, monounsaturated-rich diets that contain high oleic peanuts has
been shown to improve human serum lipoprotein profiles (32) and human serum
lipid profiles (33). In free-living subjects with impaired glucose tolerance, MUFA
diets also seemed to improve glucose metabolism (33).
Several studies have examined the effect of MUFA-containing diets on risk for
cardiovascular disease (34). MUFA-containing diets lower both plasma cholesterol
and triacylglycerol concentrations, which has favorable effects on the cardio-
vascular disease risk profile (35). MUFAs can also modify the lipoprotein profile
and the mechanism by which fatty acids affect the immune response, which
in turn will alter the development of the atherosclerotic lesion by limiting arterial
thrombus formation (36). The effects of MUFA on human immune system
responses have also been considered. Animals fed MUFA-rich diets, such as those
containing olive oil, exhibit a suppressed in vivo immune response; the reactions in
humans subjects are far more subtle than those reported in animal studies, however
(37). The level of MUFA contained in animal feeding studies is much higher than
that achievable in human studies, so MUFA still has a negligible effect on the
modulation of immune function in humans (37).
2.4. Polyunsaturated Fatty Acids (PUFA)

In a comparison of rats fed diets containing either safflower oil or fish oil, it was
shown that neither fish oil nor safflower oil intake resulted in an induced increase in
phospholipid hydroperoxides and TBARS in rat organs. Thus, the supplementation
of a rich source of n-3 PUFA, such as fish oil, compared with an n-6 PUFA containing
safflower oil, did not significantly alter lipid peroxidation in rat organs (38). Other
studies have investigated the comparative hypocholesterolemic effects of fish oil
and soybean or safflower oil, among others, such as evening primrose oil (EPO)
and oenothera biennis linn oil (OBLO), in cholesterol-fed rats after long-term
feeding (39). Both normotensive and hypertensive rats fed menhaden oil-containing
diets exhibited lower plasma cholesterol and triacylglycerol concentrations than
counterparts fed either saturated (e.g., butter) or n-6 (e.g., soybean oil) diets
(20, 21). In other studies where both EPO and OBLO contain n-6 PUFA,
gamma-linolenic acid (GLA), it was found that OBLO (linoleic GLA) and
EPO (linoleic GLA) caused the lowest serum total cholesterol, VLDL-c IDL-
c LDL-c concentrations (39). EPO has also been found to have both antisecretory
and antiulcerogenic effects in rats by inhibiting gastric mucosal damage induced by
pylorus ligarion, NSAIDS or hypothermic restraint ulcers (40). Crude extracts of
meals from evening primrose sources have been compared with crude extracts
of meals obtained from borage, for effect of concentration-dependent scavenging
of reactive oxygen species and 1,1-diphenyl-2-picrylhydrazl (DPPH) free radical
(41). Along with EPO, other GLA oils, such as black current, borage, and fungal
oils, have been noted for their blood pressure lowering effect in spontaneously
hypertensive rats. This is likely caused by the alterations in fatty acid profiles of
hepatic and vascular tissue induced by GLA oils (42). A similar result was not
observed in spontaneously hypertensive rats fed menhaden oil (21). However, inter-
esting comparative effects of feeding rapeseed (canola) oil and soybean oil on blood
pressure of rats showed that systolic blood pressure of rats fed canola oil diets was
higher than that of rats fed soybean oil diets. One explaination for this finding was
that the intake of canola oil increased plasma sodium and lipid levels and decreased
potassium levels compared with soybean oil intake (43). Although the beneficial
effects of GLA have been investigated in depth, it is unclear which oils are the
best sources of GLA. For example, rats fed equal amounts of GLA, obtained either
from transgenic canola plant or from borage plant, reacted similarly in growth and
hepatic metabolism of n-6 fatty acids (44).
In other animal studies, where GLA was compared with alpha-linolenic acid
(ALA), ALA gave similar effects as GLA in increasing fatty acid oxidation activity
in rat livers. GLA and ALA differ, however, in the mechanism of action, as
evidenced by the different affect on individual fatty acid oxidation enzymes
involved in fatty acid oxidation (45). These results were confirmed by studies
that found that, in opposition to dietary saturated fatty acids, both n-6 and n-3 fatty
acids, such as GLA and ALA, inhibit the increase of serum total cholesterol and
VLDL IDL LDL-c concentrations (46). The n-3 long chains PUFAs, EPA,
and DHA have also been found to affect lipid peroxidation. The ex vivo intake
of these highly purified n-3 fatty acids caused an immediate increase in chylo-
micron peroxidation in plasma (47). Moreover, feeding n-3 PUFA-rich fish oil in
the form of menhaden oil reduced both RBC and heart GSH-Px activities (20).
The same n-3 PUFA source, when fed at twice the energy equivalent, also can
reduce RBC glutathione content and result in enhanced lipid oxidation in RBC,
heart, and liver tissues (21). More specifically, n-3 PUFAs EPA and DHA have
triacylglycerol-lowering effects (48). EPA and DHA have also been shown to
reduce the incidence of mammary tumors that developed in rats. In this case,
DHA was found to be slightly more effective than EPA (49).
The relationship between n-3 PUFA and protection against coronary heart
disease is closely related to the hypocholesterolemic and hypertriaclglyceridemic
responses to consuming this PUFA source (50). Intake of n-6 and n-3 PUFA has
been shown to reduce hepatic mRNA and protein levels and the synthesis and
activity of G6PHD. The result (51, 52) is reduced hepatic lipogenesis and plasma
triacylglycerol content. In addition, n-3 PUFA reduces platelet aggregation and
exhibits antithrombotic and fibrinolytic activies, among other functions (53).
Several other studies have also been performed to investigate the antiatherogenic
potential of GLA (54). GLA has been noted for its positive effects in the
modification of atherosclerotic lesions in apolipoprotein E knockout mice. The
reduced atherosclerotic lesion size from feeding GLA or n-3 was related to a
suppressed smooth muscle cell proliferation in vivo and retarded development of
diet-induced atherosclerosis (55). In humans, GLA supplementation may also
affect cancer cell proliferation via the modification of fatty acid composition. A
relationship appears to exist between GLA-induced tumor cell death and the distri-
bution of fatty acids in tumor cells (56). GLA also has anti-inflammatory properties
in humans. The addition of GLA in vitro suppressed IL-1 beta release from human
monocytes stimulated with LPS. GLA simultaneously reduced the amplification
process of IL-1 beta and left the initial IL-1 beta response to LPS intact (57).
The anti-inflammatory and immune parameter effects of GLA have also been
reported in rats where GLA induced improvement of inflammatory disorders
through the regulation of eicosanoid production. High doses of GLA may exert an
anti-inflammatory affect by suppressing leukotriene B4 release and by strengthening
the gut immune system, therefore improving responses to allergic reactions
(58). As GLA is also known to have anticancer properties, studies have been
performed on the effect of GLA on the expression of maspin and the motility
of cancer cells. Maspin is a tumor suppressor and GLA has been found to
upregulate the expression of mapsin, thereby reducing the motility of cancer cells
(59). Serum levels of phospholipids dihomo-gamma-linolenic acid were also found
to be inversely associated with the risk of death caused by lung cancer (60). In dia-
betic rats, GLA was found to have beneficial, restorative effects on nerve conduc-
tion velocity, Na, K ATPase activity, and membrane fatty acid composition (61).
Moreover, bleomycin-induced lung fibrosis in hamsters was altered as a result of
elevations in tissue PGE1 and 15-HETrE, both of which have anti-inflammatory
properties (62).
In other studies, very long-chain n-3 PUFA, such as those found in fish oil, were
suspected of having a hepatotoxic potential in rabbits, which, in turn, was associat-
ed with atherosclerosis (63). The feeding of fish oil to rabbits produced an n-3
PUFA concentration dependent increase in aortic plaque surface area, indicating
a potential positive relationship between severity of liver pathology and aortic
plaque surface area. When fish oil and corn oil were compared for relative effects
on chemical-induced hepatic enzyme-altered foci in rats, it was found that dietary
fish oil inhibited hepatic enzyme-altered foci formation compared with corn oil.
This is important because of possible stimulation of the hepatic detoxification
system and enhancement of lipid peroxidation caused by excessive intake of fish
oil (64). Conversely, marine oil containing n-3 polyunsaturated fatty acids (PUFAs),
eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) will induce hepatic
peroxisomal b-oxidation and upregulate hepatic antioxidant enzymes (e.g., catalase
and glutathione peroxidase and reductase activities) and GSH concentration (65).
Different doses of arachidonic oil (C20: 4 (n-6)) have recently been compared
with high doses of C20: 4 (n-6) fish oil with DHA for in utero exposure effects in
rats. Higher dosages of C20: 4 (n-6) and C20: 4 (n-6) DHA decreased alkaline
phosphatase activity, lowered serum cholesterol, triacylglycerol, and phospholipids
concentrations, while increasing creatinine and urea concentrations and also
adrenal, spleen, and liver weights of pups (66). ARASCO, a microfungal source
of triglyceride oil enriched in C20: 4 (n-6) was evaluated for its safety in rats.
Results showed that high doses of ARASCO increased C20: 4 (n-6) levels in the
brain, heart, and liver, thus being readily incorporated into tissue lipids, but
without developmental, histopathological, or neuropathological consequences
(67). Consistent with these findings regarding the nontoxicity of ARASCO, no
genotoxic effects were observed (68).
2.5. Sterols
2.5.1. Cholesterol Experiments from laboratory animal trials (21, 69, 70) have
supported epidemiological studies (71) that link hypercholesterolemia and
hyperlipoproteinameia, two risk factors for CVD, with dietary cholesterol intake
or atherogenic fatty acid ratios. Common to many of these studies are the findings
that consumption of diets rich in cholesterol or saturated fat will result in a
reduction of LDL receptors and elevation of LDL cholesterol and total cholesterol.
Studies in nonhuman primates have indicated that the cellular events and tissue
morphological changes that occur during the progression of atherosclerosis are simi-
lar, whether animals are fed diets inducing modest levels of hypercholesterolemia
or inducing extremely high levels of cholesterolemia (72). Studies to characterize
events involved in the initiation of atherogenesis have suggested the possible
involvement of cytotoxic cholesterol oxides in this disease process (73). Moreover,
cholesterol and its oxide derivatives have been identified as a significant component
of arterial plaque composition in humans (74) and in various animal species (75).
There are relatively few studies that have examined the significance of dietary
cholesterol intake on aortic plaque composition in atherosclerosis susceptible
animals. The characteristic changes in plasma lipids and aortic plaque composition
in the atherosclerosis susceptible Japanese quail model has enabled the effects of
cholesterol feeding to be evaluated in regard to susceptiblity to atherosclerosis
(69). Herein, the Wistar rat, which is a noted animal model resistant to atherosclero-
sis, will be used for comparison with the atherosclerosis susceptible quail in many
aspects of cholesterol metabolism and development of atherosclerosis.
Japanese quail exhibit greater plasma total cholesterol concentrations than rat
counterparts when fed a similar, basal low-cholesterol diet (Table 1). Species dif-
ferences in plasma lipids are more pronounced when fed an atherogenic diet (e.g.,
cholesterol/cholic acid supplemented and saturated fat-containing diet), as evidenced
by the marked elevations observed in quail plasma total cholesterol concentrations
TABLE 1. Plasma Lipids and Aortic Plaque Score and Area Covered in Wistar Rats and
Atherosclerosis Susceptible Japanese Quail Fed Low and High Cholesterol Diets.
Total Plaque Area Covered

Cholesterol Triacylglycerides2 Score (%)3
Cholesterol Level
(% by weight): 0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5
Animal species:
Quail 230.4 9{ 2049 95*{ 128 10 359 31 N.D. 3.7 0.2{ N.D. 61 10{
Rat 118.2 5 240 5* 122 18 87 6 N.D. N.D. N.D. N.D.
1
Values represent mean SEM, n 8. Animals fed tallow fat, semisynthetic diets with low (0.05%) and high
(0.5%) cholesterol added.
2
mg/dL.
3
Plaque score based on scale of 0 (N.D.) clean surface; 1 5 plaques; 2 620 plaques; 3 >20
plaques; 4 massive atheromas observed. Values represent two judges evaluating in a blinded protocol.
4
Area covered (%) percent of aortic epithelium covered by plaque, range 0 (N.D.)100% (69).
*
A significant (p 0.05) difference between cholesterol levels.
{
Significant difference between specie.
compared with the rat (Table 1). Dramatic elevations observed in plasma cholesterol
of quail fed cholesterol/cholic acid supplemented diets are directly associated
with the level of dietary cholesterol fed to birds and severity of atherosclerosis
lesions (Table 1). Although the rat also exhibited increased plasma cholesterol
levels when fed the atherogenic diet, absolute cholesterol plasma levels are
much lower than the quail, thus possibly reflecting the resistance of this species
to the induction of hypercholesterolemia from cholesterol feeding. The concomitant
hypertriacylglyceridemia observed in quail fed the atherogenic diet, did not occur
in the rat fed a similar atherogenic diet and contributes further to explaining the role
of dyslipidemia to the susceptibility of quail to atherosclerosis.
As a result of hypercholesterolemia, an intracellular accumulation of cholesteryl
esters occurs, which is associated with the initiating events of atherogenesis, in
particular, foam cell generation, which involves the intracellular accumulation of
large amounts of cholesterol within macrophages and smooth muscle cells of
aorta (76).
Modified (oxidized) lipid species have been identified in the plasma lipoproteins
and aortic plaque of atherosclerotic humans (74, 77) and animal models (78, 79).
Furthermore, the presence of COPs in circulating lipoprotein has been demonstrated
in healthy humans (80) and monkeys. (73) Oxidized LDL has been proposed to have
a role in foam cell formation (81) as well as having various proatherogenic properties,
such as cytotoxicity and chemotactic activity (73, 82).
2.5.2. Bile Salts Bile salts have been linked to colorectal carcinogenesis by a
variety of mechanisms (83). On the other hand, bile salt-dependent lipase (BSDL),
a digestive enzyme secreted by the pancreas, which hydrolyzes dietary lipid esters,
can have a positive effect against atherosclerosis (84). This is because BSDL
activity increases with the level of LDL-c and is also positively linked to serum
concentration of ApoB100 and ApoA-I. BSDL is associated with LDL in part by
a specific interaction with ApoB100, although there appears to be no interaction
with ApoA-I. As the increase in LDL-c is a risk factor for atheroma, the associated
increase in BSDL, which can metabolize atherogenic LDL, shows that BSDL can
have the positive effect against atherosclerosis (84). In another study, in which the
cytotoxicity of bile salts against biliary epithelium (BDE) in isolated perfused
rat liver was evaluated, it was found that in vitro BDE cells are not damaged by
taurine-conjugated bile salts or glycine-conjugated bile salts, but they are very
sensitive to the cytotoxicity of hydrophobic unconjugated bile salts (85).
2.5.3. Phytosterols The relationship between total dietary phytosterol content

and the fatty acid composition of the diet decreases with increasing saturated fatty
acids, whereas the total dietary phytosterol content increases with increasing PUFA
(86). Phytosterols consist of a mixture of cell membrane constituents that include
free sterols, esterified steryls (e.g., esterified to phenolics), steryl glycosides, and
acylated steryl glycosides (Table 2).
They are found in varying concentrations in a number of vegetable oils, cereal-
based products, and nuts where they function to regulate membrane fluidity
TABLE 2. Principle Sources of Plant Sterols.1
Sterol Source Concentration (g/kg)
Oil
Rapeseed 5.139.79
Corn 8.0915.57
Wheat germ 19.70
Rice bran 32.35
Cereals
Rye 1100
Barley 830
Wheat 760
Oats 520
1
Ref. (87).
and permeability. Plant sterols (e.g., -sitosterol, stigmasterol, campesterol, -7

stigmasterol, and -5-avenasterol) are present in plant unsaponifiable matter
(Figure 2). The main sterol in vegetable oils is 4-desmethylsterol, where sitosterol
is the principle sterol and camperstserol, stigmasterol, brassica-, and -venasterol
are also present in lower concentrations. Total sterol content of soybean oil and
safflower oil is 0.29% and 0.39%, respectively, and varies with the source materials
(Table 3). This compares with the plant sterol content of 4.1% in rice bran and 10%
in rubber seed oils.
A typical Western diet contains approximately 100300 mg and 2050 mg of
plant sterol and plant stanol, respectively. The relationship between total dietary
phytosterol content and the fatty acid composition of the diet decreases with
increasing saturated fatty acids, whereas the total dietary phytosterol content
increases with increasing PUFA (89). Fortification of lipid foods, such as
margarine, with plant sterols will dramatically increase the daily intake of
phytosterols and significantly lower serum cholesterol (90). The dietary consumption
of large amounts of plant sterols will interfere with cholesterol absorption, thereby
leading to an increased daily neutral steroid excretion.
Fat-soluble plant stanols have been approved for fortification into margarine.
The sterol solubility is obtained by trans-esterification with rapeseed oil fatty acids
in the margarine (91). Plant stanol esters lower serum cholesterol more effectively
than insoluble free stanol counterparts, which are either in a crystalline, microcrys-
talline, powered, or homogenized form (92). The trans-esterification process also
ensures palatably, with the incorporation of stanols into nutritional fats that are
common to a Western diet, such as margarine and mayonnaise (91). Two FDA-
approved margarines, Take Control (Unilever) and Benecol (Johnson & Johnson)
reached the marketplace in May of 1999. The claim on the Finnish margarine,
Benecol, states that 23 servings per day of the product, containing 1.5 g plant
stanol esters per serving, will lower cholesterol in 2 weeks. In animal studies, -
sitosterol feeding interferes with exogenous and endogenous cholesterol absorption,
resulting in the interruption of enterolymphatic circulation of cholesterol, an impor-
tant regulator of hepatic cholesterol synthesis. Less than 1 g/d of sitostanol esters
OH OH
Campesterol Sitosterol
OH OH
Campesterol Stigmasterol
OH
Stigmasterol
Figure 2. Plant sterols found in oil seed, vegetable, and wood sources. Of the more than 40
plant sterols identified, b-sitosterol, stigmasterol, and campesterol are the most abundant.
incorporated into mayonnaise significantly lowered cholesterol absorption (92), and

a median intake of 23 g/d of plant stanol esters have been recommended to lower
serum LDL cholesterol by 1015% (93). The mechanism of action of phytosterol-
induced reduction of serum cholesterol involves interference of the relatively inso-
luble sterol with micellar solubility of cholesterol in the intestinal lumen. Plant sta-
nol esters also mix easily with the oil phase of intestinal content, which, in turn,
disrupts cholesterol absorption (93). Plant stanols can also prevent the absorption
of other plant sterols (91).
A compensatory increase in cholesterol synthesis attributed to the phytosterol-
induced lowering of serum cholesterol occurs, as indicated by a rise in serum
TABLE 3. GC Quantitation of Cholesterol and Cholesterol Oxidation Products in Aortic

Tissue from Wistar Rats and Atherosclerosis Susceptible Japanese Quail Fed Low and
High Cholesterol Diets.1
Sterol2: (mg/g) Cholesterol 7b-hydroxycholesterol 7-ketocholesterol

Cholesterol level
(% by weight): 0.05 0.5 0.05 0.5 0.05 0.5
Animal species:
Quail 3.19 0.91 11.48 2.13* N.D. 0.24 0.03 N.D. 0.27 0.04
Rat 0.93 0.04 0.91 0.49 N.D. N.D. N.D. N.D.
1
Values represent mean SEM, n 8. Animals fed tallow fat, semisynthetic diets with low (0.05%) and high
(0.5%) cholesterol added.
2
Values expressed on basis of tissue wet weight. See (69) for methodology.
N.D. none detected.
*
A significant (p 0:05) difference between cholesterol levels across a row.
cholesterol precursors, lathosterol and desmosterol. This compensatory rise in serum

cholesterol caused by absorption inhibition is not necessarily detrimental because
the net effect is still a lowering of serum cholesterol (93). The methyl or ethyl side
chain on the cholesterol molecule greatly reduces the intestinal absorptibility
of plant sterols. The efficiency of sitostanol absorption is only 03%, with serum
levels being almost undetectable. Campestanol, the other major plant stanol,
also has a very low absorption efficiency compared with the unsaturated version,
vampestanol (93).
Recently, a seven-part study was conducted that served as a safety evaluation
of phytosterol esters. It was shown that phytosterols do not bind to the estrogen
receptor (ER) and do not stimulate transcriptional activity of the human ER in a
recombinant yeast strain. There was also no indication of estrogenicity from
an uterotrophic assay conducted in immature female rats (94). In a rat study,
the no-observed-adverse-effect level (NOAEL) of phytosterol esters was 8.1%
following a daily oral administration for 90 days, equivalent to a dose of
6.6 g/kg/day. No treatment-related changes that were of toxicological significance
occurred in this subchronic study (95). In the ensuing two-generation reproduction
study in rats with phytosterol esters that followed, it was found that no effect on
the reproduction of parental generation rats, or on the development of second-
generation pups and their sexual maturation of the pups, occurred at 8.1% concen-
tration of phytosterol esters (96). The fecal concentrations of bile acids and neutral
sterols from healthy normolipidaemic volunteers consuming a controlled diet,
either with or without a phytosterol ester-enriched margarine, indicated that a
high intake of phytosterol esters increased the amount of neutral sterols in the feces,
but did not result in increased formation of bile acids or sterol metabolites (97).
Concomitantly, the effect of feeding a controlled diet, either with or without the
phytosterol ester-enriched margarine, on fecal short-chain fatty acid and microflora
content, fecal bacterial enzyme activity, and serum female sex hormones, indicated
that the daily consumption of 8.6 g phytosterol did not affect bacterial profile or
metabolic activities of the gut microflora. Moreover, there was no biologically
relevant effect on serum female sex hormones (98). Examination of potential

mutagenic activity of phytosterols (99) and the comparative absorption and tissue
distribution of phystosterols in the rat (100) disclosed that the most highly absorbed
sterol was cholesterol, followed by campesterol, b-sitosterol, and stigmasterol,
follwed by b-sitostanol and campestanol. The absorption of phytosterols was
slightly greater in females than in males (100). None of the phytosterols and
phytosterol esters tested showed any evidence of mutagenic activity in bacterial
mutation assays, in vitro chromosome aberration assays, or in vitro mammalian
cell gene mutation assays (99).
A positive effect of phytosterols from tall oil phytosterols showed a dramatic
protection against atherosclerosis in mice (101). Mice fed a cholesterol-enriched
diet containing a phytosterol mixture exhibited reduced plasma cholesterol levels
and decreased formation of atherosclerotic lesions. In humans, phytosterol treatment
resulted in an average of 10% reduction of total cholesterol and a 13% reduction
in LDL-c (102).
The safety and tolerability of etherified phytosterols, evaluated with the
administration to healthy adult men and women in a reduced-fat spread and salad
dressing, indicated that subjects that consumed reduced-fat spread and salad
dressing, providing 0 g/day, 3 g/day, 6 g/day, or 9 g/day of phytosterol esters for
8 weeks, displayed normal blood concentrations of all fat-soluble vitamins and
no differences in serum vitamin responses among the four groups. Also, total
cholesterol, LDL-c, and HDL-c responses did not significantly differ among the
groups, although the total cholesterol: HDL-c response in the 9 g/day group was
significantly different from control group response. In short, phytosterol esters
are well tolerated in humans and show no sign of adverse effects at daily intake
up to 9 g/day for 8 weeks (103). Diets enriched with different plant sterol mixtures
derived from wood source (e.g., Reducol from Forbes Medi-Tech, Inc. Vancouver,
Brithish Columbia) have also been used to lower serum cholesterol. ReducolTM
is a proprietary mixture of phytosterols composed of b-sitosterol, sitostanol,
campesterol, and campestanol derived from tall oil pitch and processed into a
crystalline powder. Feeding Reducol at a level of 1.8 g per day (e.g., 6% sterols
in margarine) lowered LDL cholesterol by approximately 15% compared with
placebo (104). The consumption of phytosterol mixtures increases cholesterol
synthesis and lowers cholesterol absorption.
The effects of consuming sitostanol ester margarine in subjects with moderate
hypercholesterolemia revealed that subjects showed unaffected Vitamin D and
retinol concentrations and a-tocopherol/cholesterol proportion, but lower serum
beta-carotene levels. Alpha-tocopherol and carotenes were related to serum
cholesterol and cholesterol absorption (105, 106). Moreover, the frequency of
consumption of plant stanol esters in margarines and shortenings does not signi-
ficantly affect the LDL-c lowering efficacy of the plant stanol esters (107). After
standardization for LDL-c, consistent with other findings, it was also found that
the sum of the most lipophylic hydrocarbon carotenoids, such as a-carotene,
b-carotene, and lycopene, were only slightly, though not significantly, lowered by
the consumption of the plant stanol esters in margarine and shortening.
The ensuing reduction of the most lipophylic antioxidants attributed to

phytosterol intake has also been related to a decrease in LDL (108). The increase
in dietary carotenoids, when humans consume plant sterols or stanols, is effective
in maintaining carotenoid concentrations (109). The addition of more daily
servings of high carotenoid-containing fruits or vegetables when consuming sterol
or stanol ester-containing spreads maintains plasma carotenoid concentrations
while lowering LDL-c (109).
Efforts have attempted to pinpoint the relationship between plant sterol
incorporation in human keratinocyte plasma membrane and modulation of
membrane fluidity (110). Data from in vitro trials on uptake of plant sterols and
membrane lipid fluidity in human keratinocytes suggest that, in the presence of
sitosterol, the mean fluidity of the membrane is regulated, whereas stigmasterol
triggers a looseness of molecular packing of phospholipids acyl chains. Phytoster-
ols also have anticancer dietary components (111); for example, it was found that
phytosterols indirectly (in vivo as a dietary supplement) and directly (in tissue cul-
ture media) inhibited the growth and metastasis of human prostate cancer PC-3
cells in mice (112). These findings are not limited to prostate cancer, as it was found
that phytosterols might also function in possible protective mechanisms in breast
and colon cancer. Phytosterols appear to have an effect on membrane structure
and on the function of tumor and host tissue, on the signal transduction pathways
that regulate tumor growth and apoptosis, on the immune function of the host, and
on the cholesterol metabolism of the host (111).
2.6. Fat-Soluble Vitamins (A, E, D, and K)

2.6.1. Vitamin A Several studies and reviews have examined the effects of
Vitamin A deficiency and toxicity (113). Endogenous retinoid toxicity has a
possible role in the pathophysiology of primary biliary cirrhosis (PBC), which is
a chronic, cholestatic disease, characterized by progressive destruction of the
small intrahepatic bile ducts and portal inflammation, leading to cirrhosis and
fibrosis. The major signs and symptoms of PBC resemble the manifestations of
hypervitaminosis A. Thus, the hypothesis has been made that exposure to excess
endogenous retinoids contributes to the pathogenesis of PBC and may be the cause
of some of the symptoms of PBC (114). Retinol has an effect on hepatic and renal
drug-metabolizing enzymes (115). Retinol pretreatment potentiates paracetamol-
induced hepatoxicity in BALB/c mice, an organ-specific response. And the
potentiation of paracetamol-induced hepatoxicity is independent of CYP450 and
glutathione (115).
Moderate to high doses of Vitamin A given to pregnant mice can result in signi-
ficant craniofacial, cardiac outflow, and thymic abnormalities (116). These results
signal the potential for the induction of birth defects in offspring of women
ingesting even moderate to low amounts of supplemental Vitamin A during an
early gestational period. Animal studies have been used in order to assess safe
dose levels of Vitamin A during human pregnancy (117). In one study, the safe level
of Vitamin A during human pregnancy was set at a range of 25,00037,000 IU/day
(118). In nonpregnant humans, the safe, nonteratogenic dose level of Vitamin A has
been determined to be 30,000 IU/day (119). Weaker teratogens than retinol are
retroretinoids, 14-hydroxy-4, 14-retroretinol (14-HRR), and anhydroretinol (AR)
(120). The low teratogenicity may be caused by facts that 14-HRR and AR do
not contain the terminal carboxylic group involved in binding and activation of
the retinoic acid nuclear receptors, and they are not metabolized to acidic retinoids.
Vitamin A, retinol, and derivative retinal, also causes oxidative DNA damage
via superoxide generation (121). Beta-carotene, for example, is related to a higher
incidence of lung cancer. Retinol and retinal can cause cellular DNA cleavage. This
is probably caused by the dismutation of superoxide to H2O2, generated by the auto-
oxidation of retinoids in the presence of endogenous metals. Thus, retinol and ret-
inal have capacities to exhibit pro-oxidant activity, which may lead to carcinogen-
esis of beta-carotene supplements (121). Retinol supplementation also induces
DNA damage and modulates iron turnover (122). In animal studies, retinol caused
cellular DNA damage involving cellular iron accumulation. These characteristics
could be responsible for the increased incidence of lung cancer associated with reti-
noid supplementation (117, 122). Lycopene, lutein, and zeaxanthin are considered
potential antioxidants for the oxidation of carotenoids by free radicals (123).
Recently, concerns have been raised as to the retinol equivalents calculated by
pro-vitamin A carotenoid conversion factors (124). It has been suggested that these
values should be treated with caution until more data on absorption of carotenoids
from foods is known.
2.6.2. Vitamin E (Tocopherols) In an oral toxicity study using a tocotrienal

preparation in rats, the NOAEL for tocotrienal was 0.19% (125). In the presence
of copper (II) ions, a-tocopherol induces oxidative damage to DNA as copper-
dependent reactive oxygen species formation occurs from molecular oxygen,
thus resulting in DNA base oxidation and backbone cleavage (126). This pro-
oxidant activity will induce tumor formation and act as a complete tumor
promoter in animals. However, a product of a-tocopherol oxidation/reduction,
a-tocopherolhydroquinone, (TQH2) has antioxidant properties, as it may scavenge
peroxyl radicals primarily by electron transfer to form TQ and secondarily by
addition-elimination to form the epoxyquinones (127).
Vitamin E has been found to be useful in the storage of irradiated fresh meat.
Feeding calves a diet rich in PUFA and supplemented with Vitamin E helped to
control lipid peroxidation and oxymyoglobin oxidation of ground beef cold pasteur-
ized with electron-beam irradiation during storage (128). Thus, color retention and
lipid peroxidation are controlled by the presence of a critically high concentration
of a-tocopherol (TOH) (129). As tocopherol has a peroxy-radical scavenging
function (130), the concentration of TOH decreases and the concentrations of
alpha-tocopherolquinone and 2,3-epoxy-alpha-tocopherolquinone increase when
TOH oxidation occurs in meat, such as in beef and bovine muscle microsomes.
Vitamin E has also been noted for the function of tocopherylquinones as highly
cytotoxic agents. Tocopherylquinones escape multidrug resistance in acute lympho-
blastic leukemia cell lines (131). Ascorbate and a-tocopherol also have several poten-
tial antiatherogenic mechanisms, such as the inhibition of LDL oxidation, the inhi-
bition of leukocyte adhesion to the endothelium, and the inhibition of vascular endo-
thelial dysfunction (132). Alpha-tocopherol acts as either antioxidant or pro-oxidant
of lipid peroxidation in LDL, but may only protect against atherosclerosis in com-
bination with Vitamin C. Tocopherols and tocotrienols also have antiproliferative and
apoptotic effects on normal mouse mammary epithelial cells (133), as mammary
epithelial cells uptake of tocotrienols is greater than tocopherols. This suggests that
tocotrienols have greater biopotency than tocopherols partly because of greater cel-
lular accumulation. In all, g- and d-tocotrienols may have important roles in mod-
ulating normal mammary gland growth, function, and remodeling. The mechanism
underlying the antiatherogenic properties of Vitamin E is its function to decrease
the uptake of modified LDL and suppresses aceyl-CoA: cholesterol acyltransferase
(ACAT) activity, resulting in less cholesterol esterification in macrophages (134).
2.6.3. Vitamin D Vitamin D has potential roles in the prevention of some can-
cers, osteoarthritis progression, multiple sclerosis, and hypertension (135). Feeding
low levels of 1,25-hydroxy-vitamin D3 (1,25(OH)2D3) supplementation produces
no toxicity in laying hens; however, feeding very high levels produces clear toxic
symptoms (136). The feeding of excessive amounts of vitamin D3 to rats will lead
to bone breakdown and increased levels of zinc in the blood (137). In pot-bellied
pigs, Vitamin D toxicity was expressed as anorexia, weight loss, lethargy, polyuria,
polydipsia, vomiting, tenesmus, and tremors (138). Similar effects have been
observed in weanling pigs as Vitamin D toxicity caused serum calcium and blood
urea nitrogen concentrations to increase and a decrease in serum phosporus (139).
Excess Vitamin D3 is toxic, particularly to vascular tissues; a notable pathological
feature being arterial calcification (140). Excess Vitamin D is arteriotoxic and it can
induce arterial calcification through upregulation of 1,25(OH)2D3 receptors and
increased calcium uptake in arterial smooth muscle cells (141). Vitamin D toxicity
may also be expressed with the wasting and calcification of soft tissues in cattle
after ingestion of a plant, Solanum glaucophyllum (Sg), which contains high levels
of 1,25-dihydoxy-vitamin D3. The Sg-intoxicated cattle showed atrophy of the
epidermis, severe involution of hair follicles, sebaceous and sweat glands, and
reduced cellular proliferation. In humans, outbreaks of hypervitaminosis D have
been linked to the overfortification of milk from home-delivery dairies (142).
Vitamin D has protective effects on diet-induced epithelial cell hyperpro-
liferation (143). Increasing dietary Vitamin D, along with calcium, will prevent
hyperproliferation. Amino bisphosphonate ibandronate has been found to prevent
Vitamin D toxicity (144) by inhibiting Vitamin D-induced calcification of arteries,
cartilage, lungs, and kidney in rats. Thus, in the future, ibandronate may be used
to treat patients exposed to toxic levels of Vitamin D. Another protective effect
of Vitamin D is the function of 1-alpha-dihydroxy-vitamin D3 (1,25(OH)2D3), in
particular, in the reduction of the proliferation of human cancer cells (145). It has
been shown to increase differentiation in human colon cancer cells, preventing
colonic hyperproliferation and oxidative stress. 1,25-D3 also has other immunosup-
pressive effects. As a therapy, it prolongs the survival of renal allografts and
preserves graft function in rats (146). These effects are even more apparent when
the Vitamin D therapy is combined with cyclosporine A. Vitamin D has several
possible noncalcaemic roles, including its role in the immune system and, in
particular, on T cell-mediated immunity (142). The role of Vitamin D compounds
as selective immunosuppressants is illustrated by an affinity to either prevent or
suppress autoimmune disease.
2.6.4. Vitamin K Recently, dihydro-vitamin K1 has been identified as a dietary

form of Vitamin K produced during the hydrogenation of Vitamin K1-rich vege-
table oils (147). Children have the highest intake of dihydro-vitamin K1 (30%
of total Vitamin K intake), followed by a progressive decrease in percentage
contribution with age. The hydrogenation of vegetable oils decreases the absorption
and biological activity of Vitamin K bone (148). The hydrogenated form of
Vitamin K, dihydrophylloquinone, is less bioavailable than phylloquinone and
has no measurable biological effect on measures of bone formation and resorption.
A high-performance liquid chromatographic method for the determination of
phylloquinone and menaquinones in foods of animal origin has recently been
developed (149). Dietary Vitamin K intakes are associated with hip fractures, but
not with bone mineral density in elderly men and women (150). Vitamin K may
also play a role in the etiology of colon cancer (151). It has been proposed
that bile acids (e.g., deoxycholic acid), K vitamins, iron (II) complexes, and
oxygen interact to induce an oncogenic effect in the colon by the generation of
free radicals. This may be caused by the function of the reduced K vitamins in
initiating generation of superoxide radical (O2-), leading to an Fe(II)-mediated
Fenton reaction in colon stem cells.
3. ADVERSE EFFECTS OF SOME NATURAL CONSTITUENTS

IN FATS AND OILS
3.1. Erucic Acid

The chain elongation of 22:1n-9 (erucic acid) causes an increase in 24:1n-9 platelet
sphingomyelin content (153). This effect of erucic acid has been proven in the
transient decrease in platelet counts and increase in platelet size in newborn piglets
fed canola oil. Similar findings regarding the effect of erucic acid on platelet counts
has been reported in human studies on patients with adrenoleukodystrophy (154).
Erucic acid treatment causes decreased platelet counts and morphologic and
platelet sizing measurements, suggesting that erucic acid also affects the physical
properties of platelets. Speculation has also been prevalent that the erucic acid
content in oils, like canola oil, may cause the accumulation of triglyceride in the
heart when fed to infants. Recent animal studies have shown, however, that the
modest accumulation of erucic acid associated with feeding canola oil is not
associated with biochemical evidence of heart triglyceride accumulation (155).
Other animal studies have reported that an erucic acid ethyl ester diet induces a
ADVERSE EFFECTS OF SOME NATURAL CONSTITUENTS 531
marked increase in free fatty acids and in triglyceride content, as well as marked dif-
ferences in the fatty acid pattern in triglycerides, free fatty acids, and diglycerides,
but only marginal differences in phospholipids (156). The question still remains as
to whether dietary erucic acid can be hepatotoxic in pregnancy. Although the
erucic acid content in rapeseed oil, for example, is associated with more weight
gain and higher proportions of erucic acid in the heart when compared with corn
oil, rapeseed oil dietary content also causes lower bile flow in pregnant hamsters
(157).
3.2. Unconventional Oils

3.2.1. Rice Bran Oil Rice bran oil (RBO) and its main components have affinity
toward improving the plasma lipid pattern of rodents, rabbits, nonhuman primates,
and humans by decreasing total plasma cholesterol and TG levels and increasing
HDL-c levels. Other potential properties of rice bran oil and gamma-oryzanol
include modulation of pituitary secretion, inhibition of gastric acid secretion,
antioxidant action, and inhibition of platelet aggregation (158). Oryzanol is
contained in the nonsaponifiable lipid fraction of rice bran oil. Three major
components of gamma-oryzanol are cycloartenyl ferulate, 24-methylene-cycloartanyl,
and campesteryl ferulate (159). In an aqueous model system, the nonsaponifiable
fraction in rice bran has been shown to inhibit cholesterol autoxidation (160). In
hamsters, oryzanol decreases cholesterol absorption and aortic fatty streaks
(161). Oryzanol treatment results in significant decreases in plasma total cholesterol
and the sum of IDL-c, LDL-c, and VLDL-c. Thus, oryzanol is at least partly
responsible for the cholesterol-lowering action of RBO. The blending of RBO
with other oils, such as safflower oil or sunflower oil, improves lipid profiles by
reducing TC, TG, and LDL-c and increasing HDL-c (162). However, where the
blending of RBO with safflower oil may magnify a hypocholesterolemic activity
(163), the blending of RBO with sunflower oil does not have the same effect
(164). This difference is likely caused by the different triacylglycerol structures
of safflower and sunflower oil.
3.2.2. Rubber Seed Oil Rubber seed oil (RSO), which has a high C18:3, n-3
content (6), has a lower alcoholysis rate than linseed oil, but a higher alcoholysis
rate than soybean oil and melon seed oil (165). Studies on the epoxidation of RSO by
peroxyacetic acid generated in situ have shown that increase in the process tem-
perature increases the rate of epoxide formation (166). The optimum alcoholysis
temperature for RSO is 245 2 C.
3.2.3. Ricinoleic Acid Ricinoleic acid (RA) can increase mucosal permeability
and cause cytotoxicity. It is also associated with the release of eicosanoids, a
platelet-activating factor, and nitric oxide (NO). RA disrupts normal intestinal
motility and a combination of these effects accounts for the laxative action of
RA (167). Consistent with these findings, RA will increase nitric oxide synthatase
activity in the rat ileum and colon (168), which likely accounts for the involvement
of NO in the laxative action of RA. Although RA possesses capsaicin-like dual

pro-inflammatory and anti-inflammatory properties, unlike capsaicin, RA does
not induce an inward current in dorsal root ganglia neurons and it does not have
algesic properties in vivo (169). Thus, RA may be viewed as a new capsaicin-
like, nonpungent, anti-inflammatory agent suitable for peripheral application.
3.2.4. Cyclopropenoid Fatty Acids (CPFA) Animal studies have shown that
higher percentages of CPFA in oils are related to retarded growth (170). For example,
heating baobab oil, thereby reducing its CPFA content, will caused the oil to
increase the cytosolic glutathione transferase activity in rats fed this oil source.
These mechanisms of CPFA action might be related to alterations of membrane
lipid composition or microsomal proteins. CPFA are classified as toxic nonoils
that are found in cottonseed oil (171); albeit cottonseed oil has been found to be
an ingredient that is used safely in cosmetic formulations if established limits on
gossypol, heavy metals, and pesticide concentrations are not exceeded.
3.2.5. Structured Triacylglycerols Exposure to dietary structured triacylglyce-

rols containing docosahexaenoic acid from birth will positively affect the visual
and auditory performance and tissue fatty acid profiles of rats (172). Although
the feeding of this specifically structured oil will change the fatty acid profiles of
rats and cause a higher level of 22:6n-3 in rat brain tissue, these changes will not
result in differences in learning ability. However, positive changes in visual function
and in auditory brainstem response have been observed. In the brain, 22:6n-3 levels
increased in brain phosphatidylcholines and phophatidylserines (173); 22:6n-3 also
increases in the adipose tissue, suggesting that the surplus of dietary 22:6n-3 is
stored. Structured triacylglycerols differ from conventional triacylglycerols, because
the medium-chain fatty acids esterified to the glycerol skeleton are absorbed in
the intestines via portal blood as free fatty acids (174); absorption pathways are
also conventional with long-chain triacylglycerols.
Recent animal studies that looked at the effects of the structured lipid,
SALATRIM, in enriched biscuits on serum and liver lipid concentration in rats,
reported that SALATRIM could be used as a fat substitute in biscuits without
influencing lipid metabolism (175). Studies have also shown that SALATRIM
can help prevent accumulation of TG and TC in the liver and in white adipose
tissue that is induced by a high lard diet fed to rats (176). Synthesis of structured
tricylglycerols containing caproic acid by lipase-catalyzed acidolysis have
been performed in a batch reactor, with a solvent-free system of structured
triacylglycerols containing short-chain fatty acids by Lipozyme RM IM-catalyzed
acidolysis between rapeseed oil and caproic acid and optimized using response
surface methodology (177).
The effects of Caprenin, another structured lipid, on chylomicron fatty acid com-
position and postprandial serum lipid concentrations have also been studied (178).
It was found that there is a very low uptake of C8:0, C10:0, and C22:0 into chylo-
microns. Moreover, a postprandial lipemia after caprenin is comparable with that
produced by other dietary fats as opposed to a fat-free meal. There is considerable
BIOACTIVE LIPID-SOLUBLE CONSTITUENTS 533
contribution of endogenously derived fatty acid to chylomicron lipids, and there

are equal effects of saturated fatty acids on pre- and postprandial concentrations
of plasma cholesterol.
Olestra is a different example, which is comprised of a fat-derived product made
from sucrose that is esterified with fatty acids to generate a virtually unavailable
form of calories caused by its very poor digestibility in the gastrointestinal tract.
Nevertheless, the similar physical properties of regular oils and fats enabled it to
be used as a low-calorie fat source that can be used in high-temperature cooking.
A regular intake of Olestra will reduce the uptake of carotenoids (e.g., b-carotene)
and potentially other fat-soluble vitamins. Radio-labelled Olestra experiments have
confirmed the fact that Olestra is essentially not absorbed and heating the sucrose
polyester will not improve bioavailability (179). Other acute and subacute toxicity
studies have indicated that no related effects attributed to Olestra intake in regard
to common toxicological endpoints, such as time-to-tumor or tumor incidence,
clinical chemistry, haematology, weight changes in specific organs, and tissue
morphology, occur when fed up to 10% of the diet (180). Although Olestera has
been shown to have no effect on water-soluble vitamins, formulations required
the addition of Vitamins A, D, E, and K to avoid depletion in individuals that
consume foods containing Olestra.
4. BIOACTIVE LIPID-SOLUBLE CONSTITUENTS
4.1. Phytoestrogens
The Committee on Toxicity of Chemicals in Foods, Consumer Products and the
Environment recommended in 2002 more scientific studies in understanding the
relative risk:benefit of phytoestrogens to human health. Studies have reported
that isoflavone-containing diets, when fed to C57BL/6 mice, resulted in reduced
cholesterol levels, but had no effect on cholesterol levels or on the susceptibility
of LDL to oxidative modification in LDLr-null mice (LDL receptor-deficient)
(181). Plant estrogen isoflavones also have potentially antiatherogenic effects, in
addition to antioxidative and antiproliferative properties (182). Both soy-derived
isoflavones and esterified isoflavones reduce in vitro oxidation susceptibility of
LDL. The lipophilic phytoestrogen derivatives can be incorporated into LDL,
thereby increasing the oxidation resistance and antiproliferative efficacy ex vivo.
It is unclear, however, whether the cholesterol-lowering effect of a soy-rich diet
may be associated with the presence of isoflavones (183).
Coumestrol and genistein are principle phytoestrogens, which induce micronuclei
containing acentric fragments and DNA strand breaks. Coumesterol also induces
hypoxanthine guanine phosphoribosyltransferase mutations in cells where
genestein is marginally active at this endpoint (184). The phytoestrogen, daidzein,
has no effect, nor does it cause significant toxicity on the reproductive tract of
animals or provide a protective effect against chemically induced mammary cancer
(185). Anitoxidant activity has been shown to include the interference of advanced
glycation end-products-mediated oxidative DNA damage of vascular smooth

muscle cells (186). They are also potentially useful against vascular diseases, where
reactive oxygen species are involved in hypertension.
Coumesterol exhibits both mutagenic and clastogenic properties in AHH-1
TK/- human lymphoblastoid cells (187). Earlier studies found genestein to
have both anticarcinogenic and antiproliferative activities; however, other studies
indicate that this phytoestrogen may actually enhance the development of colon
cancer (188) and is carcinogenic in neonatal mice (189). Thus, the biological
effects of genestein may be organ specific. In the male mouse reproductive tract,
genestein exerts estrogen-like effects, comparable with those present in soy-based
diets. In neonatal animals, however, higher doses are needed to show estrogen-like
effects (190). Neonatal exposure to genestein in mice during critical periods of
differentiation may actually increase the incidence of uterine adenocarcinoma
and mammary tumorigenesis (189, 191).
4.2. Monoterpenes
The cumin herb has been investigated as a new source of essential oil. It contains
considerable amounts of oxygenated monoterpenes and small amounts of monoter-
penoid and sesquiterpene hydrocarbons (192). In terms of the antioxidant activity of
monoterpenes, one recent study tested 100 pure components of essential oils for
antioxidant effectiveness and found that phenol consitutents possessed the highest
antioxidant activity (193). Additionally, the monoterpene hydrocarbons, terpinolene
and a- and g-terpinene, showed significant protective action. A number of dietary
monoterpenes have chemopreventive activity against rat mammary cancer and the
monoterpenes act through multiple mechanisms in chemoprevention of mammary
and other cancers (194). The diterpenes found in rosemary leaves have also been
found to have antioxidant properties (195). Carnosic acid is a major phenolic
diterpene present in rosemary leaves, with lesser amounts of 12-methoxycarnosic
acid and carnosol . The antioxidant potency of carnosic acid was more than twice
that of any other compound tested. A water-soluble extract of rosemary and its
purified component, rosemarinic acid, had an effect on the xenobiotic-metabolizing
enzymes in rat liver (196). The induction of xenobiotic-metabolizing enzymes
by the water-soluble extract could be attributed to flavones, monoterpenes, or an
additive effect of all components, as evidenced by the fact that a water- soluble
rosemary extract selectively induced cytochrome P450 and enhanced detoxification
enzymes (197). Examinations of the monoterpene d-limonene have shown that
d-limonene produces tumors only in kidneys of male rats in association with
hyaline-droplet nephropathy, which is because of the accumulation of the
rat-specific, low-molecular-weight protein a2u-globulin in P2 segment cells of
renal proximal tubules (198). There is, however, no risk of cancer for humans
from d-limonene, because the binding of d-limonene to a2u-globulin would not
occur in human cells.
The reproductive toxicity of a-terpinene has also been reported (199). In rats,
embryo and fetal toxicity occurs with doses greater than 30 mg/kg/d. Maternal
CHEMICAL REACTIONS IN FATS 535
toxicity occurs at 125 mg/kg/d. b-Myrcene also has an effect on rat fertility and
general reproductive performance (200). In male and female rats, exposure to it
will increase liver and kidney weights. In postnatal rats, exposure causes days of
appearance of primary coat, incisor eruption, and eye opening to be slightly
delayed. The NOAEL for b-Myrcene has been set at 300 mg/kg body weight.
The monoterpene, limonene, also causes additive toxicity in human lung cells.
The detoxification of limonene in human lung cells occurs mainly by mechanisms
not involving the glutathione system because 1,2-epoxide is not the active
compound in limonene toxicity (201).
5. CHEMICAL REACTIONS IN FATS
5.1. Hydrogenation and Isomerization Reactions in Fats and Oils

Moderate amounts of trans-fatty acids from partially hydrogenated soybean oil
(PHSBO) have an inhibitory effect on PUFA formation (202). In rodents, tissue
contents of MUFA increase and PUFA decreases after consuming hydrogenated
dietary fat (203). In concordance, total cholesterol (TC) and LDL-c increased after
partially hydrogenated fat feedings. Partially hydrogenated fish oil (PHFO) contains
a high amount of trans-fatty acids. Totally hydrogenated fish oil (THFO) contains
minimal trans-fatty acids, but a high content of very long-chain saturated fatty acids
(VLCSFA). When the absorption and metabolism of VLCSFA from THFO was stu-
died in rats, a low absorption of VLCSFA was observed (204). The PHFO and
THFO groups had reduced serum TC, and the PHFO group had increased serum
TAG.
Hydrogenated fat consumption affects cholesterol synthesis in moderately
hypercholesterolemic women (205). Plasma TC and LDL-c levels increase
with increasing degrees of hydrogenation or saturated fat intake. The mecha-
nism by which hydrogenated fat influences plasma lipid levels involves
impairment of the catabolic pathway of cholesterol. Hydrogenated fat con-
sumption also affects acylation-stimulating protein levels and cholesterol
esterification rates in moderately hypercholesterolemic women (206). The
alterations in the circulating lipid levels observed with the consumption of
hydrogenated fat-rich diets can be explained partly by changes in acylation-
stimulating protein activity and newly synthesized cholesterol. The degree
of hydrogenation of dietary fish oil has been shown to have an effect on
the trans-fatty acid content and enzymatic activity of rat hepatic microsomes
(207). Partially hydrogenated vegetable oils have an early mortality effect in
stroke-prone spontaneously hypertensive rats (SHRSP) (208). Incorporation
and metabolism of trans-20:5 in endothelial cells leads to an effect on
prostacyclin synthesis (209). In these cases, the survival time-shortening
activity of partially hydrogenated soybean oil is exerted by the cis- and
trans-isomers of oleic acid, or by an unidentified factor generated during
partial hydrogenation.
5.1.1. Conjugated Linoleic Acid (CLA) CLA, present naturally in ruminant

lipid sources or derived from thermal processing of oils, may have the capability
to prevent cancer and heart disease, improve immune function, and alter body
composition for treating obesity or building lean body mass (210). The biologically
active isomers of CLA have antiproliferative effects on human colorectal and
prostatic cancer cells, with the trans-10, cis-12-CLA isomer exhibiting the greatest
potency against colorectal cancer proliferation. The cis-9,trans-11 and trans-10,cis-
12 isomers were moderately effective against prostate cancer (211). An oxidative
mechanism appears to be involved in the growth-suppressive effects of cis-9, trans-
11 CLA (212). The mechanism of CLA in decreasing colon tumor incidence in
rats may possibly involve increased apoptosis (213). Aside from its anticarcino-
genic properties, CLA has also been found to have cytotoxic effects on the
antioxidant enzyme defense system in rat hepatocytes (214). The cytotoxic effects
of CLA, as described by lactate dehydrogenase leakage and decreased gluconeo-
genesis, were not mediated by a pro-oxidant action in the hepatocyte, and CLA
affected membrane integrity, metabolic function, cellular lipid composition,
lipid peroxidation, and activity of antioxidant enzymes. Although CLA has been
reported as peroxisome proliferators in mice, it has not been found to act as
peroxisome proliferators in rats (215). In general, CLA has been found to have a
lack of toxicity in rats, thus supporting the potential determination for the
GRAS status of CLA (216). In humans, dietary exposure to CLA may reduce
the proportion of body fat and affect fatty acid metabolism (217). The short-term
consumption of CLA does not exhibit antithrombic properties in humans (218),
nor does it offer health benefits regarding the prevention of atherosclerosis (219),
altering blood cholesterol, nor lipoprotein levels. Although CLA does not appear to
act as an antioxidant, its affinity to decrease polyenoic fatty acid concentrations
could decrease the formation of highly cytotoxic lipid oxidation products such as
MDA (220).
5.1.2. Trans-Isomers and Cancer A study conducted in postmenopausal

women suggested an association between risk of breast cancer and the level of
hydrogenated oil derived mono-trans-fatty acids was stored in the adipose tissue
(221). It was also found that trans-fatty acid might cause colorectal neoplasia by
interfering with the cell membrane function or eicosanoid metabolism (222).
Increased adenoma prevalence was associated with the consumption of sweetened
baked goods, oils, and condiments.
5.1.3. Trans-Isomers and Coronary Heart Disease An increased risk of deve-

loping heart disease has been linked to an intake of trans-fatty acids (223). The
replacement of dietary saturated fatty acids by trans-fatty acids, for example,
lowers serum HDL cholesterol and impairs endothelial function in healthy men
and women (224). It also impairs flow-mediated vasodilation and decreases the
activity of serum paraoxonase, which is an HDL-bound esterase that may protect
against atherosclerosis (225).
5.1.4. Effects of Trans-Isomers on Neonatal Growth Plasma trans-fatty

acids have been inversely related to birth weight and head circumference,
but these results are unclear (226). What is clearer is the hypothesis that dietary
trans-fatty acids inhibit the biosynthesis of long-chain PUFA with 20 and
22 carbon atoms, and thus affect infant development. These findings are supported
by studies showing an inverse correlation of plasma trans-fatty acids with n-3
and n-6 long-chain PUFA in infants. Trans-fatty acids may inhibit the desatu-
ration of linoleic acid to arachidonic acid and of alpha-linolenic acid to DHA
(227). Heightened concern exists over trans-fatty acids and infant development,
as a result of research showing the importance of n-3 and n-6 fatty acids (228).
These studies have shown that trans-fatty acids inhibit the delta-6 desaturation
of linoleic acid. In fetal, infant, and maternal tissues, there exists an inverse
relationship between trans-fatty acids and measures of growth and development.
Animal studies on the effect of trans-fatty acids and behavioral development of
pre- and postnatal mice have reported that, although reversal learning in a T-water
maze was slower in trans-fatty acid- supplemented groups, the long-term effects
of trans-fatty acids on behavioral development and neural function needs more
investigation (229).
5.2. Heterocyclic Amines (HAs)

Problems associated with the determination of HAs in cooked foods and with
human exposure have been reported. The estimated daily intake of HAs, in different
studies, ranges from 015 mg/person/day (230). In a study that looked at the hetero-
logous expression of human N-acetyltransferases 1 and 2 and sulfotransferase 1A1
in Salmonella typhimurium for the mutagenicity testing of heterocyclic amines, it
was found that both O-acetylation and O-sulfonation were important determinants
for HA genotoxicity in humans (231).
The heterocyclic amine, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
(PhIP), is a mutagenic and carcinogenic heterocyclic amine (232). PhIP is
metabolically activated by cytochrome P450-mediated N-hydroxylation, followed
by phase II esterification. The mutagenic metabolite reacts with DNA-forming
adducts, proteins, and other cellular constituents, resulting in unstable products
that are degraded to 2-amino-1-methyl-6-(5-hydroxy) phenylimidazo[4,5-b]-
pyridine (5-OH-PhIP). Urinary 5-OH-PhIP can, therefore, be used as a bio-
marker for the genotoxic dose of PhIP. Studies show that 2-amino-3-8,-
dimethylimidazo[4,5-f]quinoxaline (MeIQx) is readily available to tissues for
humans and rodents, and adduct levels are linear with an administered dose,
except at high chronic doses where adduct levels plateau slightly (233).
HAs content in preprocessed meat cuts produced in Canada were present in 16
different types of processed meat cuts (234). The highest mutagenic activity was
found in a smoked turkey breast sample; four samples had low mutagenic activity,
and 11 samples were not mutagenic. The only HA found in the samples with muta-
genic activity was MeIQx. Conclusions have been made that the consumption of
meat cuts does not present a serious health risk from HAs contaminants. In the
American diet, the estimated daily total HAs intakes for children is 11 ng/kg/d
and for adults it is 7 ng/kg/d, with PhIP estimated to account for 65% of each intake
(235). Pan-fried meats are the largest source of HAs in the average American diet,
with chicken being the largest source of HAs among different meat types. In pork,
the HAs type and level will vary with the pork product, cooking method, and done-
ness level (236), but the main HAs found in pork are IQ, MeIQ, MeIQx, DiMeIQx,
and PhIP. In chicken, HAs content increases with increasing cooking temperature
(237). PhIP formation starts accelerating at temperatures >200 C. Recent studies
have not found any correlation between color developments and HAs content in
chicken (237). In pan-fried meat patties, HAs have been shown to induce bacterial
mutagenicity and animal carcinogenicity and may be a risk factor for human cancer
(238). Fast-food meat products appear to contribute to only a small percent of
the estimated daily dietary intake of HAs (239).
The food-derived mutagen 2-amino-9 H-pyrido-[2,3-b]indole (A(a)C) has been
reported to exhibit weak mammary gland carcinogenicity in mice (240). This observ-
ation may partly be associated with AaC-DNA adduct formation in the mammary gland
epithelium. However, although the acute feeding of MelQx in mice has produced
hepatic tumors, chronic feeding did not cause outward signs of toxicity, although
there is a slight increase in sister chromatid exchanges at 400-ppm MeIQx (241).
The carcinogenic effects of HAs can be extended to investigating various
cancers in humans. When the duration of meat cooking is taken to be a marker
of HAs content, the significantly raised and exposure-related increase in stomach
cancer risk occurring in humans was associated with preferences of well-cooked
meat as opposed to rare meat. This conclusion was based on the fact that HA
content is greater in well-cooked meat (242). For example, the type of meat,
method, and extent of cooking has been attributed to a two-fold increase in risk
of colon or rectum tumors among those believed to have highest intake of HAs
(243). Notwithstanding this, the differences in human cancer risk for HA ranges
more than a thousand-fold between individuals based on exposure and genetic
susceptibility (244).
One animal study reported that either the tumor-promoting effects of MeIQx or
PhIP were weak, or else the dose of BaP (initiation treatment) in the study was too
high and masked the effects of MeIQx or PhIP (245). Several studies have con-
firmed that PhIP does affect mammary carcinogenesis in various strains of mice
and rats (246). Upon the investigation of the formation of mutagenic/carcinogenic
heterocyclic amines in dry-heated model systems, meats, and meat drippings, there
were nine HAs found at concentrations greater than 0.1 ng/g in model systems,
meat, or pan residues (247). Finally, various additives have been shown to have ef-
fects on the formation of heterocyclic amines in fried fish fiber (248). For example,
HAs formation is retarded by the addition of a high level of sugar, and it is
increased with increasing levels of MSG. Antioxidants do not show any consistent
effect on HAs formation. Coconut oil, lard, and soybean oil all contribute to high
levels of HAs. During heating, HA loss increases with both increasing temperature
and heating time (249).
5.3. Lipid Oxidation Products (LOPs)

Lipid oxidation in both food systems and biological tissues exhibit the same
temporal three-stage pattern of initiation, propagation, and termination
(Figure 3).
In Vitamin E-controlled autoxidation of methyl linoleate, the 11-hydroperoxy deri-
vative (11-hydroperoxylinoleate) was identified as the next most prominent primary
initial peroxidation product after the 9- and 13-hydroperoxides (250). Feeding a
high fish oil diet and associated amounts of n-3 polyunsaturated fatty acid does
not propagate increased levels of phospholipid hydroperoxides or TBARS in rat
organs (251). The formation of 8-oxo-20 -deoxyguanosine in DNA of intact human
diploid fibroblast cells by lipid hydroperoxides are likely due to the generation
of reactive species other than superoxide radicals and hydrogen peroxide (252).
Feeding oxidized oil to rats causes several changes in lipid and fatty acid metabo-
lism, including: a reduced rate of desaturation of linoleic acid and alpha-linolenic
acid by microsomal delta4-, delta5-, and delta6-desaturase; a reduced ratio MUFA/
SFA in the liver, suggesting reduced delta9-desaturation; small increases in, liver
Initiation Propagation Termination
R + LH RH + L R + R RR
L + O2 LOO
RHR
R = free radical Lag phase duration

L = lipid alkyl radical
LOO = lipid alkyl peroxyl radical
Figure 3. A schematic diagram of temporal pattern of lipid oxidation in food and biological
systems.
weight; and reduced tocopherol concentrations in both liver and plasma. In addi-
tion, reduced lipid concentrations in plasma and an increased ratio between
phospholipids and cholesterol are biomarkers for the oxidative stress that occurs
in the liver (254).
The interaction of lipid peroxides with cellular proteins may contribute to cellu-
lar aging. Tubulin, the building block of microtubules, is a potential target for
cellular aging, and very low concentrations of phosphatidylcholine hydroperoxides
are sufficient to interfere with tubulin and microtubule function (255). When rats are
fed a thermally oxidized corn oil diet, higher concentrations of lipid peroxides
appear in the liver and kidney (256), which corresponds to an increased liver
weight, but decreased body weight gain. The iron redox cycle also catalyzes mem-
brane lipid peroxidation and oxymyoglobin oxidation, as evidenced by the oxida-
tion of ferrous ions and ascorbic acid with auto-oxidation of myoglobin, and the
generation of lipid peroxides by lipid radicals, which, in turn, contributes to the loss
of oxymyoglobin stability (257). Aldehyde lipid oxidation products (LOPs) also
have an effect on myoglobin (258). Aldehyde LOPs alter myoglobin stability by
increasing oxymyoglobin oxidation, decreasing the ability of metmyoglobin to be
enzymatically reduced, and enhancing the pro-oxidant activity of metmyoglobin.
The acceleration of oxymyoglobin oxidation is triggered by alpha,beta-unsaturated
aldehydes via covalent attachment (259). Polyamines and spermidine, along with
sulfhydryl-containing compounds glutathione and thioctic acid, decrease headspace
hexanal, a saturated aldehydic lipid oxidation product (260). The generation of lipid
peroxyl radicals from edible oils and associated biological activities results in a need
to quench free radical components. For example, LOOH generated alkylperoxyl
radical (LOO*), following reactions with various heme compounds, such as myoglo-
bin, cytochrome c, or hemin, exhibits cytotoxicity and causes DNA damage (261).
Lipid hydroperoxides also modify proteins during myocardial ischaemia (262) and
can contribute to the pathophysiology of ischaemic injury. Oxidation reactions,
however, may not only represent toxicologic events, but rather modulate cell activ-
ity and function (263). For example, 4-hydroxynonenal (a LOP present in ox-LDLs)
has a role in cell signaling by upregulating AP-1 transcription factors with an induc-
tion of a series of genes. Protein modifications are also initiators of oxidant-induced
signal transduction pathways. Moreover, evidence exists that implicates a dietary
source of plasma lipid peroxides, which becomes elevated in the postprandial state
(264); a potential contributing factor for the correlation between postprandial
hyperlipidemia and increased risk of CVD. There is also evidence in animal studies
of proatherogenic properties of oxLDL and presence in atherosclerotic lesions
(265). Reactive aldehydes generated from lipid peroxidation are involved in
CVD (266). Another example lies with the role of oxidative stress in the patho-
physiology of asthma (267). Lipid peroxidation, as determined by plasma iso-
prostanes, is related to disease severity in mild asthma. Tumor cell lines are
sensitive to PUFA and to associated oxidation products (268). This sensitivity
depends on the antioxidant defense mechanism, as well as on culture conditions.
Hydroperoxy docosahexaenoic acid is a major metabolite, responsible for the cyto-
toxicity of DHA.
Uncovering the molecular structures of LOPs using high-resolution, two-

dimensional 1H and 13C nuclear magnetic resonance techniques have further
uncovered the molecular structures of LOPs (269). Advances in chromatography
using a combination of normal-phase high-performance chromatography with
mass spectrometric detection of nonvolatile LOPs has been used to separate TG
oxidation products (270, 271).
5.4. Cholesterol Oxidation Products (COPs)

COPS have been identified and quantitated in a number of food sources that include
egg, beef, pork, and butter (6). COPs exhibit cytotoxicity to a wide variety of cells
leading to angiotoxic and atherogenic effects (272). They alter vascular permeabil-
ity to albumin, modify prostaglandin synthesis, and stimulate platelet aggregation.
COPs also alter the functionality of LDL receptors, modify cholesterol ester accu-
mulation in various cells, and enrich the LDL particle in cholesterol esters. Oxyster-
ols are also mutagenic and carcinogenic. When the cytotoxicities of cholesterol and
a mixture of b-sitosterol/campesterol and related oxides were compared in the
C57BL/6 macrophage cell line, 5 a,6a-epoxide or cholesterol oxides caused the
greatest cell damage, followed by b-sitosterol/campesterol oxides, cholesterol,
and b-sitosterol (273). When dietary COPs are absorbed and incorporated into
rat lymph chylomicrons, postprandial lipoprotein particle size and composition
are influenced (274). These changes may affect the clearance of chylomicrons
from plasma, the arterial delivery of COPs, and the possible deposition in arterial
lesions.
The oxysterol 7-ketocholesterol is an important COP involved in athero-
sclerotic lesions and macrophage foam cells (275). There is no direct evidence
in humans that COPs contribute to atherogenesis, but it has been found that
COP levels are elevated in LDL subfractions that are considered potentially
atherogenic (276). In addition, raised levels of 7b-hydroxycholesterol may be
associated with an increased risk of atherosclerosis. Arterial injury by COPs
causes endothelial dysfunction and arterial wall cholesterol accumulation
(277). Even under normocholesterolemic conditions, COPs can cause endothelial
dysfunction, increased macromolecular permeability, and increased cholesterol
accumulation. These are all factors believed to be involved in the develop-
ment of atherosclerotic lesions. The atherogenic potential of COPs has been
demonstrated by in vitro cell culture (73, 278), as well as in animal feeding
studies (279). Japanese quail fed either purified cholesterol or oxidized
cholesterol exhibited greater plasma and liver cholesterol concentrations in
association with increased severity of atherosclerotic lesions when fed the
oxidized cholesterol (279).
Several COPs have been identified as having cytotoxic, angiotoxic, carcino-
genic, and mutagenic bioactivities (280, 281), all or some of which may play
a role in the initiation or proliferation of atherosclerotic plaque. Specifically,
25-hydroxycholesterol and cholestane-3,5a,6-triol are particularly toxic to
cultured rabbit aortic smooth muscle cells (278). The 25-hydroxycholesterol
has been found to be preferentially transported in VLDL (34.7%) and LDL

(55.1%), with HDL (10.2%) containing a much smaller proportion of the
oral dose administered to squirrel monkeys (281). Thus, potentially angiotoxic
COPs may be present in the lower density lipoprotein fractions of animals
that develop atherosclerotic lesions. In human studies, detectable amounts of
COPs, namely the cholesterol-5,6a- and -epoxides, the 7-hydroxycholesterol
isomers, and 7-ketocholesterol, were reported in arotic tissue (75, 80). These
same COPs have also been recovered in the LDL extracted from human
atherosclerotic plaques (74). The presence of 7-hydroxycholesterol in
tissues may be caused by the auto-oxidation of cholesterol, whereas the
7a-hydroxycholesterol isomer is known to be a product of in vivo oxidation
in bile acid synthesis (280). Other workers have been able to confirm the
identity of several COPs (cholest-3,5-diene-7-one, cholestanetriol, 7-hydroxy-
cholesterols, 7-ketocholesterol, 24-hydroxy-, 25-hydroxy-, and 26-hydroxy-
cholesterol) in human aortic specimens (75). In the Japanese quail and rat
species comparison study shown herein, the pattern of COPs present in aorta
of atherosclerosis susceptible (e.g., quail) was related to the hypercholesterole-
mia and aterial plaque score, not seen in the resistant (rat) species (Table 4).
Using GC-MS, 7-hydroxycholesterol and 7-ketocholesterol were identified
and quantitated in quail aortic tissue exhibiting plaque formation (Figure 4).
These COPs are identical to those reported from aortic plaques from both
humans as well as other animal models, thereby validating the use of the ather-
osclerosis susceptible Japanese quail for research in the area of experimental
atherogenesis.
The cytotoxicity and apoptosis-inducing potential of commonly occurring
oxysterols might not be dependent on the sole generation of an oxidative stress
(282). The potential genotoxicity of cholesterol oxidation products in two mam-
malian fibroblast cell lines showed that none of the COPs detected affected
baseline levels of DNA strand breaks or sister chromatid exchanges (283).
Thus, in this case, COPs were not genotoxic. Rather, a Ca2 influx through
plasma membrane channels could be an important signal in the mechanism of
COP-induced apoptosis (284). Thus, inflammatory cytokines may increase the
cytotoxicity of LOPs.
TABLE 4. Classification of Plant Sterols.1
Primary Series Examples
Cholesterol 4-desmtheyl sterols Campesterol, stigmesterol, b-sitosterol,

5 avenasterol, 7 stigmasterol,
7 avenasterol,
4-methyl-cholestane 4-monomeathyl sterols Atrastandienoligramisterol
Lanostane 4,40 -dimethylsterol Cycloarterol,
1
Ref. (90).
ADVERSE PRODUCTS FROM OVERHEATED FATS AND OILS 543
1
2
a c
1
2
b d
1
2
3
4
15 20 25 30 15 20 25 30
time (min.) time (min.)
Figure 4. Gas chromatograph-FID chromatogram of Wistar rat (A&B) and atherosclerosis-
susceptible Japanese quail (C&D) aorta derivatized nonsaponifiables. (A) 0.05% low cholesterol,
rat; (B) 0.5% high cholesterol, rat; (C) 0.05% low cholesterol, quail; (D) 0.5% high cholesterol
quail aorta. 1 internal standard, 5 a-cholestane; 2 cholesterol; 3 7b-hydroxycholesterol;
4 7-ketocholesterol. Samples were analyzed according to Yuan et al. (69).
6. ADVERSE PRODUCTS FROM OVERHEATED FATS AND OILS
Some investigators have found gender and organ-specific toxicity in normal and
malnourished rats fed thermoxidized palm oil (TPO) (285). The hearts of the first
offspring of male and female rats were enlarged, whereas the lung, liver, and
kidneys of first filial female offspring were reduced in size. This observation sug-
gests the toxicities of TPO could be cumulative for female offspring. TPO also
induces reproductive toxicity in healthy and malnourished rats (286). Fetotoxicity

was observed in TPO-fed rats, where the neonatal birth weights and litter size were
decreased; pregnancy rates in TPO-fed rats were also decreased (55%). The dangers
of thermally oxidized dietary fat for colon carcinogenesis in rodents are another
relevant effect of heating fats and oils (287). In such animal studies, extensively
oxidized beef tallow increased the number of animals with aberrant crypt focus
and average ACF per colon. It has also been shown that certain fractions of heated
fats (total polar materials) cause growth retardation, as well as increasing liver
and kidney weights, and can cause disorders of the enzyme system if fed in high
doses (288). An effect of oxidized dietary oil on plasma cholesterol and thyroid hor-
mone concentrations has been reported in miniature pigs fed on a hyperlipidaemic
diet (289). Consistent with rat studies, feeding oxidized oil reduced concentrations
of cholesterol in plasma and in LDL and HDL fractions. It also reduced the con-
centration of a-tocopherol in the plasma, reduced the concentration of tri-iodothyr-
onine in the plasma, and elevated the ratio between thyroxine and tri-iodothyronine.
The only time that long-term feeding effects of heated and fried oils on lipids and
lipoproteins in rats does not show any deleterious effect on growth, plasma, and
tissue lipid profile of rats is when the conditions of the heating or frying are not
too drastic and the oils are not heat-abused (290). In humans, studies have shown
impaired endothelial function following a meal rich in used cooking fat (291). For
example, the ingestion of a meal rich in fat previously used for deep-frying in com-
mercial fast food restaurant resulted in impaired arterial endothelial function. For
red palm oil, the effects of chronic consumption of fresh and heated red palm oil on
lipid profile and lipid peroxidation are comparable on serum cholesterol as well as
on lipid peroxidation, but prolonged heating of oil increases LDL-c levels, which
may make it more atherogenic (292).
The feeding of fried fish that had been fried in the same batch of oil over several
days to weanling rats confirmed several adverse effects of reheating dietary oil
(293). The weanling rats eventually exhibited a decreased feed consumption and
weight gain, a decrease in total lipid and cholesterol content of liver, but an increase
in total lipid and cholesterol in heart and serum cholesterol levels. These biochem-
ical changes corresponded to initial stages of cell damage in liver and kidney tis-
sues. More specifically, cyclic fatty acid monomers from heated oil modify the
activities of lipid synthesizing and oxidizing enzymes in rat livers (294). The feed-
ing of these cyclic fatty acid monomers has shown evidence of a peroxisome
proliferator-like effects. They also induced a coordinated regulation between
activities of the lipogenic enzymes studied (delta9-desaturase, phosphatidate phos-
phohydrolase) and peroxisomal oxidation. A dose-dependent decrease of delta9-
desaturase with cyclic fatty acid intake accompanied by a similar decrease of
MUFA level in the liver and the increase in gamma-linolenic acid level suggests
an increase in delta6-desaturase with cyclic fatty acid intake. In short, cyclic fatty
acid monomers affect different aspects of lipid metabolism, including a phenotypic
peroxisome proliferator response. The frying process has formed cyclic fatty acids
(295), and structures formed from oleic, linoleic, and linolenic acids in heated
frying oils have been determined.
TOXIC SUBSTANCES PRODUCED DURING SMOKING, CHARBROILING 545
Aromatic amines from cooking oil fumes are known to be carcinogenic for blad-
der cancer. Fume samples from three commercial cooking oils commonly used in
Taiwan were mutagenic in the presence of an S-9 mix (296). Exposure to cooking
oil fumes, such as those from safflower oil, vegetable oil, and corn oil, may also
increase exposure to PAHs, which, in turn, has been linked to an increased risk
of lung cancer (297). Frying sunflower oil can lead to a potentially toxic product,
causing decreased food efficiency ratio, growth retardation, and changes in liver
fatty acid composition if fed (298). The use and abuse of frying oil on the quality
of frying oil discarded by 16 catering establishments revealed that many discarded
oils contained free fatty acids that were higher in concentration than the recom-
mended safe level (299). A recent quality and sensory evaluation of used frying
oil from restaurants observed that sensory parameters are reliable indicators of
the quality of used frying oil (300). There appears to be a good correlation between
sensory evaluation and the actual analysis of total polar components and oxidized
acid levels. The frequent addition of fresh oil throughout the frying process of fro-
zen foods, such as prefried potatotes that undergo deep-fat frying in sunflower oil,
minimizes fatty acid changes (301).
7. TOXIC SUBSTANCES PRODUCED DURING SMOKING,

CHARBROILING, AND BARBECUING OF FOODS
The analysis of 200 food items for benzo[a]pyrene (B(a)P) content indicated that
the highest levels of B(a)P were found in grilled/barbecued, very well-done steaks
and hamburgers and in grilled/barbecued, well-done chicken with skin (302). The
formation of polycyclic aromatic hydrocarbons in the smoke from heated model
lipids and food lipids reflects model lipids as being more susceptible to smoke
formation than food lipids during heating (303). Methyl linolenate lipid produced
the highest amount of PAHs, followed by methyl oleate and methyl stearate.
Soybean oil generated larger amounts of PAHs than canola or sunflower oils, and
benzene-like compounds were found to be possible precursors for PAHs formation.
Numerous studies have been conducted to monitor the formation of PAHs in differ-
ent smoked meat products and smoke flavor additives (304306). Smoke flavorings
obtained from different types of wood render different concentrations of PAHs
(307). Flavoring from poplar wood appears to give the highest number and concen-
trations of both total and carcinogenic PAHs, but the storage of smoke flavorings
in polyethylene flasks reduces the concentration of some PAHs. The removal of
PAHs from water by migration into polyethylene is a process that has also been
investigated (308). PAHs are primarily absorbed on the polyethylene surface with
subsequent migration into the bulk polymer. This transportation of PAHs through
the bulk can be described by Fickian laws of diffusion and is consistent with the
theory of depth absorption of PAHs in polyethylene. Changes in total fat content,
fatty acid composition, tocopherol, ascorbic acid, pH, and oxidation often occur in
Atlantic salmon in response to either cold smoking (20 C or 30 C) or electrostatic
smoking (309). The leaner the fish, the higher percentile loss in fillet fat with these
(a) 12 1
11 2
10
3
9
8 4
7 6 5
(b) (d)
OH OH
OH OH
ROO ROO P450
(c) (e)
O O
OH OH
OH OH
Figure 5. Reactive oxygen species derived from co-oxidation catalyst of benzy(a)pyrene
(B(a)P) diol oxidation. (a) B(a)P; (b) ()B(a)P diol; (c) ()anti-diolepoxide B(a)P enantiomer; (d)
()-B(a)P diol; (e) ()-anti-diolepoxide enantiomer. ROO peroxyl radical; P-450 monoxygen-
ase. The ()B(a)P-7,8 diol enantiomer is generated from (B(a)P in vivo from both ROO and
cytochrome P-450 (6).
smoking processes. Regardless of smoking temperature, ascorbic acid decreased

about 80% from the fresh value in cold-smoked fish, whereas dry-salted and
electrostatically smoked fish lost only about 10% of the fresh ascorbic acid content.
Electrostatically smoked fish had a smaller drop in fillet pH than cold-smoked
fillets.
Investigations that examined the direct mutagenicity of the polycyclic aromatic
hydrocarbon-containing fractions of smoked and charcoal-broiled foods treated
with nitrite in acid solution indicated that nitrite could convert most samples to
direct-acting mutagens toward both strains (310). The treatment of PAHs with
nitrite in acid solution produced some non-N-nitroso direct-acting mutagens, which
may belong to nitro-PAHs. Thus, consumption of charcoal-broiled and smoked
foods, along with nitrite, is generally not recommended. Animal studies have
shown that B(a)P treatments will induce squamous cell carcinoma, papiloma, and
hyperplasia in the forestomach at incidences of 18% and 91%, respectively (311).
B(a)P also produced malignant lymphoma with an incidence of 18%, an increased
POTENTIAL HAZARDS FROM GOVERNMENT 547
leukocyte count and decreased erythrocyte count, and a decreased body weight. In
addition, bronchiolar-alveolar hyperplasia in the lung at incidences of 18% and 9%,
respectively, were induced. Further risks of frying foods have been identified, such
as those caused by the airborne particulates generated during frying of beef, fish,
and pork, which can induce carcinogen-metabolizing cytochromes P450 1A1 and
1B1 in human lung-derived cell line CL5 (312). The mutagenic risk posed by sim-
ple, well-characterized mixtures of priority PAHs are estimated to be the sum of the
risks posed by the mixture components (313). Peroxyl- radical and cytochrome
P450 dependent oxidation reactions contribute to co-oxidation processes and
increases both the number and the reactive toxicity of products derived from
auto-oxidation of xenobiotic agents, such as benzo(a)pyrene (Figure 5).
8. POTENTIAL HAZARDS FROM GOVERNMENT

APPROVED ANTIOXIDANTS
A review of butylated hydroxyanisole (BHA) and butylated hydroxy toluene (BHT)

experimental studies on potential genotoxicity and carcinogenicity concluded that
BHA and BHT pose no cancer hazard and may be anticarcinogenic at the current
levels of food additive use (314). However, both carcinogenic and anticarcinogenic
properties have been reported for BHA and BHT. The association between the
dietary intake of BHA and BHT and stomach cancer risk was studied in a
Netherlands cohort study that started in 1986 (315). There was no significant asso-
ciation found between usual intakes of BHA and BHT and stomach cancer risk. In a
rat study, the forestomach carcinogenicity of BHA was compared in males of the
F344, SHR, Lewis, and Sprague-Dawley rat strains (316). Forestomach squamous
cell papillomas and hyperplasias were developed in all rats given BHA, but the
incidence of squamous cell carcinomas (SCCs) was different, with the highest in
the SHR strain. Cytotoxicity was also the most severe in the SHR strain. Thus,
rat strain differences in BHA forestomach carcinogenesis can exist and sensitivity
to cytotoxicity may be an important factor. The growth modulation of enzyme-
altered preneoplastic liver foci (EAF), by BHT feeding to rats bearing enzyme-
altering hepatic preneoplastic lesions, has shown that the induction of EAF and
BHT treatment resulted in a reduction in the natural killer cell activity of spleno-
cytes (317). This response could be a contributing factor in the enhancement of
rodent liver neoplasia by BHT (318). When BHT was fed to rats, an increased liver
weight corresponding to a gradual vacolization of liver cells, cytoplasmic disinte-
gration, and hepatocellular necrosis, were detected without any sign of tumorigeni-
city . In mice, BHT causes lung injury and promotes tumor formation. Thus, a
recent study looked at the hydroxylation of a tert-butyl group on BHT, which yields
6-tert-butyl-2-[20 -(20 -hydroxymethyl)-propyl]-4-methylphenol (BHTOH) (319).
BHTOH is more potent than BHT. BHT, BHTOH, and other BHT metabolites
were compared in respective affinities to kill nontumorigenic and tumorigenic
mouse and human lung cell lines. BHTOH was the most toxic and induced
apoptosis the greatest. In other animal studies, feeding broiler chickens BHT for
six weeks resulted in hyperaemia, sinusoidal distension, focal necrosis, and mono-
nuclear cell infiltration, along with fatty changes in liver, hyperaemia, tubular
degeneration, fibrosis, and cloudy swelling in the kidneys (320). Thus, BHT is
hepatotoxic and nephrotoxic in broilers. The effects of BHA and BHT on DNA
adduct formation and arylamines in N-acetyltransferase activity in PC-3 cells
(human prostate tumor) in vitro showed that the higher the concentration of
BHA or BHT, the higher the inhibition of N-acetyltransferase (NAT) activity
(321). Exposure to BHA or BHT also decreased DNA adduct formation in PC-3
cells. A Brazilian study that attempted to estimate yields of the theoretical
maximum daily intake of phenolic antioxidants BHA, BHT, and TBHQ, found
that the current acceptable daily intakes of BHA, BHT, and TBHQ were unlikely
to be exceeded by the average Brazilian consumer (322). Oxidative DNA damage
and apoptosis can be induced by metabolites of BHT such as BHT-quinone,
BHT-OOH, and BHT-CHO (323). It is noteworthy that where BHT-OOH parti-
cipates in oxidative DNA damage directly, BHT-quinone causes DNA damage
through H2O2 generation, which leads to internucleosomal DNA fragmentation.
8.1. Allylbenzenes
Allylbenzene analogs, including safrole, eugenol, and estragol, are flavor com-
pounds derived from essential oils and metabolized by biotransformation in the
liver to form potentially reactive electrophilic intermediates. The major route of
bioactivation is via the hydroxylation of the 10 carbon atom on the allylic side chain.
The 20 ,30 -allylic epoxide derivatives of allylbenzene, estragole, eugenol, and safrole
have been used to determine the genotoxicity of epoxidation at the allylic double
bond for allylbenzene and its analogs (324). The epoxide formation at the allylic
double bond is a potentially genotoxic bioactivation pathway for allylbenzene ana-
logs. Although the epoxidation pathway of allylbenzene poses a potential genotoxic
threat to humans, no actual genotoxicity occurs as a result of its further metabolism.
The essential oils of dill (Anethum graveolens L.), peppermint (Mentha
piperita L.), and pine (Pinus sylvestris L.) have been found to be cytotoxic for
human lymphocytes (325). These oils have different levels of activity for chromo-
some aberration and sister chromatid exchange in human lymphocytes, with dill
seeds usually being the most active. In contrast to this activity, a dose-dependent
increase in mutation frequency has been found in pine and dill herb oils, where
dill seed oil has been found to be almost inactive. Peppermint oil has been found
to have a dose-independent effect on mutations. In terms of the genotoxicity of
alkenylbenzenes in rodents, it has been accepted that a -and b-asarone are hepato-
carcinogenic, whereas two other alkenylbenzenes, myristicin and elimicin, are not.
Elimicin and a - and b-asarone, but not myristicin, are genotoxic in an unscheduled
DNA synthesis assay using rat hepatocytes (326). Simple allylbenzenes, such as
safrole and estragole, are activated by 1-hydroxylation and sulfation, and this is
the likely mechanism of the genotoxicity of elimicin. Moreover, the propenyl ana-
logues isosafrole, anethole, and methylisoeugenol, which cannot undergo 1-hydro-
xylation, are not genotoxic. Another comparative induction of unscheduled DNA
POTENTIAL HAZARDS FROM GOVERNMENT 549
synthesis in cultured rat hepatocytes by allylbenzenes and 10 -hydroxy metabolites

10 -hydroxyestragole, -methyleugenol, and safrole indicated greater genotoxicity
than parent compounds, the allylbenzenes (327).
Pretreatment of mice with trans-anethole and eugenol produced antigenotoxic
effects against cyclophosphamise (CPH), procarbazine (PCB), N-methyl-N-nitro-
N-nitrosoguanidine (MNNG), and urethane (URE). Trans-anethole also inhibits
the genotoxicity of ethyl methane sulfonate (EMS) (328). Both trans-anethole
and eugenol gave a dose-related antigenotoxic effect against PCB and URE. Under
in vitro conditions, trans-anethole does not increase the mutant frequency in the
Salmonella/microsome test, but a dose-related response occurs in the L5178Y
mouse lymphoma TK/- assay with metabolic activation (329). Trans-anethole
also induces chromosome aberrations in vitro in Chinese hamster ovary cells, but
does not induce chromosome aberrations. A continuous intake of high doses of
trans-anethole in rats will induce a continuum of cytotoxicity, cell necrosis, and cell
proliferation (330). Regardless, trans-anethole has been reaffirmed as GRAS
(generally recognized as safe) because of a functional metabolic detoxication
pathway in humans, and its lack of mutagenic or genotoxic potential.
Safrole has been found to be a weak hepatocarcinogen, which is linked to the
formation of safrole-DNA adducts. Safrole treatment will induce oxidative damage
in rat hepatic tissue (331) and is a genotoxic carcinogen in rat liver in vivo (332).
These cytogenetic effects may result from covalent DNA modification in the rat
liver. The threshold cytotoxic concentration found for safrole is 0.10 mug/ml safrole
(333).
8.2. Migration of Packaging Materials

The effects of different plastic films (PET, PVC, PP, and PS) on the stability of
olive, sunflower, and palm oils at 24 C and 37 C during 60 days storage, indicated
that changes in peroxide and TBA values were higher in plastic bottles than in glass
(334). Plastic permeability has a major effect on oil stability, with both BHA and
BHT capable of leaching out from plastic films into oils. PVC has the highest sta-
bility of oil samples, followed by PET, PP, and PS. Increasing storage temperature
accelerates the oxidation and limited stability of vegetable oils.
The exposure of pregnant female rats to different concentrations of bis (2-Ethyl-
hexyl) phthalate (DEHP) in drinking water produced a decrease in the kidney and
testes weights, and increase in liver weight (335). Signs of histological damage in
kidneys, liver, and testes were observed, and pups exposed to the highest DEHP
concentration had an increase in time required to perform a learned avoidance
test. The NOAEL for DEHP and DNOP, respectively, are estimated to be 50 ppm
or 3.7 mg/kg/day for DEHP and 500 ppm or 36.8 mg/kg/day for DNOP (336).
Unfortunately, it has been reported that retail packaging of small portions of cheese,
even in a low-migration PVC cling film, may lead to consumer intakes of DEHA
close to or above the tolerable daily intake of 0.3 mg/kg body weight (337). Investi-
gators have evaluated the significance of the potential contamination of polyethy-
lene terephthalate bottles when reused as food packaging material (338). Changes
in product flavor, caused by aroma sorption and the transfer of undesirable flavors
from packaging to foods, are important mechanisms of deterioration when foods are
packaged in polymer-based materials (339). Product considerations include sensi-
tivity to flavor and related deteriorations, color changes, vitamin loss, microbial
activity, and amount of flavor available. Storage considerations include tempera-
ture, time, and processing method.
The rate of migration of styrene from general-purpose polystyrene indicates a
relatively weak dependence of the diffusion coefficient on the residual styrene con-
tent and a strong dependence on temperature. Monitoring the level of styrene
migration from PS cups in different foods indicates that the styrene migration is
strongly dependent on fat content and storage temperature (340). Styrene mono-
mers, styrene dimers, and styrene trimers that migrated from polystyrene containers
into instant food exhibit no endocrine-disrupting effects that include apparent estro-
genic, androgenic, anti-androgenic, and thyroid activity (341). Blends of nylon 6
and ethylene-co-vinyl alcohol have been used as innovative food-packaging mate-
rials (342).
9. MANUFACTURING HAZARDS IN PROCESSING

CRUDE OILS AND FATS
9.1. Pesticide Residue, PAH Contamination,

and Polychlorinated Biphenyls
Samples of crude fish oil refined and at each step of the refining process have been
analyzed for organochlorine pesticides and PCB (343). The deodorization step
causes a decrease in the amount of contaminants, especially the most volatile com-
pounds. Also, the concentrations of less volatile organochlorine pesticides and PCB
are reduced to about half the concentration in the crude oil. Cooked lake trout has
significantly less PCB and PAH residue than raw (344). Interestingly, smoking
results in a greater losses of pesticides and PCBs than other cooking methods,
but PAH showed greater formation during smoking. Consumption of large quanti-
ties of contaminated fish oils by children may lead to behavioral and neurological
effects (345). In food samples obtained from Catalonia, a toxic potency of poly-
chlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs),
and polychlorinated biphenyls (PCBs) was reported (346). TEQ (PCB) contribution
varied from 27% in olive oil to 81% in mussel samples. Thus, regulation of TEQ
content in foods should include not only the TEQ(PCDD/F), but also TEQ(PCB). In
January 1999, in Belgium, 500 tonne of feed contaminated with PCB and dioxins
were distributed to animal farms, resulting in contamination of the food chain
(347). The contamination was caused by transformer oil, rather than other environ-
mental sources, and the mean intake/kg body weight for Belgians was estimated to
be maximally 25,000 ng PCB and 500 pg international TEQ dioxins. An estimated
total number of cancers resulting from this incident ranged between 40 and 8000.
Neurotoxic and behavioral effects in neonates were expected, but not quantified,
MANUFACTURING HAZARDS IN PROCESSING CRUDE OILS 551
therefore leading to a reassessment of PCDD/PCDFs and Co-PCBs toxicity in con-

taminated rice bran oil responsible for the Yu-Cheng disease (348). There is evi-
dence for PCBs as neurodevelopmental toxicants in humans from previous studies,
such as two U.S. prospective longitudinal studies that analyzed cord, maternal
serum, and maternal milk samples, and a Taiwanese study on infants exposed to
PCB and polychlorinated dibenzofurans because of maternal ingestion of highly
contaminated rice oil (349). The response to shellfish contamination following an
oil spill on the Oregon coast and some of the complication factors affecting evalua-
tion of potential health risks from the consumption of oil-contaminated shellfish has
been reported (350).
9.2. Toxic Oil Syndrome

Oil refined by ITH and distributed by RAELCA was the principle, and probably
only, oil responsible for the TOS epidemic that occurred in Spain in 1981 (351).
Regarding this incident, the oil contaminants have been detected and a possible
chemical link between oil contaminants and those detected in L-tryptophan, impli-
cated in the eosinophilia-myalgia syndrome (EMS), has been found (352). TOS is
strongly associated with the consumption of oils containing fatty acid esters of 3-
(N-phenylamino)-1,2-propanediol (PAP), but it is unknown whether PAP esters are
simply markers of toxicity of oils or have the capability to induce the disease (353,
354). One study assessed the effect of several refining process variables on the for-
mation of PAP esters (355). These variables include storage time prior to refining as
well as elevated refining temperatures. Among the survivors of TOS, the prevalence
of some chronic conditions (e.g., sclerodermia, neurologic changes) is high twenty
years after its onset. 3-(N-phenylamino)-1,2-propanediol PAP esters are absorbed in
the gi tract and are distributed and stored in different organs, particularly in the liver
and brown adipose tissue (356). PAP in these organs showed different patterns of
fatty acids, indicating the ability of the gi tract to modify the fatty acid composition
of the parent PAP. Also, some PAP esters, when a long acyl chain was present in the
sn-1 position of the molecule, showed an inhibitory effect on the PAF synthesis.
This is an important observation in line with the systemic nature of TOS. A recent
study of a cohort of 758 TOS patients found an increased prevalence of cardio-
vascular risk factors, such as arterial hypertension (34%), dyslipemias (44%), over-
weight (40%), obesity (27%), carbohydrate intolerance (9%), and diabetes mellitus
(9.4%) (357). The most common reported symptoms were cramps (78%), arthral-
gias (78%), and paresthesias (70%). Only 2.8% of patients reported to be asymp-
tomatic. The analytical results most common were changes in lipidic and
carbohydrate metabolism, overt or subclinical hypothyroidism (6.6%) and respira-
tory changes in patients with no previous pulmonary disease, changes in spirometry
(6%), diffusion test (8%), and hypoxemia (18%).
When guinea pigs were fed diets supplemented with oil related to TOS, the
TOS-related oil produced a decrease in lipid peroxidation products with minimal
alterations in phospholipids fatty acid composition of liver microsomes (358).
TOS is an exogenously induced autoimmune disease in humans, believed to be a
result of the ingestion of oleic acid anilides. When the in vitro effects of anilides on
splenocytes and T cells in A/J and B10.S mice were compared, it was shown that
anilides are able to affect the immune system in a strain-dependent way and may
therefore take part in inducing TOS in humans and mice (359).
10. CONCLUSIONS
The combined increased awareness by the consumer concerning the health and
safety of fat consumption and the technological advances made by the food industry
to provide a safe and nutritious product with sensory appeal for human consump-
tion has resulted in a relatively safe and highly scrutinized food matrix. There are,
however, situations of untold consumer exposure to invisible lipid-soluble xenobio-
tic agents that enter this particular food matrix from both intentional addition, or as
a consequence of environmental contamination or reactions of labile constituents.
In these cases, risk assessment strategies are required to predict potential harm to
consumers exposed to tolerable intakes of these materials. On the other hand, fats
and oils contain many important nutrients and extranutrients that have important
roles in maintaining healthy lifestyles. In particular, with the advent of functional
foods and nutraceuticals, many specific food products that contain lipid nutrients or
lipid-soluble bioactive constituents will reach the consumer with the purpose of
enhancing health and wellness.
REFERENCES
1. J. E. Kinsella, B. Lokesh, and R. A. Stone, Amer. J. Clin. Nut., 52, 1 (1990).

2. M. B. Katan, S. M. Grundy, P. Jones, M. Law, T. Miettinen, and R. Paoletti, Mayo Clin.
Proc., 78, 965 (2003).
3. A. C. Carr, B. Z. Zhu, and B. Frei, Circ. Res., 87, 349 (2000).
4. H. Frandsen, H. Frederiksen, and J. Alexander, Food Chem. Toxicol., 40, 1125 (2002).
5. M. D. Easton, D. Luszniak, and E. Von der Geest, Chemosphere, 46, 1053 (2002).
6. D. D. Kitts, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, 5th ed., John
Wiley & Sons, Inc., New York, 1996, pp. 215280.
7. L. Y. Sheen, H. W. Chen, Y. L. Kung, C. T. Liu, and C. K. Lii, Nutr. Cancer., 35, 160
(1999).
8. J. E. Thmlinson, R. Nakayama, and D. Holten, J. Nutr., 118, 408 (1988).
9. J. F. Focant, C. Pirard, and E. De Pauw, Chemosphere, 54, 137 (2004).
10. E. N. Frankel, Chem. Phys. Lipids, 44, 73 (1987).
11. I. Staprans, D. A. Hardman, X. M. Pan, and K. R. Feingold, Diabetes Care., 22, 300
(1999).
12. I. Hayam, U. Cogan, and S. Mokady, J. Nutr. Biochem., 4, 563 (1994).
13. R. Wilson, K. Lyall, L. Smyth, C. E. Fernie, and R. A. Riemersma, Free Rad. Biol. Med.,
32, 162 (2002).
REFERENCES 553
14. I. Hayam, U. Cogan, and S. Mokady, Biosci. Biotech. Biochem., 61, 1011 (1997).
15. P. M. Chao, C. Y. Chao, and C. Huang, J. Nutr., 131, 3166 (2001).
16. C. M. Yang, C. W. Kendall, D. Stamp, A. Medline, M. C. Archer, and W. R. Bruce, Nutr.
Cancer, 30, 69 (1998).
17. K. Traul, A. Driedger, D. Ingle, and D. Nakhasi, Food Chem. Toxicol., 38, 29 (2000).
18. N. B. Cater, Food Chem. Toxicol., 35, 428 (1997).
19. Y. Tsai, S. Park, J. Kovacic, and J. Snook, Lipids, 34, 895 (1999).
20. A. Lichtenstein, and U. Schwab, Atherosclerosis, 150, 227 (2000).
21. Y. V. Yuan, D. D. Kitts, and D. V. Godin, J. Nutr., 128, 1620 (1998).
22. Y. V. Yuan and D. D. Kitts, J. Nutr., 133, 679 (2003).
23. P. Khosla and K. C. Hayes, Biochem. Biophys. Acta., 1210, 13 (1993).
24. A. Pronczuk, P. Khosla, and K. C. Hayes, FASEB J., 8, 1191 (1994).
25. F. Hu, R. van Dam, and S. Liu, Diabetologia, 44, 805 (2001).
26. J. Lovejoy, Current Atheroscler. Rep., 1, 215 (1999).
27. M. Manco, G. Mingrone, A. Greco, E. Capristo, D. Gniuli, A. De Gaetano, and
G. Gasbarrini, Metabolism, 49, 220 (2000).
28. B. Vessby, M. Uusitupa, K. Hermansen, G. Riccardi, A. Rivellese, L. Tapsell, C. Nalsen,
L. Berglund, A. Louheranta, B. Rasmussen, G. Calvert, A. Maffetone, E. Pederen,
I. Gustafsson, and L. Storlien, Diabetologia, 44, 312 (2001).
29. F. Perez-Jimenez, J. Lopez-Miranda, and P. Mata, Atheroscleor., 163, 385 (2002).
30. D. OBryne, S. OKeefe, and R. Shireman, Lipids, 33, 149 (1998).
31. B. Halvorsen, A. Rustan, L. Madsen, J. Reseland, R. Berge, P. Sletnes, and
E. Chrstiansen, Annals Nutr. Metab., 45, 30 (2001).
32. D. OBryne, K. Knauft, and R. Shireman, Lipids, 32, 687 (1997).
33. E. Sarkkinen, U. Schwab, L. Niskanen, M. Hannuksela, M. Savolainen, K. Kervinen,
A. Kesaniemi, and M. Uusitupa, European J. Clin. Nutr., 50, 592 (1996).
34. P. Kris-Etherton, J. Nutr., 129, 2280 (1999).
35. P. Kris-Etherton, T. Pearson, Y. Wan, R. Hargrove, K. Moriarty, V. Fishell, and
T. Etherton, Amer. J. Clin. Nutr., 70, 1009 (1999).
36. C. Aguilera, M. Ramirez-Tortosa, M. Mesa, and A. Gil, Recent Res. Develop. Lipids
Res., 4, 369 (2000).
37. P. Yaqoob, Eur. J. Clin. Nutr., 56, S3:S9 (2002).
38. K. Ando, K. Nagata, R. Yoshida, K. Kikugawa, and M. Suzuki, Lipids, 35, 401 (2000).
39. M. Fukushima, T. Matsuda, K. Yamagishi, and M. Nakano, Lipids, 32, 1069 (1997).
40. O. A. Al-Shabanah, Food Chem. Toxicol., 35, 769 (1997).
41. M. Wettasinghe and F. Shahidi, Food Chem., 70, 17 (2000).
42. M. M. Engler and M. B. Engler, Nutr. Res., 18, 1533 (1998).
43. Y. Naito, K. Kasama, H. Yoshida, and N. Ohara, Food Chem. Toxicol., 38, 811 (2000).
44. J. D. Palombo, S. J. DeMichele, J. W. Liu, B. R. Bistrian, and Y. S. Huang, Lipids, 35,
975 (2000).
45. T. Kumamoto and T. Ide, Lipids, 33, 647 (1998).
46. M. Fukushima, T. Ohhashi, S. Ohno, H. Saitoh, K. Sonoyama, K. Shimada, M. Sekikawa,
and M. Nakano, Lipids, 36, 261 (2001).
47. J. B. Hansen, R. K. Berge, A. Nordoy, and K. H. Bonaa, Lipids, 33, 1123 (1998).
48. P. Weber, Lipids, 34, S269 (1999).
49. M. Noguchi, Food Chem. Toxicol., 35, 633 (1997).
50. A. Nordoy, R. Marchioli, H. Arnesen, and J. Videback, Lipids, 36, S127S129 (2001).
51. C. J. Huang and M. L. Fwu, J. Nutr., 122, 1182 (1992).
52. S. D. Clark and D. B. Jump, Prog. Lipid Res., 32, 139 (1993).
53. L. Demaison and D. Moreau, Cell Mol. Life Sci., 59, 463 (2002).
54. Y. Y. Fan, K. S. Ramos, R. S. Chapkin, in 91st AOCS annual Meeting and Expo,
San Diego, California, 2001.
55. Y. Y. Fan, K. S. Ramos, and R. S. Chapkin, J. Nutr., 131, 1675 (2001).
56. S. Hrelia, A. Bordoni, P. Biagi, C. Rossi, L. Bernardi, D. Horrobin, and A. Pession,
Biochem. Biophys. Res. Commun., 225, 441 (1996).
57. R. K. Furse, R. G. Rossetti, and R. B. Zurier, J. Immunol., 167, 490 (2001).
58. S. Kaku, K. Ohkura, S. Yunoki, M. Nonak, H. Tachibana, M. Sugano, and K. Yamada,
Prostaglandins Leuko. Essential Fatty Acids, 65, 205 (2001).
59. W. Jiang, S. Hiscox, D. Horrobin, R. Bryce, and R. Mansei, Biochem. Biophys. Res.
Commun., 237, 639 (1997).
60. J. Simon, J. Fong, J. Bernert, and W. Browner, Amer. J. Epidemiol., 148, 854 (1998).
61. T. Coste, M. Pierlovisi, J. Leonardi, D. Dufayet, A. Gerbi, H. Lafont, P. Vague, and
D. Raccah, J. Nut. Biochem., 10, 411 (1999).
62. V. A. Ziboh, M. Yun, D. M. Hyde, and S. N. Giri, Lipids, 32, 759 (1997).
63. J. Ritskes-Hoitinga, P. Verscuren, G. Meijer, A. Wiersma, A. van de Kooij, W. Timmer,
C. Blonk, and J. Weststrate, Food Chem. Toxicol., 36, 663 (1998).
64. Y. Ko, C. Lij, C. Ou, J. Liu, W. Lin, and H. Chen, J. Agric. Food Chem., 48, 4144 (2000).
65. A. Demoz, N. Willunsen, and R. K. Berge, Lipids, 27, 968 (1992).
66. R. Hempenius, B. Lna, and R. Hagitt, Food Chem. Toxicol., 38, 127 (2000).
67. E. Koskelo, K. Boswell, L. Carl, S. Lanoue, C. Kelly, and D. Kyle, Lipids, 32, 397
(1997).
68. L. Arterburn, K. Boswell, T. Lawlor, M. Cifone, H. Murli, and D. Kyle, Food Chem.
Toxicol., 38, 971 (2000).
69. Y. V. Yuan, D. D. Kitts, and D. V. Godin, Br. J. Nutr., 80, 89 (1998).
70. G. E. Mott, E. M. Jackson, C. A. McMahan, and H. C. McGill, J. Nutr., 122, 1397 (1992).
71. Lipid Research Clinic Program, J. Amer. Med. Assoc., 251, 351 (1984).
72. J. Masuda and R. Ross, Arteriosclerosis, 10, 164 (1990).
73. S. K. Peng, C. B. Taylor, J. C. Hill, and R. J. Morin, Atheroslerosis, 54, 121 (1985).
74. U. P. Steinbrecher and M. Lougheed, Arterio. Thromb., 12, 608 (1992).
75. R. J. Morin and S. K. Peng, in Biological Effects of Cholesterol Oxides, CRC Press, Inc.,
76. P. B. Addis, Food Nutr. News, 62, 1 (1990).
77. M. D. Stringer, P. G. Gorog, A. Freeman, and V. V. Kakkar, Brit. Med. J., 298, 281 (1989).
78. M. E. Rosenfeld, W. Palinski, S. Yia-Herttuala, S. Butler, and J. L. Witztum, Arterio-
sclerosis, 10, 336 (1990).
REFERENCES 555
79. W. Palinski, M. E. Rosenfeld, S. Yla-Herttuala, G. C. Gurtner, S. S. Socher, S. W. Butler,

S. Parthasarathy, T. E. Carew, and D. Steinberg, Proc. Natl. Acad. Sci. USA, 86, 1372
(1989).
80. S. Dzeletovic, O. Breuer, E. Lund, and U. Diczfalusy, Anal. Biochem., 225, 73 (1995).
81. M. T. Quinn, S. Parthasarathy, L. G. Fong, and D. Steinberg, Proc. Natl. Acad. Sci. USA,
84, 2995 (1987).
82. U. P. Steinbrecher, Can. J. Cardiol., 9, 3B (1993).
83. P. Debruyne, E. Bruyneel, X. Li, A. Zimber, C. Gespach, and M. Mareel, Mut. Res., 480,
359 (2001).
84. N. Caillol, E. Pasqualini, E. Mas, A. Valette, A. Verine, and D. Lombardo, Lipids, 32,
1147 (1997).
85. A. Benedetti, D. Alvaro, C. Bassotti, A. Gigliozzi, G. Feretti, T. La Rosa, A. Di Sario,
L. Baiocchi, and A. Jezequel, Hepatology, 26, 9 (1997).
86. K. Phillips, M. Tarrago-Trani, and K. Stewart, Food Chem., 64, 415 (1999).
87. V. Piinoren, D. G. Lindsay, T. A. Meittinen, J. Toivo, and A. M. Lampi, J. Sci. Food
Agric., 80, 939 (2000).
88. F. Shahidi and V. K. S. Shukla, Inform, 7, 1227 (1996).
89. T. A. Miettinen and H. Gylling, Curr. Opin. Lipidol., 10, 9 (1999).
90. P. Jones and M. Raeini-Sarjaz, Nutr. Rev., 59, 21 (2001).
91. H. Gylling, P. Puska, E. Vartiainen, and T. A. Miettinen, J. Lipid Res., 40, 593 (1999).
92. T. A. Niettinen and H. C. Gylling, Curr. Opin. Lipidol., 10, 9 (1999).
93. T. T. Nguyen, J. Nutr. Sci., 129, 2109 (1999).
94. V. Baker, P. Hepburn, S. Kennedy, P. Jones, L. Lea, J. Sumpter, and J. Ashby, Food Chem.
Toxicol., 37, 13 (1999).
95. P. Hepburn, S. Horner, and M. Smith, Food Chem. Toxicol., 37, 521 (1999).
96. D. Waalkens-Berendsen, A. Wolterbeek, M. Wijnands, M. Richold, and P. Hepburn,
97. J. Weststrate, R. Ayesh, C. Bauer-Plank, and P. Drewitt, Food Chem. Toxicol., 37, 1063
(1999).
98. R. Ayesh, J. Weststrate, P. Drewitt, and P. Hepburn, Food Chem. Toxicol., 37, 1127
(1999).
99. A. Wolfreys and P. Hepburn, Food Chem. Toxico., 40, 461 (2002).
100. D. Sanders, H. Minter, D. Howes, and P. Hepburn, Food Chem. Toxicol., 38, 485
(2000).
101. M. Moghadasian, Food Chem. Toxicol., 35, 528 (1997).
102. M. Moghadasian, H. Mohammed, and J. Frohlich, J. Medicine, 107, 588 (1999).
103. M. Davidson, K. Maki, D. Umporowicz, K. Ingram, M. Dicklin, E. Schaefer, R. Lane,
J. McNamara, J. Ribay-Mercado, G. Perrone, S. Robins, and W. Franke, J. Amer. College
Nutr., 20, 307 (2001).
104. P. J. H. Jones, F. Y. Nianios, M. Raaeini-Sarjaz, and C. A. Vanstone, Amer. J. Clin. Nutr.,
69, 1140 (1999).
105. H. Gylling, P. Puska, E. Vartiainen, and T. Miettinen, Atheroscler., 145, 279 (1999).
106. H. Relas, H. Gylling, and T. Miettinen, Br J. Nutr., 85, 141 (2001).
107. J. Plat, E. van Onselen, M. van Heugten, and R. Mensink, Eur. J. Clin. Nutr., 54, 671
(2000).
108. J. Plat and R. Mensink, Metabolism: Clinical Experimental, 50, 520 (2001).
109. M. Noakes, P. Clifton, F. Ntanios, W. Shrapnel, I. Record, and J. McInerney, Amer.
J. Clin. Nutr., 75, 79 (2002).
110. M. Mora, C. Tourne-Peteilh, M. Charveron, B. Fabre, A. Milon, and I. Muller, Chem.
Phys. Lipids, 101, 255 (1999).
111. A. Awad and C. Fink, J. Nutr., 130, 2127 (2000).
112. A. Awad, C. Fink, H. Williams, and U. Kim, Eur. J. Cancer Prevention, 10, 507 (2001).
113. R. Russell, Amer. J. Clin. Nutr., 71, 878 (2000).
114. J. Erickson and A. Mawson, J. Theoretical Biol., 206, 47 (2000).
115. B. Bray, M. Goodin, R. Inder, and R. Rosengren, Food Chem. Toxicol., 39, 1 (2001).
116. G. Mulder, N. Manley, J. Grant, K. Schmidt, W. Zeng, C. Eckhoff, and L. Maggio-Price,
Teratology, 62, 214 (2000).
117. B. Benoist, J. de Martines, and T. Goodman, Food Nutr. Bull., 22, 320 (2001).
118. A. Hendrickx, P. Peterson, D. Hartmann, and H. Hummler, Reproductive Toxicol., 15,
445 (2001).
119. U. Wiegand, S. Hartmann, and H. Hummler, Intern. J. Vit. Nutr. Res., 68, 411 (1998).
120. G. Mao, M. Collins, and F. Derguini, Toxicol. Appl. Pharmacol., 163, 38 (2000).
121. M. Murata and S. Kawanishi, J. Biol. Chem., 275, 2003 (2000).
122. F. Dal-Pizzol, F. Klamt, M. Frota, Jr., L. Moraes, J. Moreira, and M. Benfato, Free Rad.
Res., 33, 677 (2000).
123. A. Woodall, S. Lee, R. Weesie, M. Jackson, and G. Britton, Biochim. Biophys. Acta,
1336, 33 (1997).
124. K. Scott and D. Rodriquez-Amaya, Food Chem., 69, 125 (2000).
125. H. Nakamura, F. Furukawa, A. Nishikawa, M. Miyauchi, H. Son, T Imazawa, and
M. Hirose, Food Chem. Toxicol., 39, 799 (2001).
126. N. Ymashita, M. Murata, S. Inoue, M. Burkitt, L. Milne, and S. Kawanishi, Chem. Res.
Toxicol., 11, 855, 1998).
127. D. Liebler and J. Burr, Lipids, 35, 1045 (2000).
128. P. Poon, P. Y. Y. Wong, P. Dubeski, T. D. Durance, and D. D. Kitts, J. Sci. Food Agric., 83,
542 (2003).
129. E. Makinen and A. Hopia, Lipids, 35, 1215 (2000).
130. C. Faustman, D. Liebler, and J. Burr, J. Agri. Food Chem., 47, 1396 (1999).
131. D. Cornwell, K. Jones, Z. Jiang, L. Lantry, P. Southwell-Keely, I. Kohar, and
D. Thornton, Lipids, 33, 295 1998).
132. A. Carr, B. Zhu, and B. Frei, Circ. Res., 87, 349 (2000).
133. B. McIntyre, K. Briski, M. Tirmenstein, M. Fariss, A. Gapor, and P. Sylvester, Lipids, 35,
171 (2000).
134. H. Shige, T. Ishikawa, M. Suzukawa, M. Nishiwaki, T. Yamashita, K. Nakajima, T. Ito,
K. Higashi, M. Ayaori, A. Yonemura, P. Nestel, and H. Nakamura, Lipids, 33, 1169
(1998).
135. R. Vieth, Amer. J. Clin. Nutr., 69, 842 (1999).
REFERENCES 557
136. M. Terry, M. Lanenga, J. NcNaughton, and L. Stark, Vet. Human Toxicol., 41, 312
(1999).
137. Y. Guven, S. Yildirim, and C. Kasapoglu, Biol. Trace Element Res., 37, 85 (1993).
138. J. Wimsatt, S. Marks, T. Campbell, J. Johnson, and R. Nachreiner, J. Vet. Internal.
Medicine, 12, 42 (1998).
139. D. Li, H. Liu, P. Xi, Y. Chen, and Y. Li, J. China Agric. Univ., 6, 87 (2001).
140. S. Rajasree, P. Umashankar, A. Lal, P. Sarma, and C. Kartha, Life Sci., 70, 1777
(2002).
141. E. Gimeno, E. Costa, M. Gomar, A. Massone, and E. Portiansky, J. Vet. Medicine
Physiol. Pathol. Clin. Med., 47, 201 (2000).
142. S. Blank, K. Scanlon, T. Sinks, S. Lett, and H. Falk, Amer. J. Public Health, 85, 656
(1995).
143. X. Lexun, M. Lipkin, H. Newmark, and W. Jiarmin, J. National Cancer Inst., 91, 176
(1999).
144. P. Price, J. Buckley, and M. Williamson, J. Nutr., 131, 2910 (2001).
145. E. Kallay, P. Bareis, E. Bajna, S. Kriwanek, E. Bonner, S. Toyokuni, and H. Cross, Food
Chem. Toxicol., 40, 1191 (2002).
146. C. Redaelli, M. Wagner, D. Gunter-Duwe, Y. Tian, P. Stahel, L. Mazzuchelli, R. Schmid,
and M. Schilling, Kidney Intern., 61, 288 (2002).
147. S. Booth, J. Pennington, and J. Sadowski, Lipids, 31, 715 (1996).
148. S. Booth, A. Lichtenstein, M. Obrien-Morse, N. McKeown, R. Wood, E. Satzman, and
C. Gundberg, Amer. J. Clin. Nutr., 74, 783 (2001).
149. T. Koivu-Tikkanen, V. Ollilainen, and V. Piironen, J. Agric. Food Chem., 48, 6325
(2000).
150. S. Booth, K. Tucker, C. Honglei, M. Hannan, D. Gagnon, L. Cupples, P. Wilson,
J. Ordovas, E. Shaefer, B. Dawson-Hughes, and D. Kiel, Amer. J. Clin. Nutr., 71, 1201
(2000).
151. M. Valko, H. Morris, M. Mazur, P. Rapta, and R. Bilton, Biochim. Biophys. Acta General
Subjects, 1527, 161 (2001).
152. J. Kramer, F. Sauer, E. Farnsworth, D. Stevenson, and G. Rock, Lipids, 33, 1 (1998).
153. T. Kickler, W. Zinkham, A. Moser, J. Shankroff, J. Borel, and H. Moser, Biochem. Mol.
Med., 57, 125 (1996).
154. T. Green and S. Innis, Lipids, 35, 607 (2000).
155. A. Marzo, S. Curti, P. Marzo, and R. Lisciani, Bollettino Chimico Farmaceutico, 136,
651 (1997).
156. H. Reyes, J. Ribalta, I. Hernandez, M. Arrese, N. Pak, M. Wells, and R. Kirsch, Hepatol,
21, 1373 (1995).
157. A. Cicero and A. Gadd, Phytotherapy Res., 15, 277 (2001).
158. Z. Xu and J. Godber, J. Agri. Food Chem., 47, 2724 (1999).
159. J. Kim, J. Godber, J. King, and W. Prinyawiwatkul, J. Am. Oil Chem. Soc., 78, 685
(2001).
160. N. Rong, L. Ausman, and R. Nicolosi, Lipids, 32, 303 (1997).
161. T. Sunitha, R. Manorama, and C. Rukmini, Plant Foods Hum. Nutr., 51, 219 (1997).
162. M. Sugano and E. Tsuji, J. Nutr., 127, 521S (1997).
163. M. Sugano, K. Koba, and E. Tsuji, Anticancer Res., 19, 3651 (1999).
164. I. Igwe and O. Ogbobe, J. Appl. Polym. Sci., 78, 1826 (2000).
165. F. Okieimen, O. Bakare, and C. Okieimen, Indust. Crops Prod., 15, 139 (2002).
166. T. Gaginella, F. Capasso, N. Mascolo, and S. Perilli, Phytotherapy Res., 12, S128
(1998).
167. N. Mascolo, F. Barbato, L. Grumetto, F. Quaglia, and F. Capasso, Intern. J. Pharma-
conosy, 35, 364 (1997).
168. C. Vieira, S. Fetzer, S. Sauer, S. Evangelista, B. Averbeck, M. Kress, P. Reeh, R. Cirillo,
A. Lippi, C. Maggi, and S. Manzini, Naunyn-Schmiedebergs Arch. Pharmacol., 364, 87
(2001).
169. C. Vieira, S. Evangelista, R. Cirillo, A. Lippi, C. Maggi, and S. Manzini, Mediators
Inflamm., 9, 223 (2000).
170. A. Andrianaivo-Rafehivola, M. Siess, and E. Gaydou, Food Chem, Toxicol., 33, 377
(1995).
171. Anonymous, Intern. J. Toxicol., 20, Suppl 2:21 (2001).
172. M. Christensen, S. Lund, L. Simonsen, U. Hass, S. Simonsen, and C. Hoy, J. Nutr., 128,
1011 (1998).
173. M. Christensen and C. Hoy, Lipids, 32, 185 (1997).
174. H. Mu and C. Hoy, J. Lipid Res., 42, 792 (2001).
175. M. Nonaka, Y. Nagamori, K. Yamada, and M. Sugano, J. Jap. Soc. Food Sci. Technol.,
46, 638 (1999).
176. S. Terada, M. Iwami, H. Tanizawa, and T. Shimazu, J. Jap. Soc. Nutri. Food Sci., 54, 339
(2001).
177. D. Zhou, X. Xu, H. Mu, C. Hoy, and J. Adler-Nissen, J. Agric. Food Chem., 49, 5771
(2001).
178. J. Snook, S. Park, G. Wardlaw, R. Jandacek, D. Palmquist, L. Myoung-Sook, and
J. Hoover, Nutr. Res., 16, 925 (1996).
179. K. W. Miller, K. D. Lawson, D. H. Tallmadge, B. L. Madison, J. R. Okenfuss, P. Hudson,
S. Wilson, J. Thorstenson, and P. Vanderploeg, Fund. Appl. Toxicol., 24, 229 (1995).
180. W. M. Lafranconi, P. H. Long, J. E. Atkinson, A. L. Knezevich, and W. L. Wooding, Food
Chem. Toxicol., 32, 789 (1994).
181. E. Kirk, P. Sutherland, S. Wang, A. Chait, and R. LeBoeuf, J. Nutr., 128, 954 (1998).
182. M. Tikkanen and H. Adlercreutz, Biochem. Pharmacol., 60, 1 (2000).
183. C. Sirtori, Food Chem. Toxicol., 35, 428 (1997).
184. S. Kulling and M. Metzler, Food Chem. Toxicol., 35, 605 (1997).
185. C. Lamartiniere, J. Want, M. Smith-Johnson, and I. Eltoum, Toxicol Sci., 65, 228 (2002).
186. K. Mizutani, K. Ikeda, T. Nishikata, and Y. Yamori, J. Hypertension, 18, 1833 (2000).
187. O. Domon, L. McGarrity, M. Bishop, M. Yoshioka, J. Chen, and S. Morris, Mutation
Research Fundamental Molecular Mechanisms of Mutagenesis, 474, 129 (2001).
188. C. Rao, Food Chem. Toxicol., 35, 1228 (1997).
189. R. Newbold, E. Banks, B. Bullock, and W. Jefferson, Cancer Res., 61, 4325 (2001).
190. L. Strauss, S. Makela, S. Joshi, I. Huhtaniemi, and R. Santii, Mol. Cell. Endocrinol., 144,
83 (1998).
REFERENCES 559
191. L. Hilakivi-Clarke, E. Cho, I. Onojafe, M. Raygada, and R. Clarke, Oncology Reports, 6,

1089 (1999).
192. S.A. El-Sawi and M. Mohamed, Food Chem., 77, 75 (2002).
193. G. Ruberto and M. Baratta, Food Chem., 69, 167 (2000).
194. P. Crowell, Breast Cancer Res. Treat., 46, 191 (1997).
195. S. Richheimer, M. Bernart, G. King, M. Kent, and D. Bailey, J. Amer. Oil Chem. Soc., 73,
507 (1996).
196. P. Debersac, M. Vernevaut, M. Amiot, M. Suschetet, and H. Siess, Food Chem. Toxicol.,
39, 109 (2001).
197. P. Debersac, J. Heydel, M. Amiot, H. Goudonnet, Y. Artur, M. Suschetet, and M. Siess,
198. J. Whysner and G. Williams, Pharmacol. Therapeut., 71, 127 (1996).
199. F. Araujo, Food Chem. Toxicol., 34, 1021 (1996).
200. F. Paumgartten, R. De-Carvalho, C. Souza, K. Madi, and I. Chahoud, Brazilian J. Med.
Biol. Res., 31, 955 (1998).
201. V. Rolseth, R. Djurhuus, and A. Svardal, Toxicol., 170, 75 (2002).
202. A. Bysted, G. Holmer, and P. Lund, J. Amer. Oil Chem. Soc., 75, 225 (1998).
203. N. Morgado, J. Sanhueza, A. Gallequillos, A. Garrido, S. Nieto, and A. Valenzuela,
Annals Nutr. Metab., 43, 310 (1999).
204. L. Granlund, L. Larsen, E. Christiansen, and J. Pedersen, Br. J. Nutr., 84, 681 (2000).
205. N. Matthan, L. Ausman, A. Lihtenstein, and P. Jones, J. Lipid Res., 41, 834 (2000).
206. N. Matthan, K. Cianflone, A. Lichtenstein, L. Ausman, M. Jauhiainen. and P. Jones,
J. Lipid Res., 42, 1841 (2001).
207. N. Morgado, A. Galleguillos, J. Sanhueza, A. Garrido, S. Nieto, and A. Valenzuela,
Lipids, 33, 669 (1998).
208. M. Miyazake, M. Huang, S. Watanabe, T. Kobayashi, and H. Okuyama, Nutr. Res., 18,
1049 (1998).
209. C. Loi, J. Chardigny, C. Cordelet, L. Leclere, M. Genty, C. Ginies, J. Noel, and
J. Sebedio, Lipids, 35, 911 (2000).
210. L. Whigham, M. Cook, and R. Atkinson, Pharmacol. Res., 42, 503 (2000).
211. J. Palombo, A. Ganguly, B. Bistrian, and M. Menard, Cancer Lett., 177, 163 (2002).
212. R. Devery, A. Miller, and C. Stanton, Biochem. Soc. Trans., 29, 341 (2001).
213. H. Park, J. Ryu, Y. Ha, and J. Park, Br. J. Nutr., 86, 549 (2001).
214. H. Cantwell, R. Devery, M. OShea, and C. Stanton, Lipids, 34, 833 (1999).
215. P. Jones, L. Lea, and R. Pendlington, Food Chem. Toxicol., 37, 1119 (1999).
216. J. Scimeca, Food Chem. Toxicol., 36, 391 (1998).
217. A. Smedment and B. Vessby, Lipids, 36, 773 (2001).
218. P. Benito, G. Nelson, D. Kelley, G. Bartolini, P. Schmidt, and V. Simon, Lipids, 36, 221
(2001).
219. P. Benito, G. Nelson, D. Kelley, G. Bartolinit, P. Schmidt, and V. Simon, Lipids, 36, 229
(2001).
220. S. Livisay, S. Zhou, C. Ip, and E. Decker, J. Agri. Food Chem., 48, 4162 (2000).
221. L. Kohlmeier, Food Chem. Toxicol., 35, 1228 (1997).
222. W. McKelvey, S. Greenland, and R. Sandler, Epidemiol., 11, 469 (2000).

223. P. Pientinen, Food Chem. Toxicol., 35, 854 (1997).
224. N. de Roos, M. Bots, and M. Katan, Arterio. Thomb. Vascular Biol., 21, 1233 (2001).
225. N. de Roos, Trans fatty acids, HDL-cholesterol, and cardiovascular disease risk: Effects
of dietary changes on vascular reactivity, Wageningen Universiteit, Wageningen, the
Netherlands, 2001.
226. M. Craig-Schmidt, Lipids, 36, 997 (2001).
227. S. Elias and S. Innis, Amer. J. Clin. Nutr., 73, 807 (2001).
228. S. Carlson, M. Clandinin, H. Cook, E. Emken, and L. Filer, Jr., Amer. J. Clin. Nutr., 66,
715S (1997).
229. I. Wauben, H. Xing, D. McCutcheon, and P. Wainwright, J. Nutr., 131, 1568 (2001).
230. K. Skog, Food Chem. Toxicol., 40, 1197 (2002).
231. E. Muckel, H. Frandsen, and H. Glatt, Food Chem. Toxicol., 40, 1063 (2002).
232. H. Frandsen, H. Frederiksen, and J. Alexander, Food Chem. Toxicol., 40, 1125 (2002).
233. K. Turteltaub, R. Mauthe, K. Dingley, J. Vogel, C. Frantz, R. G. Garner, and N. Shen,
Mut. Res., 376, 243 (1997).
234. B. Stavric, B. Lau, T. Matula, R. Klassen, D. Lewis, and R. Downie, Food Chem.
Toxicol., 35, 199 (1997).
235. G. Keating and K. Bogen, Food Chem. Toxicol., 39, 29 (2001).
236. R. Sinha, M. Knize, C. Salmon, E. Brown, D. Rhodes, J. Felton, O. Levander, and
N. Rothman, Food Chem. Toxicol., 36, 289 (1998).
237. A. Solyakov and K. Skog, Food Chem. Toxicol., 40, 1205 (2002).
238. N. Tran, C. Salmon, M. Knize, and M. Colvin, Food Chem. Toxicol., 40, 673 (2002).
239. M. Knize, R. Sinha, N. Rothman, E. Brown, C. Salmon, O. Levander, P. Cunningham,
and J. Felton, Food Chem. Toxicol., 33, 545 (1995).
240. E. Snyderwine, N. Sadrieh, R. King, and H. Schut, Food Chem. Toxicol., 36, 1033
(1998).
241. J. Breneman, J. Briner, M. Ramsey, A. Kirector, and J. Tucker, Food Chem. Toxicol., 34,
717 (1996).
242. M. Ward, Food Chem. Toxicol., 35, 634 (1997).
243. N. Probst-Hensch, Food Chem. Toxicol., 35, 633 (1997).
244. J. Felton, M. Malfatti, M. Knize, C. Salmon, E. Hopmans, and R. Wu, Mut. Res., 376, 37
(1997).
245. M. Miyauchi, A. Nishikawa, F. Furukawa, K. Kasahara, H. Nakamura, M. Takahashi,
and M. Hirose, Cancer Letts., 142, 75 (1999).
246. M. Nagao, T. Ushijima, N. Watanabe, E. Okochi, M. Ochiai, H. Nakagama, and
T. Sugimura, Environ. Mol. Mutagen., 39, 158 (2002).
247. P. Pais, C. Salmon, M. Knize, and J. Felton, J. Agric. Food Chem., 47, 1098 (1999).
248. C. Tai, K. Lee, and B. Chen, Food Chem., 75, 309 (2001).
249. C. Chiu and B. Chen, Food Chem., 68, 267 (2000).
250. A. Brash, Lipids, 35, 947 (2000).
251. K. Ando, K. Nagata, R. Yoshida, K. Kikugawa, and M. Suzuki, Lipids, 35, 401 (2000).
252. T. Kaneko and S. Tahara, Lipids, 35, 961 (2000).
REFERENCES 561
253. M. Shimonov, I. Pinchuk, A. Bor, I. Beiel, M. Fainaru, M. Rubin, and D. Lichtenberg,

Lipids, 34, 255 (1999).
254. K. Eder, Lipids, 34, 717 (1999).
255. M. Kawakami, L. Ward, and H. Doi, Lipids, 35, 205 (2000).
256. B. Nwanguma, A. Achebe, L. Ezeanyika, and L. Eze, Food Chem. Toxicol., 37, 413
(1999).
257. S. Gorelik and J. Kanner, J. Agric. Food Chem., 49, 5945 (2001).
258. M. Lynch and C. Faustmann, J. Agric. Food Chem., 48, 600 (2000).
259. C. Faustmann, D. Liebler, T. McClure, and Q. Sun, J. Agric. Food Chem., 47, 3140
(1999).
260. S. Zhou and E. Decker, J. Agric. Food Chem., 47, 1932 (1999).
261. A. Kanazawa, T. Sawa, T. Akaike, S. Morimur, K. Kida, and H. Maeda, Biofactors, 13,
187 (2000).
262. G. Leonarduzzi, M. Arkan, H. Basaga, E. Chiarpotto, A. Sevenaian, and G. Poli, Free
Rad. Biol. Med., 28, 1370 (2000).
263. P. Eaton, D. Hearse, and M. Shattock, Cardiovascular Res., 51, 294 (2001).
264. F. Ursini, A. Zamburlinii, G. Cazzolato, M. Maiorino, G. Bon, and A. Sevanian, Free
Rad. Biol. Med., 25, 250 (1998).
265. D. Pratico, Trends Cardiovascular Med., 11, 112 (2001).
266. K. Uchida, Free Rad. Biol. Med., 28, 1685 (2000).
267. L. Wood, D. Fitzgerald, P. Gibson, D. Cooper, and M. Garg, Lipids, 35, 967 (2000).
268. R. Noding, S. Schonberg, H. Krokan, and K. Bjerve, Lipids, 33, 285 (1998).
269. C. Silwood and M. Grootveld, Lipids, 34, 741 (1999).
270. C. Bauer-Plank and L. Steenhorst-Slikkerveer, J. Amer. Oil Chem. Soc., 77, 477 (2000).
271. L. Steenhorst-Slikkerveer, A. Louter, H. Janssen, and C. Bauer-Plank, J. Amer. Oil
Chem. Soc., 77, 837 (2000).
272. F. Guardiola, R. Coedony, P. Addis, M. Rafecas, and J. Boatella, Food Chem. Toxicol.,
34, 193 (1996).
273. C. Adcox, L. Boyd, L. Oehrl, J. Allen, and G. Fenner, J. Agri. Food Chem., 49, 2090
(2001).
274. D. Vine, K. Croft, L. Beilin, and J. Mamo, Lipids, 32, 887 (1997).
275. M. Lyons and A. Brown, Lipids, 36, 701 (2001).
276. A. Brown and W. Jessup, Atherosclerosis, 142, 1 (1999).
277. J. Rong, S. Rangaswamy, L. Shen, R. Dave, Y. Chang, H. Peterson, H. Hodis, G. Chisolm,
and A. Sevanian, Arterio. Thromb. Vas. Biol., 18, 1885 (1998).
278. S. K. Peng, C. B. Taylor, P. Tham, N. T. Werthessen, and B. Mikkelson, Arch. Pathol.
Lab. Med., 102, 57 (1978).
279. W. E. Donaldson, Poultry Sci., 61, 2097 (1982).
280. R. L. Smith and B. H. Johnson, Free Rad. Biol. Med., 7, 285 (1989).
281. S. K. Peng, A. Savanian, and R. J. Morin, in S. K. Peng and R. J. Morin, eds., Biological
Effects of Cholesterol Oxides, CRC Press, Inc., Boca Raton, Florida, 1992, pp. 147166.
282. Y. OCallaghan, J. Woods, and N. OBrien, Cell Biol. Toxicol., 17, 127 (2001).
283. J. Woods and N. OBrien, Nutr. Cancer, 31, 192 (1998).
284. M. Ares, M. Porn-Ares, J. Thyberg, L. Juntti-Berggren, P. Berggren, U. Diczfalusy, B.

Kallin, I. Bjourkhem, S. Orrenius, and J. Nilsson, J. Lipid Res., 38, 2049 (1997).
285. E. Isong, E. Essien, O. Eka, and I. Umoh, Food Chem. Toxicol., 38, 997 (2000).
286. E. Isong, P. Ebong, E. Ifon, I. Umoh, and O. Eka, Plant Foods Hum. Nutr., 51, 159
(1997).
287. C. Yang, C. Kendall, D. Stamp, A. Medline, M. Archer, and W. Bruce, Nutr. Cancer, 30,
69 (1998).
288. G. Billek, Eur. J. Lipid Sci. Technol., 102, 587 (2000).
289. K. Eder, J. Anim. Physiol. Anim. Nutr., 82, 271 (1999).
290. K. Narasimhamurthy and P. Raina, Mol. Cell. Biochem., 195, 143 (1999).
291. M. Williams, W. Sutherland, M. McCormick, S. Jong, R. Walker, and G. Wilkins,
J. Amer. College Cardiol., 33, 1050 (1999).
292. J. Kamsiah, S. Tan, and I. Syed Zahir, Cutting Edge Technologies for Sustained
Competitiveness: Proceedings of the 2001 PIPOC International Palm Oil Congress,
Food Technology and Nutrition Conference, Hotel Istana, Kuala Lumpur, Malaysian
Palm Oil Board, Kuala Lumpur, 2001, p. 336.
293. K. Ammu, M. Raghunath, T. Sankar, K. Lalitha, and K. Devadasan, Nahrung Food, 44,
368 (2000).
294. J. Martin, F. Joffre, M. Siess, M. Vernevaut, P. Collenot, M. Genty, and J. Sebedio,
J. Nutr., 130, 1524 (2000).
295. W. Christie and G. Dobson, Eur. J. Lipid Sci. Technol., 102, 515 (2000).
296. T. Chiang, P. Wu, S. Liao, L. Wang, and Y. Ko, Food Chem. Toxicol., 37, 125 (1999).
297. T. Chiang, P. Wu, and Y. Ko, Environmental Res., 81, 18 (1999).
298. S. Lopez-Varela, F. Sanchez-Muniz, and C. Cuesta, Food Chem. Toxicol., 33, 181 (1995).
299. J. McSavage and S. Trevisan, Food Serv. Technol., 1, 85 (2001).
300. N. Vahcic and M. Hruskar, Food Technol. Biotechnol., 37, 107 (1999).
301. A. Romero, F. Sanchez-Muniz, and C. Cuesta, J. Sci. Food Agric., 80, 2135 (2000).
302. N. Kazerouni, R. Sinha, C. Hsu, A. Greenberg, and N. Rothman, Food Chem. Toxicol.,
39, 423 (2001).
303. B. Chen and Y. Chen, J. Agric. Food Chem., 49, 5238 (2001).
304. W. Rauter, Ernaehrung, 21, 447 (1997).
305. P. Simko, J. Chromatogr., 770, 3 (2002).
306. J. Kjallstrand and G. Petersson, Food Chem., 74, 85 (2001).
307. M. Guillen, P. Sopelana, and M. Partearroyo, J. Agri. Food Chem., 48, 5083 (2000).
308. P. Simko, P. Simon, and V. Khunova, Food Chem., 64, 157 (1999).
309. M. Espe, R. Nortvedt, O. Lie, and H. Hafsteinsson, Food Chem., 77, 41 (2002).
310. K. Kangsadalampai, C. Butryee, and K. Manoonphol, Food Chem. Toxico., 35, 213
(1997).
311. A. Hakura, J. Sonoda, Y. Tsutsui, T. Mikami, T. Imade, M. Shimada, S. Yaguchi,
M. Yamanaka, M. Tomimatsu, and K. Tsukidate, Reg. Toxicol. Pharmacol., 27, 273
(1998).
312. H. Wang, T. Chen, P. Yang, Y. Ma, C. Yu, and T. Ueng, Food Chem. Toxicol., 40, 653
(2002).
REFERENCES 563
313. P. White, Mutat. Res. Genetic Toxicol. Environ. Mutagenesis, 515, 85 (2002).
314. G. Williams, M. Iatropoulos, and J. Whysner, Food Chem. Toxicol., 37, 1027 (1999).
315. A. Botterweck, H. Verhagen, R. Goldbohm, J. Kleinjans, and P. van Den Brandt, Food
Chem. Toxcol., 38, 599 (2000).
316. S. Tamano, M. Hirose, H. Tanaka, A. Hagiwara, and T. Shirai, Food Chem. Toxicol., 36,
299 (1998).
317. H. Tryphonas, F. Lacroix, E. Lok, P. Jee, D. Clayson, S. Hayward, D. Miller, and
R. Mehta, Food Chem. Toxicol., 37, 671 (1999).
318. A. Safer and A. al-Nughamish, Histol. Histopathol., 14, 391 (1999).
319. L. Dwyer-Nield, J. Thompson, G. Peljak, M. Squier, T. Barkder, A. Parkinson, J. Cohen,
D. Dinsdale, and A. Malkinson, Toxicol., 130, 115 (1998).
320. G. Rao, K. Parthasarathi, and A. Sundararaj, Cheiron, 29, 91 (2000).
321. C. Yeh, J. Chung, H. Wu, Y. Li, Y. Lee, and C. Hung, Food Chem. Toxicol., 38, 977
(2000).
322. G. Maziero, C. Baunwart, and M. Toledo, Food Add. Contam., 18, 365 (2001).
323. S. Oikawa, K. Nishino, S. Oilawa, S. Inoue, T. Mizutani, and S. Kawanishi, Biochem.
Pharmacol., 56, 361 (1998).
324. T. Guenthener and G. Luo, Toxicology, 160, 47 (2001).
325. J. Lazutka, J. Mierauskiene, G. Slapsyte, and V. Dedonyte, Food Chem. Toxicol., 39, 485
(2001).
326. G. Hasheminejad and J. Caldwell, Food Chem. Toxicol., 32, 223 (1994).
327. V. Chan and J. Caldwell, Food Chem. Toxicol., 30, 831 (1992).
328. S. Abraham, Food Chem. Toxicol., 39, 493 (2001).
329. N. Gorelick, Mutat. Res., 326, 199 (1995).
330. P. Newberne, R. Smith, J. Doull, J. Goodman, I. Munro, P. Portoghese, B. Wagner,
C. Weil, L. Woods, T. Adams, C. Lucas, and R. Ford, Food Chem. Toxicol., 37, 789
(1999).
331. T. Liu, C. C. Chen, C. L. Chen, and C. Chi, Food Chem. Toxicol., 37, 697 (1999).
332. H. Daimon, S. Sawada, S. Asakura, and F. Sagami, Carcinogenesis, 19, 141 (1998).
333. H. Tseng, Food Science Taiwan, 25, 799 (1998).
334. M. Tawfik and A. Huyghebaert, Food Chem., 64, 451 (1999).
335. F. Arcadi, C. Costa, C. Imperatore, A. Marchese, A. Rapisarda, M. Salemi, G. Trimarchi,
and G. Costa, Food Chem. Toxicol., 36, 963 (1998).
336. R. Ppon, P. Lecavalier, R. Mueller, V. Valli, B. Procter, and I. Chu, Food Chem. Toxicol.,
35, 225 (1997).
337. J. Peteresn, E. Naamansen, and P. Nielsen, Food Add. Contam., 12, 245 (1995).
338. V. Komolprasert and A. Lawson, J. Agric. Food Chem., 45, 444 (1997).
339. J. Hotchkiss, Food Add. Contam., 14, 601 (1997).
340. M Tawfik and A. Huyghebaert, Food Add. Contam., 15, 592 (1998).
341. K. Date, K. Ohno, Y. Azuma, S. Hirano, K. Kobayashi, T. Sakurai, Y. Nobuhara, and
T. Yamada, Food Chem. Toxicol., 40, 65 (2002).
342. P. Laurienzo, M. Malinconcino, M. Volpe, D. Luongo, V. Ranieri, and M. Scoponi,
Packaging Technol. Sci., 14, 109 (2001).
343. G. Hilbert, L. Lillemardk, S. Balchen, and C. Hojskov, Chemosphere, 37, 1241 (1998).
344. M. Zabik, A. Booren, M. Zabik, R. Welch, and H. Humphrey, Food Chem., 55, 231
(1996).
345. A. Smith and S. Gangolli, Food Chem. Toxicol., 40, 767 (2002).
346. E. Elijarrat, A. Monjonell, J. Caixach, and J. Rivera, J. Agric. Food Chem., 50, 1161
(2002).
347. N. van Larebeke, L. Hens, P. Schepens, A. Covaci, J. Baeyens, K. Everaert, J. Bernheim,
R. Vlietinck, and G. de Poorter, Environ. Health Perspec., 109, 265 (2001).
348. D. Soong and Y. Ling, Chemosphere, 34, 1579 (1997).
349. J. Jacobson and S. Jacobson, Neurotoxicol., 18, 415 (1997).
350. D. Gilroy, J. Toxicol. Environ. Health, 60, 317 (2000).
351. M. de la Paz, R. Philen, I. Borda, A. de la Camara, J. Socieas, and E. Kilbourne, Food
Chem. Toxicol., 34, 251 (1996).
352. G. Diggle, Intern. J. Clin. Practice, 55, 371 (2001).
353. H. Schurz, R. Hill, M. de la Paz, R. Philen, I. Borda, S. Bailey, and L. Needham, Chem.
Res. Toxicol., 9, 1001 (1996).
354. E. Gelpi, M. de la Paz, B. Terracini, I. Abaitua, A. de la Camara, E. Kilbourne, C. Lahoz,
B. Nemery, R. Philen, L. Soldevilla, and S. Tarkowski, Environ. Health Perspect., 110,
457 (2002).
355. M. Ruiz-Mendez, M. de la Paz, J. Abian, R. Calaf, B. Blount, N. Castro-Molero,
R. Philen, and E. Gelpi, Food Chem. Toxicol., 39, 91 (2001).
356. D. Closa, E. Golch, R. Calaf, J. Abian, J. Rosello-Catafau, and E. Gelpi, Lipids, 36, 1125
(2001).
357. M. Alvarez, P. Cano, E. de Cabo, G. de Aguinaga, I. Martinez, and P. de la Paz, Rev. Clin.
Esp., 200, 305 (2000).
358. J. Arevalo, M. Perez-Gonazlez, C. Sanchez-Bernal, and M. Llanillo, Comp. Biochem.
Physiol., 120, 111 (1998).
359. S. Bell, I. Kuntze, A. Caputo, and R. Chatelain, Food Chem. Toxicol., 40, 19 (2002).
14
Quality Assurance
of Fats and Oils
Fereidoon Shahidi
1. INTRODUCTION
The quality of fats and oils is dictated by several physical and chemical parameters
that are dependent on the source of oil; geographic, climatic, and agronomic vari-
ables of growth in the case of plant oils as well as processing and storage condi-
tions. Thus, quality assurance criteria may depend partly on the type of oil under
investigation as well as on other factors that may vary depending on the intended
use and regulations that vary from country to country (1-3).
Edible oils may originate from animals both land-based and aquatic, higher
plants, and algal sources. Regardless of the source, the extraneous matters such
as large pieces of wood, metal pieces, soil, and so one should be eliminated. For
oilseeds, these are usually passed through a magnetized sieve. However, this
process does not eliminate environmental pollutants that might exist endogenously
or have been introduced into the raw material. The physical state of food lipids,
mainly their crystallinity and whether they exist in the liquid or solid form, is
dictated primarily by the degree of saturation/unsaturation of the oil. As a result,
565
566 QUALITY ASSURANCE OF FATS AND OILS
the approximate composition of the source material must be determined. For

example, solid fat content may be estimated by low-resolution nuclear magnetic
resonance (NMR) spectroscopy. Near-infrared (NIR) spectroscopy allows determi-
nation of fat content and other components of oilseeds. The color of the oil, which
is dependent on several factors, may be determined visually or with a Lovibond
tintometer or other handheld color measuring devices (4). The color may be
from carotenoids, chlorophylls, or other components. When the harvested seeds
are immature, often chlorophyll content of the resultant oil is high and this affects
the stability of products (5). Chlorophylls are photosensitizers; hence, their pre-
sence leads to enhanced photo-oxidation of the oil. Obviously, agronomic condi-
tions, season of the harvest, and many other related factors affect the quality of
the resultant oil. In this regard, soil may affect the content of certain unwanted
minerals, such as cadmium, in the seeds, but these do not usually end up in the
oil. Furthermore, as explained earlier, other contaminants may be introduced into
the raw material during harvest, processing, and transport. Thus, fats and oils and
their source material have to be tested for the presence and level of contaminants. In
addition, diseases and pests always lead to decreased quality of oils, as reflected in
their high acid values (6). In the case of animal fat, every effort should be made to
process the raw material in the fresh state or use fresh-frozen material to ensure
premium quality of the product.
In the extraction of edible oil from seeds, cleaning and subsequent conditioning
of the seeds followed by heating during or immediately after crushing are needed as
these deactivate the endogenous enzymes and help in releasing of the oil. After
expelling part the oil, the resultant leftover material may be flaked and then
extracted with hexanes. The expelled and solvent-extracted oils are then combined
and desolventized to afford crude oil.
The edible oils, after rendering or extraction from source material, may be sub-
jected to degumming, refining, bleaching, deodorization, and possibly winterization
and blending and hydrogenation and/or addition of stabilizers/antioxidants. Discus-
sion of these steps in any detail is beyond the scope of this overview. However, each
processing step carries with it many advantages and some disadvantages. To explain
these briefly, it is essential to first examine the constituents of fats and oils in a
cursory manner.
Edible oils are composed of triacylglycerols (triglycerides) as their main compo-
nents. The phospholipids are minor components that are generally removed during
the degumming process (7). The recovered phospholipids, often called lecithin,
may possibly be dietary supplements. Free fatty acids are then eliminated during
the refining process, and bleaching of the oil leads to the removal of colored mate-
rials as well as decomposition of hydroperoxides to secondary oxidation products.
The deodorization step is then designed to remove the odorous secondary oxidation
products from the oil. However, many useful minor components present in the oils
are also removed during the deodorization process. The deodorizer-distillate is
often high in the content of tocopherols and tocotrienols that can be removed,
purified, and sold as dietary supplements or used in specialty applications. The final
oil after refining, bleaching, and deodorizing (RBD) may further be subjected to
INTRODUCTION 567
winterization, a cooling process that allows the removal of more saturated fats as
well as possible blending. Some oils may also be subjected to hydrogenation to
enhance their oxidative stability. However, hydrogenation often leads to the produc-
tion of 3050% trans-fats that are a health concern because of their potential harm-
ful effect on the cardiovascular system (8). Therefore, novel formulations with
more saturated oils in the mix has become popular (9).
Among the parameters often checked or evaluated for quality assurance of edible
oils are those related to the makeup of the oil or their properties. Table 1 sum-
marizes a list of parameters usually employed to assess quality of edible fats and
oil. However, not all parameters listed may be evaluated for each oil.
In addition to parameters listed in Table 1 that dictate the quality of fats and oils,
storage and transport conditions are of considerable importance as they determine
the final quality of the oil. Obviously, of the above factors, fatty acid composition
and oxidative stability are of utmost importance, both from nutritional and sensory
quality viewpoints. In general, intake of omega-3 fatty acids in the western world is
much less than desired. Nutritionally, one would like to have a ratio of 1:21:5 for
omega-3 to omega-6 fatty acids in the diet. However, a high content of omega-3
fatty acids in edible oils is responsible for their rapid quality deterioration. Hence,
much effort has been made to eliminate the omega-3 fatty acids, mainly linolenic
acid, from vegetable oils. However, recent trends have reflected the concern about
low intake of omega-3 fatty acids and its deleterious effects.
Adulteration of fats and oils is another matter of concern, which might occur
accidentally or deliberately. Rendering of pork fat and beef tallow in the same
equipment without proper washing is an example of accidental and unintended con-
tamination/adulturation. However, often cheaper oils have been sold in place of, or
mixed with, more expressive oils. Thus, before to the recognition of health benefits
of hazelnut oil, this oil was an adulterant in olive oil (10). As mentioned earlier,
different oils have considerably different sterol compositions. Thus, sterols could
be a means of identifying adulterants because often fatty acid compositions of
the adulterant and the original oils are similar (11-13).
In addition, depending on the intended use, the quality of oil during storage and
use must be monitored. The oils may undergo hydrolytic rancidity, autoxidation,
photo-oxidation, and thermal oxidation. The latter type of oxidation is observed pri-
marily in the frying oil and causes quality deterioration that must be monitored with
different parameters such as color, viscosity, polar components and polymers,
among others (14,15). Obviously, oils that are highly unsaturated are not suitable
for frying purposes. On the contrary, autoxidation is a process that proceeds slowly
for properly stored oils. However, if the oil is kept in clear bottles, photo-oxidation
may occur, especially when photosensitizer chlorophyll is present. Thus, parameters
of interest for quality assurance of fats and oils begin at the farm gate and continue
up to the dinner table, which includes proper holding and use of oil at home after
purchase that, despite its importance, is often ignored by most consumers.
The following sections provide some further details about determination of qual-
ity of fats and oils. Other specifics may be found in several chapters in this series
and in several other publications.
TABLE 1. Quality Parameters of Fats and Oils.
Parameter Details
Fatty acid composition and distribution Percentage of total; depends on the type of material
Relative density At 20 C or 40 C relative to water at 20 C (<1)
Refractive index At 40 C
Viscosity At 20 C
Color Visual, Lovibond or Colormet
Turbidity Visual or instrumental
Solidification point, titer, solid fat
content, and cooling curve For water-insoluble fatty acids
Odor and taste Sensory evaluation
Saponification value mg KOH/g
Iodine value (IV) g iodine/100-g sample (WIJS method)
Unsaponifiable matter g/kg
Acid value (AV) mg KOH/g

Smoke, flash and fire points C
Oxidative state
Peroxide value (PV) meq oxygen/100-g sample
Thiobarbituric acid reactive
substances (TBARS) mmol/g
para-Anisidine value (p-Anv) mg/kg
TOTOX 2PV p-AnV
OSI, Rancimat and AOM value
Polar Lipids Percentage
Polymers Percentage
Volatile mater (%) At 105 C
Phosphorus mg/kg
Iron, copper, lead, arsenic mg/kg
Cadmium mg/kg
Trans-fatty acids Percentage; measured at 10 m
Cholesterol content Percentage, mainly for animal fat
Contaminants and foreign matter,
including plasticizers (%)
Carotenoids and chlorophylls mg/kg
Squalene C30H50
Sterols GC determination
Tocols HPLC determination
Synthetic antioxidants BHA, BHT, TBHQ, PG
Antifoaming agents Dimethyl polysiloxane, singly or with silicon dioxide
Metal chelators Citric acid or citrates, phosphoric acid
Crystallization inhibitor Oxystearin
Adulterants Fingerprinting using sterols or other minor components
Abbreviations: BHA, butylated hydroxyanisole; BHT, butylated hydroxytoluene; TBHQ, tert-butylhydroquinone;

and PG, propyl gallate.
2. OIL COMPOSITION
Fats and oils contain various classes of compounds (16). These componds are
primarily neutral lipids that include triacylglycerols (triglycerides) with lower
amounts of diacylglycerols (diglycerides), monoacylglycerols (monoglycerides),
OIL COMPOSITION 569
and free fatty acids. Partial acylglycerols are produced by hydrolysis of triacylgly-
cerols. Some oils such as cottonseed oil contain about 10% diacylglycerols. The
amount of free fatty acids should be less than 0.1%, preferably less than 0.05%
in freshly refined oils. In addition, polar lipids, mainly phospholipids, and to a
lesser extent, glycolipids are present. The content of phosphorus in crude oils
may reach 500 ppm, and in refined oils, which from phospholipids, the content is
generally less than 5 ppm, and may be below 2 ppm. In addition, fats and oils, in
general, contain a small amount of unsaponifiable matter, generally at 0.32.0%
mainly tocopherols, tocotienols, phytosterols, hydrocarbons (e.g., squalene and
carotenes), among others. Phenolics such as hydroxytyrosol and oleuropein might
also be present (17). Trace metals, mainly iron and copper, and other components
often exist. The content of iron and copper in freshly refined oils should be less
than 0.1 and 0.01 ppm, respectively. Crude palm oil was 0.6 ppm for iron,
6.05 ppm for copper, 0.6 ppm for magnesium, 1.2 ppm for chromium, and
2.2 ppm for nickel (18).
The triacylglycerols of fats and oils contain a range of fatty acids, and their
arrangements on the glycerol backbone may vary, depending on the source materi-
al. High-performance liquid chromatography as well as gas liquid chromatography
may be used for separation and tentative identification of individual triacylglycerols
based on their carbon number.
In neutral oils and fats, the fatty acids are not usually randomly distributed
among different positions on the glycerol backbone and are associated in particular
patterns. As an example, saturated fatty acids such as palmitic and stearic acids are
associated with the sn-1 and sn-3 positions of soybean oil, albeit at higher propor-
tions in the sn-1 position. However, the reverse is observed at high content of satu-
rated fatty acids. Linoleic acid is preferably in the sn-2 position, whereas oleic
acid is randomly distributed among the three positions. Linolenic acid is primarily
at sn-2 followed by sn-1 and sn-3 positions. The stereospecific distribution of fatty
acids has a marked effect on the oxidative stability of the resultant oils, and their
presence at the sn-2 position helps their stability (19).
The fatty acids present in fats and oils may be analyzed after their hydrolysis and
subsequent conversion by methylation to volatile methyl esters. In this Process, dif-
ferent methylating agents may be used, and these are methanol/sulfuric acid (20) or
methanol-BF3 (21). The methyl esters so produced are then identified with gas
chromatography. Standard fatty acids methyl esters are often used for tentative
identification purposes. For determination of fatty acid isomers, including trans-
fatty acids, it is necessary to use appropriate columns and conditions for analysis.
Other parameters that are indirectly related to the composition of edible oils
include iodine value and saponification value. The iodine value is a simple chemical
constant for a fat or oil. It measures unsaturated or the average number of double
bonds in fats and oils. Iodine value is defined as the number of grams of iodine that
could be added to 100 g of oil, which is measured with the AOCS Method cd 1-25
(22). Meanwhile, saponification value is a measure of the alkali-reactive groups in
fats and oils and is defined as the mg of KOH needed to saponify 1 g of oil. Shorter
chain fatty acids give higher saponification values than do longer chain fatty acids.
3. MINOR COMPONENTS
Polar lipids. Polar lipids, mainly phospholipids, are present in fats and oils, and
these originate primary as components of cell membranes and serve biological
functions in the cells. Among phospholipids present are phosphotidylcholine
(PC), phosphatidylethanolamine (PE), and phosphatidylinositol (PI). In general,
saturated fatty acids are present at the sn-1 and unsaturated fatty acids at the
sn-2 positions of phospholipid molecules.
Sphinogolipids are also important bioactive components of all membranes. Their
hydrolysis products participate in regulation of growth, differentiation, and apopto-
sis by cells. They may also participate in reducing the cancer risk in humans; colon
and skin cancers are particularly inhibited.
The content of polar lipids is reduced during oil refining. Degumming removes
most polar lipids. However, refining, bleaching, and deodorization would also bring
about a reduction in the content of polar lipids.
4. UNSAPONFIABLES MATTER
In general, unsaponifiable matters are resent in edible oils at less than 2% (23,24),
which include tocopherols/tocotrienols, other phenolics, phytosterols, hydrocar-
bons, among others. The content of these unsaponifiable matters is varied in differ-
ent oils and depending on the extent of oil refining. Although tocols and other
phenolics as well as phytosterols are removed during different stages of oil refining,
their main reduction occurs during deodorization of oils. Thus, deodorizer distil-
lates rich in tocols and sterols may be used for production of these components,
which may ultimately be used as nutraceuticals or for other food applications.
The dominance of tocopherols, namely, alpha-, beta-, gamma-, and delta- and
the corresponding tocotrienols, depends on the type of oil under investigation.
Thus, tocotrienols occur primarily in palm and rice bran oils. Meanwhile, tocopher-
ols are more widely present in different oils. However, their proportions in different
oils is dependent on the source material. As an example, sunflower oil contains
mainly alpha-tocopherol and very small amounts of other tocopherols, whereas
soybean oil contains mainly gamma-tocopherol with decreasing amounts of del-
ta-, alpha-, and beta-tocopherols as determined by high-performance liquid chroma-
tography.
Another group of unsaponifiable matter is phytosterols, fatty acid esters of phy-
tosterols, and sterol glycosides. Again, their amount is reduced during processing.
The presence of high amounts of phytosterols in soybean germ oil has been docu-
mented, which include beta-sitosterol, campesterol, stigmasterol, and 5-avenas-
terol. Phytosterols are recognized for their cholesterol-lowering properties (25).
Phytosterols are usually analyzed with gas chromatography.
The hydrocarbons present in oils are composed mainly of squalene and
carotenoids such as beta-carotene, among other carotenes. In addition, oxygenated
COLOR AND APPEARANCE 571
derivatives of carotenoids may be present. Palm oil serves as a rich source of

carotenoids at 500700 ppm.
5. CHARACTERISTICS OF FATS AND OILS
Fats and oils pass through a series at crystallization phases at cooling (26). There-
fore, when melting such crystals, the melting points of fats and oils provides an
estimate of their degree of saturation/unsaturation that parallels the saturation/unsa-
turation pattern dictated by their fatty acid constituents. Trans-fats, when present,
have a higher melting point than do their cis-counterparts because of better packing
of trans-fatty acids when compared with their cis counterparts. The melting beha-
vior and crystal structures are major factors that are important when using such pro-
ducts in different applications such as in confectionary products.
Fats and oils often show multiple melting points. As an example, tristearin has
three melting points at 52 C, 64 C and 70 C, because fats and oils solidify in more
than one crystal form; this property is known as polymorphism. Crystallization of
fats and oils occurrs in two stages of nucleation and growth.
Titer is another variable often recorded for fats and oils (16). It measures the
solidification point of the fatty acids as per AOCS Method ce 12-59 (22).
The density of liquid oils is dependent on their fatty acid composition, minor
components, and temperature. An equation taking these into account was developed
by Pantzaris (27) using iodine value, saponification value, and temperature. The
density of liquid oils is in the range of 0.9090.921 and for solid fats varies between
0.858 and 0.893. The lower values are for more solid fats such as lard and tallow. In
a similar way, the viscosity of various vegetable oils depends on their fatty acids.
Generalized methods have been developed that allow calculation of density and
viscosity of different oils. Coupland and McClements (28) and Fisher (29) have
related viscosity and density, refraction, surface tension, and other physical proper-
ties. Viscosity of fats and oils also depends on the temperature.
The refractive index of oils depends on their molecular weight, fatty acid chain
length, degree of unsaturation, and degree of conjugation. Triacylglycerols have
higher refractive indices than do their constituent free acids. Values of refractive
index for different oils generally vary between 1.447 and 1.482.
Smoke point is another characteristic that is important if oils are with for frying.
The temperature at which smoking is observed with actual frying or heating is
measured with AOCS Method Ca 9a-48 (22). Smoke point depends primarily on
the content of free fatty acids as they are more volatile than their corresponding
triacylglycerols.
6. COLOR AND APPEARANCE
Most oils are yellow-red or amber liquids. The color is from the presence of chlor-
ophylls and carotenoids. The colored bodies are often removed during the bleaching
process. Often lighter color has been associated with better quality oils, especially
for salad oils and shortenings.
The presence of chlorophylls not only renders a green color to products, but also
they act as sensitizers for fats and oils oxidation. However, unrefined olive oils con-
tain 120 ppm of chlorophylls that are considered important as extra virgin quality
indicators for this oil.
Carotenoids are present in edible oils at different levels. These are powerful anti-
oxidants against both autoxidation and photo-oxidation. Therefore, attempts have
been made to retain them or recover them, as in the case of palm oil. However, car-
otenoids may be degraded to colorless products at high temperatures exceeding
150 C.
The color of edible oils is measured by the so-called Wesson method that is
described in the AOCS Method Ce 136-45 (22) by comparison with red and yellow
Lovibond glasses of known characteristics. The oil is placed in Lovibond containers
that are 1 or 5.25 inches, and the color superimposes a mixture of red and yellow
standards to adjust to the color of the sample. Although color is three-dimensional,
the brightness factor is not considered. Yellow is needed to allow the color to look
similar, but yellow is considered unimportant in this method and only the redness is
measured. This method is the one used by the U.S. edible oil industry. The British
standard, however, uses Lovibond tintometer. The geometry and color scales for
these two methods are different, as in the tintometric method, a series of perma-
nently colored glass standards of red, yellow, and blue are used. Each standard color
is numbered. The addition of the blue color field provides a greater degree of
brightness and greenness than in the Wesson method (30).
7. OXIDATIVE QUALITY AND STABILITY TESTS
Oxidative stability of edible oils depends primarily on their fatty acid composition
and, to a lesser extent, in the stereospecific distribution of fatty acids in the triacyl-
glycerol molecules. The presence of minor components in the oils also affects
their oxidative stability. A detailed discussion of oxidative processes in fats and
oils is provided elsewhere in this series. Oxidation may occur via different routes
and includes autoxidation, photo-oxidation, thermal oxidation, and hydrolytic
processes, all of which lead to production of undesirable flavor and products harm-
ful to health. Flavor and odor defects may be detected by sensory analysis or by
chemical and instrumental methods. However, chemical and instrumental proce-
dures are often employed in the processing and during usage of edible oils. Indica-
tors of oxidation are those that measure the primary or secondary products of
oxidation as well as those from hydrolytic processes or from thermal oxidation,
including polymers and polar components (15).
Peroxide value. Peroxide value (PV) is the most common measurement of lipid
oxidation. Hydroperoxides have no flavor or odor of their own, but they are unstable
and break down rapidly to other products such as aldehydes that have a strong, dis-
agreeable flavor and odor. Peroxide value measures the miliequivalents of oxygen
POLYMERS AND POLAR COMPONENTS 573
(hydroperoxides) per gram of oil. The iodometric AOCS Method Cd 8-53 (22) is
used. PV is most widely used for determination of edible oil quality. The maximum
PV of 0.1 and preferably less than 0.05 is expected for freshly refined oils. A per-
oxide value of higher than 10 meq/kg is considered unacceptable. Conjugated
dienes and trienes absorbing at 234 and 268 nm, respectively, are directly related
to hydroperoxides and are often used in addition or in place of PV.
8. CARBONYL COMPOUNDS
Carbonyl compounds in oxidized fats and oils are the secondary oxidation products
that originate from decomposition of hydroperoxides. They usually have low
threshold values and hence are responsible for off-flavor development in oxidized
oils. Therefore, content of carbonyl compounds corresponds with sensory data.
Anisidine value. The p-anisidine value (p-AnV) measures the amount of unsatu-
rated aldehydes in fats and oils. In this method, p-anisidine reacts with aldehydes in
acetic acid to afford a yellowish color that is measured at 350 nm. The color inten-
sity depends on the amount of aldehydes as well as on their structure. The AOCS
Method Cd 18-90 (22) has been standardized for anisidine value analysis. The
Totox value, which is 2 PV p-AnV, provides information about the current status
of oxidation as well as its history and is used by the industry.
Thiobarbituric Acid Value. The 2-thiobarbituric acid (TBA) test is a popular
method for measuring sensory oxidation products. It is based on the formation of
a colored complex between two molecules of TBA reagent with one molecule of
malonaldehyde or TBA reactive substances (TBARS). This intensity of the pink
chromogram is measured at 532 nm.
Gas Chromatographic Methods. Gas chromatographic methods may be used for
measuring volatile oxidation products. Static headspace, dynamic headspace, or
direct injection methods may be employed. Specific aldehydes may be measured
as indicators for oxidative stability of oils and fats. Thus, propanal is an and as indi-
cator for stability of omega-3 fatty acids, whereas hexanal is best for following the
oxidative stability of omega-6 fatty acids.
Free Fatty Acid/Acid Value. Hydrolytic processes lead to the formation of free
fatty acids by splitting of acylglycerols that can affect flavor. The Standard AOCS
Method Ca 5a-40 and Cd 3a-63 (22) for acid value are commonplace. Free fatty
acids are normally calculated as free oleic acids on a percentage bases. Free fatty
acids are important quality indicators during processing and storage of fats and oils.
They are also found during frying of fats and oils. The amount of moisture from
foods fried and the frying temperature are important.
9. POLYMERS AND POLAR COMPONENTS
The content of polymers and polar components in oils increases during frying
process. Size exclusion chromatography and HPLC may be used for the analysis
of such components. The content of polar lipids should not exceed about 20%.
10. ANTIOXIDANTS
Antioxidants are used widely in fats and oils products to delay oxidative processes.
Synthetic antioxidants, namely, butylated hydroxyanisole (BHA), butylated hydro-
xytoluene (BHT), tert-butylhydroquinone (TBHQ), and propyl gallate (PG), are
permitted antioxidants that are frequently used in products. Their presence and
concentration may be determined with HPLC and GC methods. Meanwhile, metal
chelators such as citric acid may be determined by HPLC analysis.
11. ADULTERATION
Adulteration of fats and oils is an old problem. Many older tests involved deter-
mination of physical properties such as refractive index, melting point, and viscos-
ity. However, color tests were later used for this purpose. Thus, Baudonin reaction
for sesame oil and the Halpben test for cottonseed oil have been noted. In both
cases, a compound characteristic to an oil determines the presence of the oil.
However, today such detections and quantitations are carried out with GC and
HPLC procedures. Thus, cholesterol and phytosterols may be determined by gas
chromatography for fingerprinting purposes; however, fatty acid analysis might
also be used for higher levels of contamination (31). Detailed discussion of issues
related to oil authentication and adulteration has taken place (11).
12. POLLUTANTS
Environmental pollutants such as pesticides and herbicides may be present in fats

and oils. In this connection, special attention should also be paid to the presence of
polychlorinated biphenyls (PCB) as well as dioxin as well as polycyclic aromatic
hydrocarbon in the oil. The presence of high levels of such unwanted matters in the
oil may render them unfit for edible purposes.
REFERENCES
1. FAO/WHO. Recommended International Standard for Edible Sunflower Seed Oil, FAO
and WHO, Rome, Italy, 1970.
2. H. W. Lawson, Standards for Fats and Oils, AVI Publishing Company, Westport,
Connecticut, 1985.
3. J. B. Rossell, in J. H. P. Tyman and M. H. Gordon, eds. Developments in the Analysis of
Lipids, Royal Society of Chemistry, London, U.K., 1994, pp. 179202.
4. R. J. Hamilton, and J. Cast, eds. Spectral Properties of Lipids. CRC Press, Boca Raton,
Florida, 1999.
5. M. S. O. Gomes, P. Sinnecker, R. T. Tanaka, and U. L. Lanfer-Marquez, J. Agric. Food
Chem., 51, 16341639 (2003).
6. B. Muik, B. Lendl, A. Molina-Diaz, B. Ortega-Calderon, and M. J. Ayora-Canada, J. Agric.
Food Chem., 52, 60556060 (2004).
REFERENCES 575
7. R. Zamora, C. Olmo, J. L. Navarro, and F. J. Hidalgo, J. Agric. Food Chem., 52, 4166
4171 (2001).
8. G. L. Johnson, R. M. Machado, K. G. Freidl, M. L. Achenbach, P. J. Clark, and S. K. Reidy,
Org. Process Res. Develop., 6, 637644 (2002).
9. G. R. List, F. Orthoefer, T. Pelloso, K. Warner, and E. Neff, in Physical Properties of Fats
and Oils and Emulsifiers, N. Willak, ed., AOCS Press, Champaign, Illinois, 2000, pp.
226237.
10. E. C. Lopez-Diez, G. Bianchi, and R. Goodacre, J. Agric. Food Chem., 51, 61456150
(2003).
11. M. Jee, in M. Jee, ed., Oils and Fats Authentication, Blackwell Publishing (CRC Press),
Boca Raton, Florida 2002, pp. 124.
12. M. H. Gordon, in M. Jee, ed., Oils and Fats Authentication, Blackwell Publishing (CRC
Press), Boca Raton, Florida, 2002, pp. 143155.
13. F. Shahidi, in K. R. Cadwallader and H. Weenan, eds., Freshness and Shelf Life of Foods,
ACS Symposium Series 836, American Chemical Society, Washington, D.C., 2002, pp.
201211.
14. J. Benedito, A. Mulet, J. Velasco, and M. C. Dobarganes, J. Agric. Food Chem., 50, 4531
4536 (2002).
15. E. G. Perkins, in A. J. St. Angelo, ed., Lipid Oxidation in Food, ACS Sympsoium Series
500, American Chemical Society, Washington, D.C., 1992, pp. 310321.
16. R. D. OBrien, in Fats and Oils: Formulating and Processing for Applications, Technomic
Publishing Co., Lancaster, Pennsylvania, 1998, pp. 181250.
17. D. Bosku, in F. D. Guston, ed., Vegetable Oils in Food Technology: Composition,
Properties and Uses. CRC Press, Boca Raton, Florida, 2002, pp. 244277.
18. K. G. Berger, and R. J. Hamilton, in R. J. Hamilton, ed., Developments in Oils and Fats,
Blackie Academic & Professional, Glasgo, U.K., 1995, pp. 192203.
19. F. D. Gustone, Vegetable Oils in Food Technology: Composition, Properties and Uses.
CRC Press, Boca Raton, Florida, 2002.
20. M. A. Khan, and F. Shahidi, J. Food Lipids, 6, 331339 (2000).
21. S. M. Budge, and C. C. Parish, Lipids, 38, 781791 (2003).
22. AOCS. Official Methods and Recommended Practices of the American Oil Chemists
Society. 4th ed., AOCS Press, Champaign, Illinois, 1990.
23. F. Shahidi, and V. J. Shukla, Inform 7, 12271232 (1996).
24. F. Shahidi, in K. R. Cadwallader, and H. Weenan, Freshness and Shelf Life of Foods, ACS
Symposium Series 836, American Chemical Society, Washington, D.C., 2002, pp. 201211.
25. P. C. Dutta, Phytosterols as Functional Food Components and Nutraceuticals, Marcel
Dekker, New York, 2004.
26. N. Widlak, Physical Properties of Fats, Oils and Emulsifiers, AOCS Press, Champaign,
Illinois, 1999.
27. T. P. Pantzaris, Fat Sci. Technol, 97, 397402 (1985).
28. J. N. Coupland, and D. J. McClements, J. Amer Oil Chem. Soc., 74, 15591564 (1997).
29. C. H. Fisher, J. Amer. Oil Chem. Soc., 75, 12291232 (1998).
30. A. A. Belbin, Inform, 4, 648654 (1993).
31. J. M. N. Marikkar, H. M. Ghazali, Y. B. Che Man, T. S. G. Peiris, and O. M. Lai, Food
Chem., 90, 2330 (2005).
15
Dietary Lipids and Health
Bruce A. Watkins,1 Yong Li,1 Bernhard Hennig,2 and Michal Toborek2
1
Purdue University
West Lafayette, Indiana
2
University of Kentucky
Lexington, Kentucky
1. INTRODUCTION
Lipids support multiple biological functions in the body. They serve as the structural
building material of all membranes of cells and organelles. Lipids are the most
efficient fuel for living organisms containing more than twice the energy content
compared with carbohydrates and proteins on a weight basis. Lipids and their deri-
vatives also serve as signaling molecules that facilitate a variety of physiological
functions. In addition, lipids are recognized as important biomarkers of disease
and are involved in several pathological conditions. The cellular activities in tissues
and organs are to some extent a result of biological actions of fatty acids mediated
by changes in the membrane bilayer structure to impact the processes of membrane-
associated receptors and signal transduction systems and ion channels. Recent
literature also demonstrates a specific role of fatty acids in gene modulation and
protein expression to influence risk of chronic disease.
In contrast to the shorter chain and more saturated fatty acids, the essential fatty
acids (EFAs), linoleic acid (LA, an omega-6 fatty acid, 18:2n-6), and a-linolenic
acid (LNA, an omega-3 fatty acid, 18:3n-3) serve as substrates for the production
577
578 DIETARY LIPIDS AND HEALTH
of polyunsaturated fatty acids (PUFAs) used in cellular structures and as precursors

for the biosynthesis of many of the bodys regulatory biochemicals [glycerolipids,
long-chain (LC) PUFAs, and eicosanoids] (1). The eicosanoids are powerful, short-
lived, hormone-like compounds synthesized from specific PUFA. In addition,
formation and dietary sources of the LC-PUFA, arachidonic acid (AA, 20:4n-6),
eicosapentaenoic acid (EPA, 20:5n-3), and docosahexaenoic acid (DHA, 22:6n-3)
balance prostanoid production. An absolute requirement of DHA is necessary for
normal neural and retinal development in the infant and young child. Biochemical
and clinical studies indicate that sources of LNA and stearidonic acid (SDA, 18:4n-3)
alter the balance of their respective LC-PUFA, principally EPA. A short-term
human study indicated that vegetable oils containing SDA could be a dietary source
of n-3 fatty acids that would be more effective in increasing tissue EPA concentra-
tions than with LNA-containing seed oils (2). However, the concentration of DHA
in tissues was not affected by dietary SDA supplementation, and this should be a
concern when DHA is intended to be enhanced, for example, in infant nutrition.
This chapter introduces the contemporary understanding of food lipids in nutri-
tion and health. The health aspects of dietary lipids and the underlying cellular and
molecular mechanisms of PUFA actions are also discussed. A primary focus is
placed on the role of dietary fat in cardiovascular disease and atherosclerosis.
1.1. The Dietary Reference Intakes

New dietary reference intakes (DRIs) for healthy individuals and populations were
recently published by the Institute of Medicine of the National Academies in 2002.
Along with setting the Adequate Intake (AI) of LA levels at 17 g/day for young
men and 12 g/day for young women, the new report provides the first comprehen-
sive recommendations for n-3 PUFA in the United States (3, 4). The DRIs for n-3
PUFAs place a primary emphasis on adequate consumption of LNA to satisfy the
principle requirement for all ages and both genders. To a lesser extent, provisions
for modest recommended intakes are made for the long-chain n-3 PUFAs, EPA, and
DHAs. The daily AI for LNA are 1.6 and 1.1 g/d for adult men and women, respec-
tively. The acceptable macronutrient distribution range (AMDR) for LNA is 0.6%
to 1.2% of daily energy intake (4). Both EPA and DHA can satisfy 10% of the
AMDR (0.06% to 0.12% energy) for n-3 PUFAs, and an optimal ratio of LA/
LNA (n-6/n-3 fatty acids) is proposed to range from 5 to 10. The AMDR for total
fat is set at 20% to 35% of energy. No DRIs are set for saturated fat and trans-fatty
acids because of their perceived adverse effects on health and the tolerable upper
intake levels (ULs) were not set for these fatty acids because of practical issues. In
light of these DRIs for fatty acids, the food supply may provide a reasonable way
for all healthy individuals and at-risk groups to achieve these intakes.
1.2. General Nutrition and Health of Lipids

Lipids (fatty acids, glycerol lipids, and cholesterol) are vital nutrients that serve as
an energy source, structural components of the living organisms, and essential to
INTRODUCTION 579
biological activities of homeostasis. Beyond their traditional role as nutrients, lipids

or fatty acid molecules also facilitate critical biochemical and physiological func-
tions as modulators of cell actions and genes. The n-6 and n-3 PUFAs have been
recognized as ligands for peroxisome proliferator-activated receptors (PPARs). The
PPARs regulate gene expression in lipid and carbohydrate metabolism (5, 6), but
emerging evidence indicates that different PPARs are involved in a much broader
capacity as biological regulators (7).
Long-chain PUFAs and their derivatives, such as prostaglandins and leuko-
trienes, are PPAR activators. The n-6 and n-3 PUFAs are the most potent PPAR
ligand fatty acids. These natural ligands work on all three types of PPARs
(PPARa, PPARg, and PPARd) by binding to PPARs, however, the affinity to bind
and expression of PPARs vary depending on the type of tissue and cell. The PUFA
ligands (LA, AA, EPA, and DHA) are known agonists or antagonists of COX-2
expression through the activation of PPARs (8).
1.3. n-3 PUFA

The biological effects of dietary lipids on human health remain a primary focus of
nutrition research as consumption recommendations are continually updated in
response to new information obtained through epidemiological, clinical, and animal
investigations. The role of n-3 PUFAs (DHA) in the development of the infant ner-
vous system and retina is clearly established (9). Moreover, implications of a ther-
apeutic effect (10, 11) on reducing cardiovascular disease and cancer risk and
actions of their derivatives as biological effectors of human pathologies further
drive biochemical and molecular investigations to elucidate the health benefits of
dietary fatty acids (12). In addition to their beneficial impact on cardiovascular
pathologies and cancers, n-3 fatty acids are also known to lessen the severity and
minimize symptoms of chronic inflammatory diseases (13, 14), including rheuma-
toid arthritis (15) and inflammatory bowel disease (16), and may even benefit in
correcting psychological disorders (17).
There are three major n-3 PUFA species present in food. These are LNA in vege-
tables, oilseeds, and nuts, and EPA and DHA in cold water fishes and algae.
Another n-3 PUFA receiving greater attention recently is SDA, which is high in
some plant oils (such as hempseed oil and black currant seed oil) but can be isolated
and concentrated from marine fish. SDA may function as an important human diet-
ary component for people with deficits in 6-desaturase activity. In human and
other mammals, fatty acids of the n-3 series longer than 18 carbons cannot be
synthesized from common carbon sources. Thus, n-3 PUFAs in the human body
are either ingested directly or formed from LNA, which renders LNA as the essen-
tial fatty acid of the n-3 series PUFA. Although it is known that the human can
make LC n-3 PUFAs starting with LNA, some evidence suggests that a supplement
of preformed LC-PUFA is beneficial, especially in early infancy (formula-feeding)
and under certain metabolic disease conditions.
James et al. (2) reported recently that consuming encapsulated SDA oil (ethyl
ester) increased EPA and DPA(n-3) (22:5n-3) in healthy male and postmenopausal
female subjects (n 15/group) in a double-blind, parallel-group design study. The

results clearly showed that SDA was more efficiently converted to biologically
active 20 and 22 carbons n-3 PUFAs and implies that vegetable oils containing
SDA could be a dietary source of n-3 fatty acids that would be more effective in
increasing tissue EPA concentrations than LNA-containing oils. The introduction of
SDA-containing oils in food manufacturing could provide a wide range of dietary
alternatives for increasing tissue n-3 PUFA concentrations to fulfill the benefits
proposed for a variety of chronic health problems.
1.4. Conjugated Linoleic Acid

Conjugated linoleic acid (CLA) is the name given to describe a group of positional
and geometric fatty acid isomers of octadecadienoic acid. The CLA isomers are
reported to have antioxidant capacity, reduce carcinogen-DNA adduct formation,
induce apoptosis, modulate tissue fatty acid composition and prostaglandin E2
(PGE2) formation, and alter the expression and action of cytokines and growth fac-
tors (18). Though numerous biological actions of CLA have been reported, the most
consistent findings include anticancer effects in rodents and cancer cells, and reduc-
tion of body fat in growing animals. In some cases, the biological responses
observed from CLA isomers were influenced by the amounts of dietary n-6 and
n-3 PUFAs (1921).
CLA isomers have been recognized as effective anticarcinogenic agents for sev-
eral types of cancers. The cytotoxic effects of CLA isomers on growth of various
human and animal-derived cancer cells seem to be mediated by lowering the
expression of the gene transcription factor Bcl-2 family members that inhibit apop-
totic cell death or induce caspase-dependent apoptosis (2226). CLA also prevented
basic fibroblast growth factor-induced angiogenesis (27), a critical process for
growth and metastasis of cancers. Evidence for the anticarcinogenic effects of
CLA isomers indicates a modifying role in PPARa action (28).
Research demonstrated that CLA isomers reduce body fat in growing animals
(2931) and its actions on fat and energy metabolism may, in part, be directed
through changes in both PPARa and PPARg (32, 33). In addition, specific effects
of CLA isomers on activity and expression of enzymes associated with anabolic
pathways of lipid metabolism are reported (34). For example, CLA was observed
to decrease the mRNA level of the 9-desaturase enzyme in both liver tissue and
hepatocyte cultures (35).
CLA may also modulate immune function by diminishing the production of an
array of pro-inflammatory products in macrophages through activation of PPARg
(32) and lowering basal- and lipopolysaccharide-stimulated IL-6 and basal tumor
necrosis factor (TNF) production in rat resident peritoneal macrophages (19).
Through activation of PPARg, CLA decreased interferon-g-induced mRNA
expression of cyclooxygenase (COX)-2, inducible nitric oxide synthase (iNOS),
TNFa, and pro-inflammatory cytokines [interleukin (IL)-1b and IL-6] in RAW
macrophage cell cultures (32). Dietary CLA isomers also reduced ex vivo PGE2
production in rat bone organ cultures (20). Similar effects of CLA on PGE2
INTRODUCTION 581
production in various biological systems have been demonstrated (3638). Based

on these aforementioned actions of CLA isomers, the likely biochemical and mole-
cular targets that integrate their potential impact on biology include PPARs, COX
enzymes, and other transcription factors.
1.5. Food Fortification

Although significant strides have been directed at reducing fat content in food pro-
ducts, certain lipid ingredients and sources of fatty acids are used to enhance the
health and nutritional quality of foods. For example, CLA isomers were enriched
in both dairy and nondairy products to convey its anticancer and antiobesity effects
that were reported repeatedly in animal studies (39). Sources of n-3 PUFAs are also
added directly to infant formula to provide sufficient DHA for normal development
of the nervous system during early infancy. In the United States, DHA was
approved by the FDA in 2001 to be added into infant formula (40, 41).
Biological enrichment of n-3 PUFAs has been reported in lamb muscle (42),
chicken meat (4346), eggs (43, 47), ewe milk (4850), and pork (43). Sources
of n-3 PUFA in the above mentioned food products included fish meal, fish oil,
vegetable oils rich in n-3 PUFA (linseed oil and canola oil), algae and algal oil,
and oilseeds. Clinical trials clearly showed improvement in visual function in
infants fed LC-PUFA (DHA and AA)-enriched formula matching that of breast-
fed infants (9). In most cases, addition of LC-PUFA did not dramatically affect
the physical and sensory quality of the food products. However, enhancing the level
of LC-PUFA resulted in elevated susceptibility to lipid peroxidation both in the
food system and in the human body. To counter this problem, increased amounts
of Vitamin E were tested and found to be satisfactory in reducing lipid peroxidation
(47, 51). Several human studies indicated that n-3 PUFA fortification provides an
effective means to increase n-3 PUFA intakes to satisfy the new DRIs for this group
of health-enhancing fatty acids (52, 53).
Numerous studies have been conducted to enrich dairy products with CLA and
n-3 PUFAs by providing sources containing high amounts of LC n-3 PUFAs to the
ration of dairy cows. Feeding fish oil (200 ml and 400 ml/head/d) to dairy cows
consistently increased milk CLA yield by as much as three-fold compared with
the control group (54). Milk from multiparous Holstein cows supplemented with
2% added menhaden oil contained higher concentrations of CLA, transvaccenic
acid, and total unsaturated fatty acids (0.68 and 2.51; 1.42 and 6.28; and 30.47
and 41.71 g/100 g of fat, respectively) compared with the milk from controls
that consisted of a 50:50 ratio of forage to concentrate. Butter made from milk
enriched with CLA inherently had higher concentrations of CLA (55). Similar find-
ings were also reported by Donovan et al. (56) that by feeding lactating cows men-
haden oil at 2% dry-matter basis, the CLA content in milkfat increased 356% (to
2.2 g/ 100 g fatty acids) compared with the milkfat of control cows. Aside from
changes in CLA content, the n-3 PUFA also increased from a trace amount to
over 1 g/100 g of milk fatty acids (LNA 0.22, EPA 0.40, and DHA 0.20 g/100g)
when a diet with 3% fish oil was fed. In another study, concentrations of n-3 PUFAs
in milkfat were increased proportionally to the fishmeal amount in the diet and the
CLA concentrations were higher with the 100% fishmeal diet than with the 100%
soybean meal diet (57).
2. PUFA BIOCHEMISTRY
2.1. PUFA Formation

In human and most of the mammalian species, n-6 and n-3 series of PUFAs are
derived either from the diet or via in vivo synthesis from precursor EFAs, i.e., LA
and LNA because of the lack of enzymes needed to form these 18-carbon PUFAs
from de novo biosynthesis. The formation of LC-PUFAs occurs in two locations
within the cell. In the endoplasmic reticulum, LA or LNA is first desaturated by
a 6 desaturase (rate-limiting reaction step) and then elongated to form 20:3n-6
and 20:4n-3, followed by another desaturation step catalyzed by a 5 desaturase
to give rise to AA and EPA, respectively. The steps of elongation yield 22:4n-6
or 22:5n-3 from n-6 and n-3 EFA. The biosynthesis of 22:5n-6 and DHA is more
complicated and involves the Sprecher pathway. In the formation of 22:5n-6 and
DHA, as well as 22:4n-6 and 22:5n-3, the PUFAs are elongated in microsomes
to 24:4n-6 or 24:5n-3, then transported to peroxisomes where another double
bond is introduced by 6 desaturase. Following the desaturation step, 2 carbons
are removed (retroconversion) by b-oxidation and the resulting 22:5n-6 or DHA
are transported to various sites to fulfill their cellular functions (58, 59). Recent stu-
dies with a selective 6 desaturase inhibitor, SC-26196, further confirmed this
pathway in human cell cultures (60). Leonard et al. (61) also identified a new elon-
gation enzyme and its gene that is specific to LC-PUFA, which provided further
support that this biosynthetic pathway for LC-PUFA occurs in the human.
The primary site for PUFA formation is in the liver, although other tissues
appear to have the capacity to generate PUFA. Brain and retina are two special tis-
sues, in which cellular phospholipids accumulate large amounts of AA and DHA.
Dietary AA and DHA supplementation significantly (5070%) reduces the conver-
sion of LA and LNA in adults (62, 63). Age is also a factor that affects ones ability
to form LC-PUFA from precursors. Carnielli et al. (64) showed that 30-day-old
low-birthweight premature infants are capable of synthesizing DHA from LNA
that was contained in the formula. It appears that, in aging, there is a gradual
decline in the rate-limiting step of 6 desaturation supporting LC-PUFA formation,
and the decrease is greater in women than in men (65).
2.2. Prostanoids and Cyclooxygenases

The biosynthesis of prostanoids involves three important enzymatic reactions. In
the case of PGE2, formation begins with the release of AA from phospholipids
by phospholipase A2 (PLA2), followed by the synthesis of PGH2 by COX, and con-
version of PGH2 to specific prostanoids (e.g., PGE2) by terminal PG synthases (66).
MOLECULAR ACTIONS 583
The resulting PGE2 is then released from the cell and exerts its effects via specific
receptors (EP1, PE2, EP3, and EP4) to regulate various biological events (66).
COX plays a central role in prostanoid biosynthesis. Two main isoenzymes
of COX have been identified to mediate the initial reactions of PG biosynthesis:
COX-1 and COX-2 (67, 68). COX-1 is constitutively expressed in most tissues
and is localized predominantly in the endoplasmic reticulum. COX-2 is typically
expressed following stimulation with growth factors or cytokines and found, prin-
cipally, in the nuclear and perinuclear membranes of the cell (69). COX-1 and
COX-2 play central roles in PGE2 production. Bone cells contain both inducible
and constitutive cyclooxygenases (COX-1 and COX-2) and they are differentially
regulated (70). Two genes have been identified to be responsible for the two iso-
forms of cyclooxygenase (67, 68). The regulation of PGE2 production is predomi-
nantly through the regulation of COX-2, rather than COX-1 (71). PGE2 was also
shown to amplify its own production through stimulation of COX-2 in the same
study (71). Recently, COX-3, a variant of COX-1 that is made from the same
COX-1 gene but retains intron-1, was discovered in cerebral cortex and heart and
seems to play a role in pain and fever development (72).
Nonsteroidal anti-inflammatory drugs (NSAIDs) nonselectively inhibit the activ-
ities of both COX-1 and COX-2. Out of concern for the gastrointestinal side effects
of NSAID, scientists have discovered and developed many specific COX-2 inhibi-
tors that facilitate a safer choice for controlling inflammatory diseases (73). For
example, NS-398, a selective COX-2 inhibitor, has been shown to reduce tension
(mechanical stress)-induced PGE2 production from periodontal ligament cell
cultures (74).
Dietary fat has also been shown to regulate the expression of COX-2. Singh et al.
(75) reported that a high-fat corn oil diet (high in n-6 PUFA) may promote colon
tumorigenesis by up-regulating COX-2 expression, whereas a high-fat diet with fish
oil (high in n-3 PUFA) may exert its antitumor effect by inhibiting COX-2 expres-
sion. Moreover, dietary saturated fatty acids down-regulated COX-2 in rat liver with
alcohol-induced disease (76). The nuclear receptor PPARg has been shown to either
down-regulate (77) or up-regulate (8) the expression of COX-2, depending on the
model system. It is reasonable to speculate that dietary fat may exert a regulatory
role in the expression and function of COX-2 to affect many physiological and
pathological processes.
3. MOLECULAR ACTIONS
3.1. Signal Transduction

Signal transduction is the sequential events that are initiated when a signal molecule
binds with the cell surface receptor and ultimately causes changes in target genes.
Fatty acids or their derivatives could affect the signal transduction system at each
step of the cascade. Many signaling molecules that play critical roles in transmitting
extracellular signals are known to be acylated to facilitate cross-membrane
translocation. These lipid molecules can also be signal mediators by themselves and
even modulate the action of transcriptional factors, such as PPARs (78).
Certain PUFAs are known for their inhibitory effects on inflammation and auto-
immune responses through regulation of signal transduction pathways. Zeyda et al.
(79) showed that, when treated with PUFA, c-Jun NH2-terminal kinase, one mem-
ber of the mitogen-activated protein kinase families, is inhibited. PUFAs have also
been shown to modulate mammary cancer cell growth through regulation of the epi-
dermal, growth-factor receptor/mitogen-activated protein kinase signal transduction
cascade (80).
3.2. Transcription Factors

Transcription factors are protein molecules that convey metabolic parameters into
nuclear transcriptional regulatory events and are fundamental components of the
gene expression modulation mechanism. These transcription factors can be induced
by endogenous and exogenous chemicals via either the signal transduction path-
ways or directly by their ligands and can directly regulate gene expression in
response to these inducer molecules to affect a number of physiological functions
such as fatty acid and glucose metabolism. Study of these signal factors will facil-
itate the elucidation of the molecular mechanisms underlying the regulation of
metabolism and its possible disorders.
3.3. Peroxisome Proliferators Activated Receptors

PPARs are gene regulators belonging to the steroid/thyroid nuclear receptor super-
family. The PPARs heterodimerize with retinoid X receptors (RXRs) to form a
PPAR/RXR heterodimer prior to interacting with target genes through their
PPAR response elements (PPREs) in the target gene promoter region. Three iso-
forms of PPARs are identified, each demonstrating different tissue prevalence
and regulatory functions. PPARa is highly expressed in brown adipose tissue and
is found in great levels in liver, kidney, heart, mucosa of digestive system, and ske-
letal muscle (81). PPARa plays an important role in fatty acid oxidation in liver.
PPARg has two subtypes: PPARg1 and PPARg2. They both are expressed in adipo-
cytes while PPARg1 is also found in other tissue and cell types, such as colon, kid-
ney, macrophages, and skeletal muscle. The main function of PPARg is regulation
of adipogenesis and adipocyte differentiation as well as the control of insulin
sensitivity.
Unsaturated or saturated fatty acids, LC-PUFA and their derivatives, such as
prostaglandins and leukotrienes, are PPAR activators. The PUFAs are natural
ligands that work on all three types of PPARs by binding to them with low affinity.
The PUFA ligands (LA, AA, EPA, and DHA) are known agonists or antagonists of
COX-2 expression through the activation of PPARs (8). EPA has been shown to
increase PPARg1 gene expression in human adipocyte cultures (82). CLA has
been shown to be a potent ligand and activator of PPARa (33, 83, 84).
MOLECULAR ACTIONS 585
Among the genes that are involved in fatty acid metabolism, some are mediated
by PPARs while others are not (85). PUFAs have been shown to suppress hepatic
lipogenic gene expression through another group of transcription factors termed
sterol regulatory element-binding proteins (SREBPs) (86). LC-PUFAs were also
shown to directly modulate the transcriptional activity of the hepatocyte nuclear
factor (HNF-4a), which plays an important role in regulation of hepatocyte differ-
entiation, ureagenesis, and lipid metabolism (8789). Other nuclear factors that are
regulated by fatty acids include liver X receptors (LXRs) and retinoid X receptors
(RXRs) (90).
3.4. Polymorphisms and Single Nucleotide Polymorphisms

Gene polymorphisms are variations in DNA sequences between individuals that
happen more frequently than mutations. In the human, many chronic disease risks
are defined by dual influences of the genetic makeup and the environment. As most
human diseases, especially those that are chronic in nature, have complex inheri-
tance patterns and variable genetic traits, it is believed that the genetic component
that predetermines the risk and outcome of these diseases are sophisticated and het-
erogeneous. Therefore, with many genes contributing to disease risk, the poly-
morphisms have a modest independent and interactive influence overall to impact
the phenotype of disease. Among the genetic variations, genetic polymorphisms,
common variants as obtained from large population and family-based association
studies, are recognized in recent years to play a role in common and complex
chronic diseases. By this concept, each of the susceptible alleles only modestly
affects the risk of the disease in an individual but can account for a large proportion
of the population-attributable risk. Among the genetic variants, single nucleotide
polymorphism (SNP, a single nucleotide substitution) is the most frequently occur-
ring gene mutations in human genomes. The search for SNPs that are associated
with disease risks is currently under wide investigation in the biotechnology and
pharmaceutical industry.
By identifying genetic susceptibility factors, it is believed that prevention of or
delay in the onset of chronic disease might be possible by understanding the role of
dietary components that modify genes or SNPs. Couture et al. (91) reported that a
reduction effect of 18.5% in low-density lipoproteins (LDL)-cholesterol from a
high-carbohydrate diet in men was attributable to apolipoprotein E (apoE) poly-
morphism E2 allele. It was also shown that human subjects carrying scavenger
receptor class B, type I (SRB-I) gene exon-1 polymorphism homozygous for allele
1 are less susceptible to the presence of saturated fatty acids in the diet than subjects
that were heterozygous for allele 2 (92). Vincent et al. (93) reported in a human
dietary intervention study that several gene polymorphisms of key proteins have
been identified and linked to variable responses of diet. Some interactions found
in this study include apoE and LDL-cholesterol and triacylglycerols, apoA-IV and
LDL-cholesterol, microsomal transfer protein and LDL-cholesterol, as well as
intestinal fatty-acid-binding proteins and triacylglycerols. Rantala et al. (94)
showed that the activity of paraoxonase-1 (PON1), an enzyme that may help prevent
lipoprotein oxidation of high-density lipoproteins (HDLs), was affected by dietary

vegetable content with high levels of vegetables in the diet resulting in a lower
amount of the enzyme activity as well as reduced cholesterol levels. This response
was modulated by the genetic variance of PON1. Therefore, future research on dietary
fats and their components may demonstrate an important association between
modifying chronic disease risk via genes and SNPs by PUFA in the individual.
4. FAT AND CHRONIC DISEASES
4.1. Association of Fat and Chronic Diseases

The food we eat plays a critical role in our overall health. It is believed that a sub-
stantial amount of chronic disease risk is diet-related and could be significantly
reduced through improvements in dietary habits, e.g., up to 70% of all cancers in
the United States are attributable to diet (95). Indeed, a recent report by the USDA
indicates that poor diet and diet-related chronic diseases contribute to five of the ten
leading causes of death (heart disease, cancer, stroke, diabetes, and arteriosclerosis)
costing the United States economy an estimated $250 billion annually (96).
Dietary fat intake has remained a key focus of nutrition research in recent years
and dietary guidelines for fat consumption have been continually updated in
response to new information gained through epidemiological and clinical studies.
Links to cardiovascular disease (CVD), degenerative and inflammatory arthritis,
cancer and osteoporosis, and the recognition of fats or their derivatives as biological
effectors of human pathologies have fueled efforts to characterize the behavior of
lipids in vivo. Initially, the association of cholesterol and saturated dietary fat with
increased risk of CVD spurred dietary recommendations to reduce the intake of ani-
mal fat and to increase the intake of plant oils. However, mounting evidence is now
showing that with increased dietary intake of plant oils, such as corn, safflower, and
soybean (especially the partially hydrogenated form), which are high in LA, the
dietary ratio of n-6/n-3 fatty acids has increased significantly during the past years
(97). The high intake of n-6 with an inadequate amount of n-3 fatty acids in the diet
could contribute to the development of some cancers and other chronic diseases.
Intake of dietary sources of n-3 fatty acids is associated with reduced incidence
and severity of inflammatory disorders, cardiovascular diseases, and some cancers
in humans (12, 98103). Populations consuming fish that are rich in n-3 fatty acids
are known to have a low incidence of atherosclerotic disorders (104). Dietary fish
oil, which is high in EPA and DHA, also was shown to reduce myocardial ischemic
damage (105) and ventricular fibrillation (106). The antitumorigenic effect of n-3
fatty acids was demonstrated in breast cancer (107), colon cancer (108110), and
pancreatic neoplasm (111). In addition to their beneficial influence on cardiovascu-
lar disorders and cancers, n-3 fatty acids are also known to decrease the severity and
minimize symptoms of inflammatory diseases, including rheumatoid arthritis (15)
and inflammatory bowel disease (16), and may be of benefit in correcting psycho-
logical disorders (17).
FAT AND CHRONIC DISEASES 587
Tissue fatty acid composition correlates well with dietary intake of foods rich in
n-3 and n-6 PUFA; therefore, it is possible to examine the consequences (positive or
negative) of major changes in dietary habits between and within ethnic populations.
For example, in Japan, the intake of n-6 fatty acids has markedly increased over the
past 40 years while the level of n-3 fatty acids consumed has remained constant.
This may be explained by a decreased popularity of fish and fresh vegetables among
younger Japanese and their adoption of American dietary habits. As a result, fatty
acid intakes for this group mirror the trends observed in average Americans (112).
Currently, the average Japanese consumes over 14 g/day of LA and 2 g/day and
1.6 g/day, respectively, of LNA and EPA plus DHA (113). As a consequence, a diet-
ary increase in n-6 intake has led to an increase in the ratio of n-6/n-3 fatty acids
from 2.8 in 1955 to over 4 by 1985.
Interestingly, the incidence of breast cancer for native Japanese and Chinese is
lower compared with Japanese and Chinese who have immigrated to the United
States. Furthermore, the incidence of breast cancer in Japanese and Chinese resid-
ing in the United States, and presumably consuming a Western diet, is similar to the
incidence in American women. These results confirm the influence of environmen-
tal factors, especially diets high in fat, on the incidence of breast cancer (114). Cancer
incidence of epithelial origin in Japan is increasing, and includes lung, stomach,
colorectal, mammary, and uterine. These cancers, that are termed Western-type can-
cers, are rising with the subsequent increase in the dietary ratio of n-6/n-3 fatty
acids. Mortality rates of these cancers are approaching precedent levels based on
epidemiological data in the United States. In addition, the incidence of rheumatoid
arthritis in Japan is much less than predicted (115) in keeping with the expected
effect of dietary n-3 fatty acids on prostanoid formation (116).
The development of these chronic, Western-type diseases is associated with an
excessive formation and function of eicosanoids derived from n-6 fatty acids. As
balance can be restored to eicosanoid biosynthesis by dietary n-3 fatty acids,
an effective strategy to diminish cardio-cerebrovascular mortality (in addition to
several other serious disorders) may be to decrease the intake of n-6 fatty acids
and replace them with n-3 fatty acids (116). Such a strategy is supported by studies
that show an increased incidence of cardiovascular diseases, specifically ischemic
heart disease, in Japanese whose diet has increasingly become more Westernized
(113, 117).
4.2. Mode of Action of n-3 PUFA Against Chronic Diseases

To advance understanding of the dynamic influence of dietary lipids, research
efforts are focusing on the importance of the balance between n-6 and n-3 fatty
acids in the human diet. What is emerging is recognition that these PUFAs modu-
late eicosanoid biosynthesis in numerous tissues and cell types, alter signal trans-
duction, and influence gene expression (87, 118). The effect of n-6 and n-3 PUFA
on CVD, cancer and bone/joint health is related to the newer discoveries of how
dietary PUFA impact health.
In the human and other mammals, eicosanoids derivatives of 20 carbon n-3 or

n-6 PUFA, such as prostaglandins, leukotrienes, and thromboxanes, act and lead to
the activation of various signaling mechanisms that have effects on numerous cel-
lular functions that may be either beneficial or detrimental. Thus, modulation of
PUFA formation and eicosanoid biosynthesis could be a feasible way to reduce
the risks of certain chronic illnesses like arthritis, diabetes, inflammation, cancer,
and cardiovascular disease (119). Therefore, although numerous mechanisms are
involved in the biological activities of the n-3 fatty acids, a unifying attribute is
the down-regulating effect on eicosanoid production from n-6 fatty acid precursors.
Eicosanoid biosynthesis requires activation of phospholipases and the availability
of the free fatty acid substrate, AA and EPA. The primary eicosanoid precursor
is AA. Unesterified AA is transformed to eicosanoids by cyclooxygenase (two iso-
forms, COX-1 and COX-2) and lipoxygenase enzymes. Several lines of evidence
indicate that the amount of eicosanoids produced from AA correlates with the
severity of inflammation in rheumatic diseases (120).
Eicosanoids derived from the n-3 PUFAs are up to 100-fold less biologically
potent for inducing pro-inflammatory cellular responses than those derived from
AA (121, 122). It is proposed that dietary n-3 PUFAs act by two different mechan-
isms to control the amount of n-6 eicosanoids maintained in tissues: first, by com-
peting for incorporation into tissue lipid esters, thus reducing the rate of tissue
formation of active n-6 eicosanoids; and second, by forming weaker n-3 eicosa-
noids (123, 124) that compete at cellular receptor sites and diminish signaling by
eicosanoids derived from n-6 PUFAs. As a result, diets with n-3 PUFAs create con-
ditions that reduce the formation and constrain the function of active n-6 eicosa-
noids in stimulated cells (116).
Dietary supplementation of n-3 PUFA-rich fish oil significantly decreased the
ability of bone tissue to produce PGE2 in animals compared with those given
high-n-6 fatty acid diets (20, 125). It is presumed that prostaglandin E3 (PGE3),
an eicosanoid derivative of EPA, could have increased in these animals given the
high-n-3 diet. Although PGE3 is as potent as PGE2 in mediating bone resorption,
EPA is a less effective substrate for cyclooxygenase than AA (126). Hence, supple-
mentation with oils high in EPA typically reduces the 2-series PG derived from AA
with a small increase in 3-series PG derived from EPA (127). In addition, PGE3 is
less inflammatory than PGE2 (128), which partly explains why n-3 PUFA is bene-
ficial for controlling inflammatory bone/joint diseases.
The modulatory effect of n-3 PUFAs on eicosanoid production could also be
achieved at the enzyme level. The action of n-3 fatty acids that decreases the pro-
duction of PGE2 could be an effect of down-regulation of COX-2 activity in local
tissues (129, 130). In a study with rats, dietary n-6 PUFA up-regulated COX-2 and,
to some extent, COX-1 expression leading to a concomitant increase in COX
enzyme activity and prostaglandin synthesis, but fats containing added menhaden
oil (high in n-3 PUFAs) had an opposite effect (131).
Experiments by Curtis et al. (129) gave evidence that the observed effects
of n-3 fatty acids were not always mediated by the alterations in eicosanoid meta-
bolism. In that study, supplementation of IL-1-stimulated cultured chondrocytes
ROLE OF DIETARY FAT IN CARDIOVASCULAR DISEASE 589
with LNA resulted in gene suppression of IL-1, TNFa, and COX-2, and one of the
cartilage-degrading enzymes, aggrecanase. The supplemented LNA was not con-
verted to the longer chain n-3 fatty acids, i.e., EPA and DHA, however, supplemen-
tation of chondrocytes with EPA has produced effects similar to those enriched
with LNA.
The LC-PUFAs are believed to possess the ability to covalently attach to a vari-
ety of proteins, thus dramatically affecting translocation, cell-to-cell signaling, and
protein function (118). Dietary fatty acids may regulate cell signaling pathways via
cell surface receptors, proximal/accessory components of receptor-mediated path-
ways, at intermediate signaling steps, and via nuclear receptors. They act as cellular
second messengers and modulators during cellular transduction of external signals.
They also modify the activities of enzymatic processes, such as those catalyzed by
phospholipases, protein kinases, G-proteins, adenylate and guanylate cyclases, as
well as ion channels and other biochemical events involved in stimulus-response
coupling mechanisms (132).
The n-3 PUFA may also alter gene expression via direct interaction with proteins
involved in gene transcription (87). Investigations suggest that a variety of PUFA
(EPA, DHA, etc.) influence transcriptional regulatory mechanisms including the
previously described PPARs, and nuclear factor kappa B (NF-kB), the SREBP,
and a PUFA response element (133). Camandola et al. (134) showed that AA sup-
plementation of human monocytes strongly stimulated nuclear translocation of
NF-kB, a transcription factor that is believed to regulate the expression of gene
products involved in inflammatory reactions. Activation of NF-kB could also be
achieved by stimulation with PGE2, but not with the n-3 PUFA, EPA. Therefore,
a low dietary ratio of n-6/n-3 fatty acids may down-regulate NF-kB-mediated
gene expression of pro-inflammatory mediators. The PUFAs could also modify
the expression of oncogene in vivo. Rats given safflower oil (high in LA) had
elevated levels of p21ras protein compared with those offered menhaden oil (high
in n-3 PUFA). The Ha-ras mRNA was also increased in rats given a diet high in n-6
PUFA and high in fat content (21% of diet) compared with those given a high-fat
diet rich in n-3 PUFA, implying a tumor promoting potential of n-6 fatty acids
(131).
5. ROLE OF DIETARY FAT IN CARDIOVASCULAR DISEASE

AND ATHEROSCLEROSIS
5.1. Introduction
Even though the mortality from coronary heart disease has declined recently, athero-
sclerosis and related vascular disorders still are the leading cause of death in the
Western world. The etiology of this disease is multifactorial, with hyperlipidemia,
smoking, diabetes mellitus, hypertension, and obesity being well-established risk
factors for the development of atherosclerosis. Dietary fat affects plasma lipids,
lipoproteins, and vascular inflammation and, thus, is linked to atherosclerosis.
Injury to or abnormal mechanisms of the vascular endothelium may be initiating

events in the etiology of atherosclerosis.
Although epidemiological studies suggest that dietary cholesterol and saturated
fatty acids increase serum cholesterol, recent evidence suggests that high intakes of
polyunsaturated fats, and, especially, fats high in n-6 PUFAs, may be equally
atherogenic because of their ability to convert easily to cytotoxic lipid peroxidation
products, thus contributing to a cellular imbalance in oxidative stress antioxidant
status and to an inflammatory response. Most of the data available in the literature
are from studies where high-fat diets were used. Conversely, low-fat diets, indepen-
dent of the fat source, may be the prudent choice in prevention and treatment of
atherosclerosis. Low-fat diets usually are high in dietary fiber, antioxidants, and
other undefined materials, all of which may protect against atherosclerosis.
5.2. Pathogenesis of Atherosclerosis

Theories: There are numerous theories for the pathogenesis of atherosclerosis.
Despite considerable research, the etiology of this disease is not well understood.
The current trend is to consider atherosclerosis as a response of the vascular wall to
a variety of initiating agents and multiple pathogenic mechanisms (e.g., hyperlipi-
demia) contributing to the development of atheromatous plaques. In fact, recent
research suggests that atherosclerosis is a chronic inflammatory disorder accompa-
nied by risk factors such as oxidized LDL, reactive oxygen species, diabetes, and
infection. It appears that the major participants in the atherosclerotic disease pro-
cess include an active vascular endothelium, smooth muscle cells, blood-borne cells
such as monocytes and macrophages, and circulating lipoproteins. The result is a
multifactorial sequence of events involving endothelial cell injury/dysfunction,
uptake of circulating blood monocytes and their differentiation into macrophages,
coupled with smooth muscle cell migration and proliferation.
The most intensely studied current hypothesis of atherosclerosis is the response
to injury hypothesis (135). The hypothesis takes into account the cellular interac-
tions that occur during the different phases of lesion initiation, development, and
progression. The initiating event appears to be injury to or dysfunction of the
endothelium via lipids or lipoprotein derivatives or via mechanical, chemical, toxic,
viral, or immunologic agents. These events may induce growth-factor secretion
and changes in endothelial cell surface adhesive glycoproteins. Monocytes are
attracted and attach to endothelial cells that will contribute directly or indirectly
to continued secretion of growth factors and other biologically active molecules.
Subsequent subendothelial migration of monocytes may lead to fatty-streak formation
and release of cytokines and growth factors. Monocytes and macrophages are major
cellular components of lesions and, thus, are likely to play a role in their initiation
and evolution. These events provide three possible sources of cytokines and growth
factors, namely from platelets, macrophages, and the endothelium. Some of the
smooth muscle cells in the proliferative lesion themselves may form and secrete
cytokines and growth factors, such as platelet-derived growth factor. It is not clear
what role dietary fat plays in the above stated events. However, hyperlipidemia, or
some component(s) of hyperlipidemic serum, as well as other risk factors, are

thought to cause endothelial injury/dysfunction, resulting in endothelial cell activa-
tion, adhesion of platelets or monocytes, increase in cytokine activity, and transmi-
gration of monocytes into the arterial intima. Once in the subendothelial space,
monocytes transform into macrophages, take up substantial amounts of lipid, and
become foam cells. These foamy macrophages, as well as other cells, also can pro-
duce cytokines and growth factors that cause migration of smooth muscle cells
from the media into the intima. These interactions then lead to fibrous plaque for-
mation and further lesion progression. In other words, once smooth muscle cells
proliferate in the intima, there is further lipid accumulation as well as elaboration
of the extracellular components of the atheromatous plaque.
Even though numerous risk factors, including hyperlipidemia, smoking, and
hypertension, seem to contribute to the development of atherosclerosis, to date it
has not been possible to link these risk factors into a common pathogenic mechan-
ism. There is evidence, however, that modulations in the level of activity of a select
set of endothelial transcription factors (e.g., endothelial NF-kB) may provide a
mechanism for linking these seemingly diverse processes with the generation of
dysfunctional endothelium and the onset of atherosclerotic lesion formation (136,
137). Stimuli known to activate the NF-kB complex include inflammatory cyto-
kines, with the common denominator apparently being reactive oxygen species.
One may speculate that oxidized lipids, when present in inappropriate levels,
may induce endothelial oxidative stress and generate excess reactive oxygen spe-
cies, which activate NF-kB and modulate endothelial gene expression. Antioxidants
and related compounds may protect against atherosclerosis by inhibiting the activa-
tion of endothelial transcription factors such as NF-kB.
Dietary Fat: There is ample evidence demonstrating that serum cholesterol is a pre-
dictor of atherosclerosis and that serum cholesterol concentrations can be modified
by varying the composition of dietary fat. Keys et al. (138), in 1957, and Hegsted
(139), in 1965, provided the first quantitative estimates of the relative effects of the
various classes of fatty acids on serum cholesterol concentrations. Both studies
indicated that saturated fatty acids increased, whereas PUFA decreased, serum cho-
lesterol. Also, monounsaturated fatty acids had no specific effect on cholesterol
concentrations. These conclusions were based on combining results from numerous
studies and then describing mathematically the relationship between dietary fatty
acid composition and serum total cholesterol concentrations. However, it is now
known that saturated fatty acids are not equally hypercholesterolemic. For example,
stearic acid (18:0) and saturated fatty acids with less than 12 carbon atoms seem to
have little or no effect on raising serum cholesterol levels. This suggests that the
cholesterol-raising properties of saturated fatty acids should be attributed solely
to lauric (12:0), myristic (14:0), and palmitic acid (16:0). However, these three satu-
rated fatty acids appear to have different effects on serum total-cholesterol concen-
trations as well. Numerous studies suggest that lauric acid is less, and myristic acid
probably more, hypercholesterolemic than palmitic acid (139, 140). Furthermore,
human studies suggest that lauric acid raises total serum cholesterol and LDL cho-
lesterol concentrations compared with oleic acid (18:1n-9). However, it is not as
potent in increasing cholesterol concentrations as is palmitic acid. Although in

normocholesterolemic men and women, dietary palmitic and oleic acids seemed
to exert similar effects on serum cholesterol and lipoprotein profiles. Much still
needs to be learned about the effects of dietary fat on serum cholesterol levels
and metabolism (141143). It is clear from these and other studies that extremes
in types of dietary fat should be avoided and that moderation in dietary fat is
advisable.
Even though serum cholesterol appears to be a risk factor for atherosclerosis,
i.e., each 1% rise in serum cholesterol is predicted to increase the risk of coronary
disease by about 2%, the effect of dietary cholesterol on serum cholesterol concen-
tration is not clear and far from being understood. It appears, however, that the aver-
age baseline consumption of cholesterol-containing foods can modulate the
magnitude of a mathematically predicted change in serum cholesterol due to
changes in dietary cholesterol (144). An increase in dietary cholesterol is expected
to have the greatest effect on serum cholesterol level when the past baseline amount
of dietary cholesterol was near zero. However, if the baseline cholesterol consump-
tion is greater than 500 mg/d, additional dietary cholesterol will contribute to little
measurable change in serum cholesterol. This suggests that people who desire
maximal reduction of serum cholesterol by dietary means may have to reduce their
dietary cholesterol to minimal levels in order to observe significant serum
cholesterol reductions. Furthermore, the responsiveness to dietary cholesterol can
be extremely variable, with some individuals being much more responsive
(hyperresponders) than others. Thus, the need to limit cholesterol intake should
apply more strictly to diet-sensitive hypercholesterolemic individuals rather than
to the population in general. Individual variations in the response to dietary choles-
terol may be mediated by differences in fat absorption efficiency, neutral sterol
excretion, conversion of hepatic cholesterol to bile acids, or modulation of key
enzymes involved in intracellular cholesterol metabolism, such as HMG-CoA
reductase.
Although regression analysis of numerous human studies suggests that cholester-
ol and saturated fatty acid intake are primary determinants of serum cholesterol, the
role of dietary fat in the development of atherosclerosis remains controversial and
not well understood. The question arises whether or not dietary saturated fats
should be replaced by unsaturated fats. Unsaturated fats, especially n-3
fatty acids, may be beneficial to human health (145, 146). Some populations,
such as the Greenland Eskimos who consume high levels of n-3 fatty acids from
fish and sea mammals, have lower incidence of coronary heart disease (147). In
patients with hyperlipidemia, n-3 fatty acids only at high doses result in a decrease
of LDL cholesterol. However, these fatty acids consistently lower serum triacylgly-
cerols in normal subjects and in patients with hypertriglyceridemia. There is clear
evidence that hypertriglyceridemia is an independent risk factor of cardiovascular
diseases such as atherosclerosis (148, 149). Increasing the intake of foods rich in
n-3 PUFA may be most critical in decreasing the risk of cardiovascular diseases,
especially in light of a recent human-derived study, suggesting that consumption
of total 18:3n-3 is inversely related to plasma triacylglycerol concentrations (150).
Diets high in n-6 and n-3 fatty acids may lead to a decrease in serum cholesterol
but replacing saturated with unsaturated lipids may not be desirable because of their
tendency to be oxidized. As mentioned earlier, mounting evidence is now showing
that with increased dietary intake of plant oils, such as corn, safflower, and soybean
oil, which are high in LA, the dietary ratio of n-6/n-3 fatty acids has increased
significantly during the past years (97). The high intake of LA-rich fats will lead
to serum hypertriglyceridemia and an increase in cellular oxidative stress. There
is evidence that most age-related diseases are initiated by elevated cellular oxidative
stress or an imbalance in the bodys oxidative stress/ antioxidant status, as well a
state of chronic low-level inflammation.
Dietary antioxidants, such as Vitamin E, might act as antiatherogenic agents by
suppressing oxidative modification of LDL and the recruitment of monocytes into
the arterial subendothelium by smooth muscle cells (151). In fact, data from sub-
jects with varying degrees of coronary atherosclerosis support the hypothesis that
high-serum PUFA levels, when insufficiently protected by antioxidants, may indi-
cate a higher risk of atherosclerosis (152). In particular, a positive relationship
between LA intake and coronary artery disease has been observed in animal and
human studies (153, 154). In fact, the potential detrimental health effect of high
intakes of LA has been termed the linoleic acid paradox (155), in which a sup-
posedly healthy fatty acid (i.e., one that lowers total cholesterol) is associated with
increasing rates of cancer and inflammatory and cardiovascular diseases. Moreover,
a low intake of LNA and other n-3 (fish) oils may further compound this paradox.
All these studies lead one to conclude that the type of fat becomes a less significant
component in the pathogenesis of atherosclerosis, when one consumes a low-fat
diet, rich in soluble fibers and natural antioxidants.
5.3. Lipoprotein Metabolism

Plasma lipoproteins are units of complex lipid and protein compositions. Lipopro-
teins function primarily as carriers of lipids in the blood. The apoprotein fractions
of the different lipoproteins play an important role in the regulation of the metabolic
fate of the different plasma lipoproteins via their role as enzymatic cofactors and
their interactions with specific receptors in cell membranes. Lipoproteins can be
separated by density into four major different classes: chylomicrons, very low-
density lipoproteins (VLDLs), LDLs, and HDLs. Both chylomicrons and VLDLs are
triacylglycerol-rich particles. Chylomicrons are of mucosal cell origin and function
mainly as carriers of lipids of exogenous dietary origin to the liver and peripheral
tissues. VLDLs, on the other hand, are of hepatic origin and transport endogenous
lipids. LDLs are generated primarily by the metabolism of VLDL and are high in
cholesterol and cholesteryl esters. HDLs are the smallest particles and contain the
highest relative amount of protein. These units are relatively rich in cholesterol and
phospholipids. High levels of LDL are associated with atherosclerosis and coronary
heart disease, whereas high levels of HDL provide protection from these diseases.
In fact, HDL particles appear to play an important role in the efflux of cholesterol
from the extrahepatic tissues.
5.4. Dietary Fat and Lipoprotein Metabolism

There is substantial evidence that indicates that dietary fat can influence signifi-
cantly not only serum levels of cholesterol and triacylglycerols but also the lipid
composition and content of lipoproteins (156159). Much attention has been placed
on the effects of diet on LDL levels, and saturated fatty acid and cholesterol itself
have been identified as the major nutritional factors that can raise serum LDL-
cholesterol levels. However, LDL cholesterol is only one of the many risk factors
for atherosclerosis, and it is not known if oxidative modification of LDL is an equally
or more important factor in the pathogenesis of atherosclerosis than total LDL
cholesterol per se. More longitudinal studies are needed to answer these questions.
If lipid peroxidation is a major risk factor for atherosclerosis, then excess consump-
tion of highly unsaturated fats may not be advisable.
The quantitative relationship between cholesterol intake and cholesterol levels is
still controversial, especially because in humans, there appears to be a high indivi-
dual variability in processing of dietary cholesterol. However, numerous animal and
human studies support the concept that dietary cholesterol can raise LDL-cholesterol
levels and change the size and composition of these particles as well. LDL
particles become larger in size and enriched in cholesterol esters. Mechanisms con-
tributing to these events include an increase in hepatic synthesis of apoB-containing
lipoproteins, increased conversion of VLDL remnants to LDL, or a decrease in the
fractional catabolic rate for LDL. Reduced LDL receptor activity due to an increase
in hepatic cholesterol content, secondary to excess dietary cholesterol, may lead to
a decreased uptake of both LDL and VLDL remnants.
In addition to dietary cholesterol, saturated fatty acids also are thought to raise
serum LDL-cholesterol levels as well as total cholesterol concentrations. The major
effect of saturated fatty acids on serum cholesterol appears to be due to a reduction
in LDL-receptor activity. It is likely that saturated fatty acids may contribute to a
cellular redistribution of cholesterol and cholesterol oxidation derivatives, leading
to a favorable environment by these lipid particles to suppress LDL-receptor synth-
esis. In addition to their effects (directly or indirectly) on LDL-receptor activity,
saturated fatty acids also may promote the synthesis of apoB-containing lipopro-
teins.
It is now known that not all saturated fatty acids are equally hypercholesterole-
mic. For example, medium-chain saturated fatty acids of carbon length 810, as
well as stearic acid (18:0), have little or no effect on serum cholesterol concentra-
tions. In contrast, evidence indicates that palmitic acid (16:0), the principle fatty
acid in most diets, can increase serum cholesterol concentrations in humans. How-
ever, in normocholesterolemic humans, dietary palmitic and oleic acids have been
shown to exert similar effects on serum cholesterol, suggesting that only humans or
animal species sensitive to dietary cholesterol and selected fats (hyperrespon-
ders) may exhibit significant changes in serum cholesterol in response to dietary
fat intake. Myristic acid (14:0) and, to a lesser extent, lauric acid (12:0), which are
relatively high in coconut oil, both can raise serum cholesterol and LDL-cholesterol
levels. Overall, it is not clear why humans respond so differently to cholesterol or
saturated fatty acids. Variations may exist at the level of fatty acid catabolism and
regulation of LDL-receptor activity.
In contrast to saturated fatty acids, unsaturated fatty acids may not be cholesterol-
emic. However, because of their ability to become oxidized and thus to contribute
to oxidative stress within a cell, some unsaturated fats could indirectly be highly
atherogenic. Oleic acid, the major monounsaturated fatty acid in the diet, often is
called a neutral fatty acid because it has a neutral or cholesterol-lowering effect
on serum cholesterol. The main classes of unsaturated fatty acids in the diet can be
divided into n-6 and n-3 PUFA. LA is the predominant n-6 fatty acid, and the parent
n-3 fatty acid is LNA. Both occur in plant oils. Fish oils contain large amounts of
LC-PUFA, e.g., 20:5n-3, which have their origins of plant sources. With regard to
cholesterol metabolism, LA may lower serum cholesterol levels by up-regulating
LDL-receptor activity or by inhibiting hepatic synthesis of apoB-containing lipo-
proteins. LC n-3 PUFA appear to have a greater influence on triacylglycerol than
on cholesterol metabolism. High intake of fish-oil-derived n-3 fatty acids reduces
triacylglycerol levels, especially when fed to individuals with hypertriglyceridemia.
The role of dietary fats on HDL metabolism is not as well understood. In
general, saturated fatty acids do not reduce HDL cholesterol, and dietary monoun-
saturated fatty acids, when substituted for saturated fatty acids, contribute to a
favorable modification of the lipoprotein ratios, i.e., a decrease in the ratio of
LDL/HDL. In contrast to monounsaturated fatty acids, a high intake of n-6
PUFA (e.g., LA) reduces HDL cholesterol concentrations, possibly by reducing
the synthesis of apoA-I, a major HDL apoprotein. The actions of n-3 fatty acids
on HDL-cholesterol levels are similar to those of LA. Even though unsaturated fatty
acids do not appear to be hypercholesterolemic in general, their HDL cholesterol-
lowering capacity might be of concern, as HDL is directly protective against ather-
osclerosis. Decreased serum HDL levels would indicate reduced removal of lipids
from the arterial wall.
5.5. Lipids and Endothelial Cell Dysfunction

As the knowledge of the pathogenesis of atherosclerosis rapidly increases, it
appears that an active vascular endothelium, smooth muscle cells, and blood-borne
cells such as monocytes and macrophages all play active roles in the atherosclerotic
disease process. Risk factors, such as elevated plasma levels of certain lipids,
prooxidants, and cytokines, may contribute to the chronic activation/stimulation
as well as to the damage of the endothelium and other vascular tissues (160). There
is evidence that supports the hypothesis that it is not only pure cholesterol and satu-
rated fats but rather oxidation products of cholesterol and unsaturated fats (and pos-
sibly certain pure unsaturated fats) that are atherogenic, possibly by causing
endothelial cell injury/dysfunction. Lipid-mediated endothelial cell dysfunction
may lead to adhesion of monocytes, increased permeability of the endothelium to
macromolecules, i.e., a decrease in endothelial barrier function, and disturbances in
growth control of the vessel wall.
5.6. Fatty Acids

Although many mechanisms for the etiology of atherosclerosis have been proposed,
endothelial injury/dysfunction clearly plays a role in the atherosclerotic disease pro-
cess. There is evidence suggesting that diet-derived lipids metabolically interact
with the vascular endothelium and may be responsible for abnormal regulatory
mechanisms and a subsequent alteration of endothelial integrity. For example,
high levels of circulating triacylglycerol-rich lipoproteins (chylomicrons and
VLDLs) have been implicated in the injury process of the endothelium. Plasma
chylomicron levels are elevated in humans after consuming a high-fat meal, and
hepatic synthesis of VLDL is increased when caloric intake is in excess of body
needs. When plasma triacylglycerol-rich lipoproteins are elevated, hydrolysis of
triacylglycerols by lipoprotein lipase, and, thus, elevated concentrations of fatty
acid anions, occurs in proximity to the endothelial surface. Such high levels of
diet-derived fatty acids can cause endothelial injury or dysfunction and, thus,
disrupt the ability of the endothelium to function as a selective barrier. This would
result in lipid deposition by allowing increased penetration of cholesteryl ester-rich
remnant lipoproteins into the arterial wall. Thus, when exposing cultured endothe-
lial cells to selected fatty acids, LA (the parent n-6 fatty acid) most markedly acti-
vated endothelial cells and disrupted endothelial barrier function (161).
Furthermore, the disruption in endothelial barrier function was exacerbated greatly
in the presence of small amounts of oxidation derivatives of unsaturated fatty acids.
This suggests that, in general, fatty acid oxidation derivatives, but not pure lipids,
are extremely cytotoxic.
It is not clear why LA and none of the saturated fatty acids that were studied
disrupted endothelial barrier function. The injurious effects of LA on cultured
endothelial cells may be mediated, in part, by the induction of peroxisomes and,
thus, by excessive hydrogen peroxide formation. In addition, enrichment of
endothelial lipids with selective fatty acids can modify specific cellular lipid pools
and alter the morphology of cultured cell monolayers. Such fatty acid-mediated
compositional changes may be sufficient to alter membrane properties, e.g., fluidity
and activities of membrane-bound enzymes. One may speculate from these and
other data that high dietary intakes of certain unsaturated fatty acids, such as
LA, might not be entirely safe.
5.7. Cholesterol
There is experimental evidence that suggests that some oxysterols, but not pure
cholesterol, are the prime cause of atherosclerotic lesion formation (162). Upon
cholesterol feeding, a strong relationship was seen between plasma oxysterols
and aortic wall oxysterols. One may speculate that the deposition of pure lipids,
such as cholesterol and its esters, may be merely a secondary process in response
to oxysterol-induced endothelial cell injury. Cell injury/dysfunction and the sub-
sequent disruption of endothelial barrier function by oxysterols (163, 164) could
initiate the early events in atherosclerosis. Such injury could allow increased uptake
of cholesterol-rich lipoproteins into the arterial wall by decreasing endothelial cell

prostacyclin production, thereby enhancing platelet adhesion and aggregation and
by increasing monocyte adhesion and infiltration. These studies suggest that oxy-
sterols, and not pure cholesterol, cause dysfunction of vascular endothelial cells.
Not all oxysterols, however, are equally cytotoxic, and different mechanisms of
endothelial cell injury by different types of cholesterol oxidation derivatives may
exist. As relatively high concentrations of oxysterols are formed in certain pro-
cessed foods, and because they are easily absorbed and transported in the blood,
cholesterol oxidation derivatives may be an important dietary risk factor in cardio-
vascular disease.
5.8. Lipids, C-Reactive Protein, Homocysteine, and PPARs

In addition to various plasma lipids and lipoprotein species, other markers of car-
diovascular disease and atherosclerosis are now considered as potent screening
tools to predict these diseases. These markers are mostly related to the discoveries
that a low-level chronic inflammation and related disorders of the immune system
are as much, and probably more, of a clinical predictor of cardiovascular pathology
than dietary fat and associated lipoproteins (165).
C-reactive protein: Evidence suggests that C-reactive protein (CRP) is a better pre-
dictor of the risk of cardiovascular events than LDL cholesterol. Such significant
clinical data are supported by laboratory and experimental evidence that demon-
strate that atherosclerosis, in addition to being a disease of lipid accumulation,
also represents a chronic inflammatory process. CRP is a hepatically derived pen-
traxin that plays a key role in the innate immune response. The proatherogenic
properties of CRP include activation of endothelial cells to express adhesion mole-
cules, secretion of inflammatory cytokines such as IL-6, and decreased expression
and bioavailability of endothelial nitric oxide synthase (166). Using high-sensitivity
assays, CRP levels of <1, 1 to 3, and >3 mg/L correspond to low-, moderate-, and
high-risk groups, respectively, for future cardiovascular events (165). Effects of
dietary fats on CRP levels is not clear, but may be related to the overall inflamma-
tory potential of a particular type of dietary fat. For example, since n-6 rich oils, and
especially oils rich in LA, are pro-inflammatory, one might predict elevated CRP
levels in populations consuming diets high in the ratio of n-6 to n-3 fatty acids.
Homocysteine: Another clinical marker of inflammation and risk factor of athero-
sclerosis is elevated plasma levels of homocysteine (167). Homocysteine, a sulfur-
containing amino acid, is an intermediate formed during the metabolism of the
essential amino acid methionine (168). In the general U.S. population, some hyper-
homocysteinemia is quite common and often due to mild nutritional deficiencies in
mostly folic acid, but also in Vitamins B12 and B6. In fact, folic acid (or folate)
deficiency in adults can increase the risk of coronary artery disease, stroke, several
types of cancer, and possibly Alzheimers and Parkinsons diseases (169). Even
though a causal role of homocysteine in cardiovascular disease remains to be
established (170), the emerging data strongly suggest that elevated plasma homo-
cysteine levels increase the risk of multiple age-related diseases, and that adequate
dietary or supplemental folate can be a primary means of normalizing homocys-

teine levels and of increasing health span.
Peroxisome proliferators-activated receptors: PPARs appear to possess potent anti-
inflammatory signaling properties (171). PPARs are transcription factors, which
regulate gene expression by binding with the retinoid receptor RXR as a hetero-
dimeric partner to specific DNA sequence elements termed PPAR-responsive ele-
ments (171). The PPARs comprise three subtypes, PPARa, PPARd (or b), and
PPARg, with distinct expression patterns and biological functions (172, 173).
PPARa is predominantly expressed in liver, heart, muscle, and kidney, where it re-
gulates fatty acid catabolism. Molecular and genetic studies have established roles
for PPARg in adipocyte differentiation, lipid storage, and glucose homeostasis.
PPARd is expressed in most tissues and is implicated in lipid homeostasis and
wound healing. All three PPARs regulate macrophage cholesterol homeostasis by
enhancing cholesterol efflux. PPARs control plasma levels of cholesterol and tri-
acylglycerols and regulate expression of key proteins involved in all stages of athero-
genesis by exerting antiatherogenic actions at the level of the vascular wall (172).
Dietary fatty acids and AA metabolites are natural activators of PPARs. For exam-
ple, PPARa and PPARg are activated by eicosanoids derived from AA via the
lipoxygenase and cyclooxygenase pathways. Similar to PPARa and PPARg,
PPARb/d is a receptor for unsaturated fatty acids. PPARa and PPARg are also
expressed in both endothelial and smooth muscle cells in vitro and in vivo in the
human atherosclerotic plaque. There is general agreement in the literature that
PPARs have anti-inflammatory actions (172), but mechanisms and the importance
of selected vascular cells and tissues in this process are not clear.
In addition to regulating gene transcription via PPAR responsive elements,
PPARs have recently been shown to modulate gene expression by interfering
with other transcription factor pathways. PPARs have been shown to down-regulate
inflammatory response genes by negatively interfering with the NF-kB, AP-1, and
STAT transcriptional pathways (172). Furthermore, by regulating antioxidant
enzyme activities, such as catalase, PPAR activators may inhibit NF-kB activation
by reducing oxidative stress. PPAR activators also may antagonize NF-kB activa-
tion through the expression of the inhibitory protein IkBa. Such repression mechan-
isms via protein-protein interactions and cofactor competition may explain, in part,
the anti-inflammatory actions of PPARs. It needs to be determined how individual
types of fatty acids, e.g., n-6 and n-3 fatty acids and their biological metabolites,
regulate PPAR signaling to direct antiatherogenic actions.
5.9. Summary and Dietary Advice

Numerous animal and human studies suggest that dietary cholesterol and certain
saturated fatty acids increase serum as well as LDL-cholesterol concentrations.
Even though humans with elevated serum cholesterol levels may be at risk,
evidence also is mounting to suggest that the complicated processes that occur
during atherosclerosis involve not only the participation of modified lipoproteins,
but also low level and chronic inflammation and related disorders of the immune
Hypertriglyceridemia
Insulin Resistance High-Fat, High-Energy Diets
Obesity
Antioxidants
FA
CD36
FA
Oxidative Stress
Antioxidants/Agonists
? FA Metabolism
Antioxidants NF-B, AP-1 PPAR
Inflammatory Cytokines
Adhesion Molecules
Endothelial Cell Activation/Dysfunction

Figure 1. Proposed role of dietary fat (in particular unsaturated fat), excess calories and
hypertriglyceridemia, in the etiology of atherosclerosis. Dietary fats, rich in certain unsaturated
lipids, are atherogenic by contributing to increased cellular oxidative stress, leading to activation
of oxidative stress-sensitive transcription factors (e.g., NF-kB), which, in turn, promote cytokine
production, adhesion molecule expression, and, ultimately, endothelial barrier dysfunction and
atherosclerosis. The ability of membrane- or intracellular-mediated receptors or transcription
factors, like CD36 or PPARs, in modulating these events is still under investigation. Certain
nutrients, chemicals, or agonists, which have antioxidant properties, may protect against
atherosclerosis by acting at any one of the progressive steps of these signaling pathways.
system. Modified lipoprotein particles (e.g., oxidatively modified LDL) cause

secretion of inflammatory cytokines from blood-borne and arterial wall cells, which
will lead to endothelial cell activation. The resulting disturbances in endothelial
integrity possibly allow increased penetration of cholesterol-rich lipoprotein rem-
nants into the arterial wall, a critical event in the etiology of atherosclerosis. Mod-
ulations in the level of activity of a select set of oxidative stress-responsive
transcription factors (e.g., endothelial NF-kB) may provide a common mechanism
for linking these diverse processes (Figure 1). Reactive oxygen species appear to be
the common denominator in the many stimuli known to activate the NF-kB com-
plex and subsequent inflammatory events. One may speculate, then, that high levels
of dietary n-6 PUFA, which are easily oxidizable, can activate oxidative stress-
responsive transcription factors, which, in turn, may promote cytokine production,
adhesion molecule expression, endothelial barrier dysfunction, and, ultimately,
accelerated atherosclerosis. Interestingly, the activation of NF-kB can be inhibited

by a variety of antioxidants (Figure 1), which suggests that certain nutrients that
have antioxidant properties may protect against atherosclerosis by interfering
with the proposed mechanisms of endothelial cell dysfunction. In summary, the
research on atherosclerosis is complex, but the advice for patients at risk is simple:
eat less fat or calories, independent of its source, and eat only enough food to match
energy needs.
REFERENCES
1. D. Spielmann, U. Bracco, H. Traitler, G. Crozier, R. Holman, M. Ward, and R. Cotter,

J. Parenter. Enteral Nutr., 12, 111S (1988).
2. M. J. James, V. M. Ursin and L. G. Cleland, Am. J. Clin. Nutr., 77, 1140 (2003).
3. Dietary reference intakes Energy, carbohydrate, fiber, fat, fatty acids, cholesterol,
protein, and amino acids, The National Academies Press, Washington, D.C., 2002.
4. Dietary reference intakesEnergy, carbohydrate, fiber, fat, fatty acids, cholesterol,
protein, and amino acids, The National Academies Press, Washington, D.C., 2002.
5. S. D. Clarke, J. Nutr., 131, 1129 (2001).
6. S. D. Clarke, Br. J. Nutr., 83, S59-S66 (2000).
7. E. D. Rosen and B. M. Spiegelman, J. Biol. Chem., 276, 37731 (2001).
8. E. A. Meade, T. M. McIntyre, G. A. Zimmerman, and S. M. Prescott, J. Biol. Chem., 274,
8328 (1999).
9. R. Uauy, D. R. Hoffman, P. Peirano, D. G. Birch, and E. E. Birch, Lipids, 36, 885 (2001).
10. S. H. Werkman and S. E. Carlson, Lipids, 31, 91 (1996).
11. E. E. Birch, D. R. Hoffman, R. Uauy, D. G. Birch, and C. Prestidge, Pediatr. Res., 44, 201
(1998).
12. U. Suchner and U. Senftleben, Infusionsther. Transfusionsmed., 21, 167 (1994).
13. P. C. Calder, Proc. Nutr. Soc., 61, 345 (2002).
14. A. Gil, Biomed. Pharmacother., 56, 388 (2002).
15. M. J. James and L. G. Cleland, Semin. Arthritis Rheum., 27, 85 (1997).
16. S. Endres, R. Lorenz, and K. Loeschke, Curr. Opin. Clin. Nutr. Metab. Care, 2, 117
(1999).
17. J. R. Hibbeln and N. Salem, Jr., Am. J. Clin. Nutr., 62, 1 (1995).
18. B. A. Watkins and Y. Li, in R. E. C. Wildman, ed., Hand Book of Nutraceuticals and
Functional Foods, CRC Press, Boca Raton, Florida, 2001.
19. J. J. Turek, Y. Li, I. A. Schoenlein, K. G. D. Allen, and B. A. Watkins, J. Nutr. Biochem.,
9, 258 (1998).
20. Y. Li and B. A. Watkins, Lipids, 33, 417 (1998).
21. Y. Li, M. F. Seifert, D. M. Ney, M. Grahn, A. L. Grant, K. G. Allen, and B. A. Watkins,
J. Bone Miner. Res., 14, 1153 (1999).
22. J. R. Liu, B. Q. Chen, Y. M. Yang, X. L. Wang, Y. B. Xue, Y. M. Zheng, and R. H. Liu,
World J. Gastroenterol., 8, 999 (2002).
23. A. Miller, C. Stanton, and R. Devery, Anticancer Res., 22, 3879 (2002).
REFERENCES 601
24. E. J. Kim, J. G. Jun, H. S. Park, S. M. Kim, Y. L. Ha, and J. H. Park, Anticancer Res., 22,
2193 (2002).
25. J. D. Palombo, A. Ganguly, B. R. Bistrian, and M. P. Menard, Cancer Lett., 177, 163
(2002).
26. M. Yamasaki, H. Chujo, Y. Koga, A. Oishi, T. Rikimaru, M. Shimada, K. Sugimachi,
H. Tachibana, and K. Yamada, Cancer Lett., 188, 171 (2002).
27. E. J. Moon, Y. M. Lee, and K. W. Kim, Oncol. Rep., 10, 617 (2003).
28. P. Thuillier, G. J. Anchiraico, K. P. Nickel, R. E. Maldve, I. Gimenez-Conti, Muga, SJ,
K. L. Liu, S. M. Fischer, and M. A. Belury, Mol. Carcinog., 29, 134 (2000).
29. K. M. Hargrave, C. L. Li, B. J. Meyer, S. D. Kachman, D. L. Hartzell, M. A. Della-Fera,
J. L. Miner, and C. A. Baile, Obes. Res., 10, 1284 (2002).
30. N. Tsuboyama-Kasaoka, M. Takahashi, K. Tanemura, H. J. Kim, T. Tange, Okuyama, H,
M. Kasai, S. Ikemoto, and O. Ezaki, Diabetes, 49, 1534 (2000).
31. A. Akahoshi, Y. Goto, K. Murao, T. Miyazaki, M. Yamasaki, M. Nonaka, K. Yamada,
and M. Sugano, Biosci. Biotechnol. Biochem., 66, 916 (2002).
32. Y. Yu, P. H. Correll, and J. P. Vanden Heuvel, Biochim Biophys Acta, 1581, 89 (2002).
33. J. M. Peters, Y. Park, F. J. Gonzalez, and M. W. Pariza, Biochim. Biophys. Acta, 1533,
233 (2001).
34. M. A. Belury, S. Y. Moya-Camarena, K. L. Liu, and J. P. Vanden Heuvel, J. Nutr.
Biochem., 8, 579 (1997).
35. K. N. Lee, M. W. Pariza, and J. M. Ntambi, Biochem. Biophys. Res. Commun., 248, 817
(1998).
36. M. Sugano, A. Tsujita, M. Yamasaki, K. Yamada, I. Ikeda, and D. Kritchevsky, J. Nutr.
Biochem., 8, 38 (1997).
37. J. A. Bulgarella, D. Patton, and A. W. Bull, Lipids, 36, 407 (2001).
38. K. L. Liu and M. A. Belury, Cancer Lett., 127, 15 (1998).
39. B. A. Watkins and Y. Li, Adv. Conjugated Linoleic Acid Res., 2, 174 (2003).
40. Agency Response Letter GRAS Notice No. GRN 000041, 5-17-2001, U. S. Food and
Drug Administration, Washington, D.C.
41. Agency Response Letter GRAS Notice No. GRN 000080, 12-11-2001, U. S. Food and
Drug Administration, Washington, D.C.
42. E. N. Ponnampalam, G. R. Trout, A. J. Sinclair, A. R. Egan, and B. J. Leury, Meat Sci.,
58, 151 (2001).
43. P. R. C. Howe, J. A. Downing, B. F. S. Grenyer, E. M. Grigonis-Deane, and W. L.
Bryden, Lipids, 37, 1067 (2002).
44. S. Lopez-Ferrer, M. D. Baucells, A. C. Barroeta, and M. A. Grashorn, Poult. Sci., 78, 356
(1999).
45. S. Lopez-Ferrer, M. D. Baucells, A. C. Barroeta, and M. A. Grashorn, Poult. Sci., 80, 741
(2001).
46. S. Lopez-Ferrer, M. D. Baucells, A. C. Barroeta, J. Galobart, and M. A. Grashorn, Poult.
Sci., 80, 753 (2001).
47. T. Grune, K. Kramer, P. P. Hoppe, and W. Siems, Lipids, 36, 833 (2001).
48. S. M. Kitessa, D. Peake, R. Bencini, and A. J. Williams, Anim. Feed Sci. Tech., 108, 1
(2003).
49. M. Mozzon, N. G. Frega, B. Fronte, and M. Tocchini, Food Technol. Biotech., 40, 213
(2002).
50. G. Papadopoulos, C. Goulas, E. Apostolaki, and R. Abril, J. Dairy Res., 69, 357 (2002).
51. A. Bendich and P. E. Brock, Int. J. Vitam. Nutr. Res., 67, 213 (1997).
52. J. A. Lovegrove, C. N. Brooks, M. C. Murphy, B. J. Gould, and C. M. Williams, Br. J.
Nutr., 78, 223 (1997).
53. J. M. W. Wallace, A. J. McCabe, P. J. Robson, M. K. Keogh, C. A. Murray, G. Marquez-
Ruiz, H. McGlynn, W. S. Gilmore, and J. J. Strain, Ann. Nutr. Metab., 44, 157 (2000).
54. P. Y. Chouinard, L. Corneau, W. R. Butler, Y. Chilliard, J. K. Drackley, and D. E.
Bauman, J. Dairy Sci., 84, 680 (2001).
55. R. J. Baer, J. Ryali, D. J. Schingoethe, K. M. Kasperson, D. C. Donovan, A. R. Hippen,
and S. T. Franklin, J. Dairy Sci., 84, 345 (2001).
56. D. C. Donovan, D. J. Schingoethe, R. J. Baer, J. Ryali, A. R. Hippen, and S. T. Franklin,
J. Dairy Sci., 83, 2620 (2000).
57. A. A. Abu-Ghazaleh, D. J. Schingoethe, and A. R. Hippen, J. Dairy Sci., 84, 1845
(2001).
58. S. M. Innis, H. Sprecher, D. Hachey, J. Edmond, and R. E. Anderson, Lipids, 34, 139
(1999).
59. A. E. Leonard, S. Pereira, H. Sprecher, and Y. Huang, Prog. Lipid Res., 43, 36 (2004).
60. S. D. Harmon, T. L. Kaduce, T. D. Manuel, and A. A. Spector, Lipids, 38, 469 (2003).
61. A. E. Leonard, B. Kelder, E. G. Bobik, L. T. Chuang, C. J. Lewis, J. J. Kopchick,
P. Mukerji, and Y. S. Huang, Lipids, 37, 733 (2002).
62. E. A. Emken, R. O. Adlof, S. M. Duval, and G. J. Nelson, Lipids, 34, 785 (1999).
63. E. A. Emken, R. O. Adlof, S. M. Duval, and G. J. Nelson, Lipids, 33, 471 (1998).
64. V. P. Carnielli, D. J. Wattimena, I. H. Luijendijk, A. Boerlage, H. J. Degenhart, and P. J.
Sauer, Pediatr. Res., 40, 169 (1996).
65. C. Bolton-Smith, M. Woodward, and R. Tavendale, Eur. J. Clin. Nutr., 51, 619 (1997).
66. M. Murakami and I. Kuda, Prog. Lipid Res., 43, 3 (2004).
67. W. L. Xie, J. G. Chipman, D. L. Robertson, R. L. Erikson, and D. L. Simmons, Proc.
Natl. Acad. Sci. U. S. A., 88, 2692 (1991).
68. D. A. Kujubu, B. S. Fletcher, B. C. Varnum, R. W. Lim, and H. R. Herschman, J. Biol.
Chem., 266, 12866 (1991).
69. I. Morita, M. Schindler, M. K. Regier, J. C. Otto, T. Hori, D. L. DeWitt, and W. L. Smith,
J. Biol. Chem., 270, 10902 (1995).
70. C. C. Pilbeam, P. M. Fall, C. B. Alander, and L. G. Raisz, J. Bone Miner. Res., 12, 1198
(1997).
71. H. Kawaguchi, L. G. Raisz, O. S. Voznesensky, C. B. Alander, Y. Hakeda, and C. C.
Pilbeam, Endocrinology, 135, 1157 (1994).
72. N. V. Chandrasekharan, H. Dai, K. L. Roos, N. K. Evanson, J. Tomsik, T. S. Elton, and
D. L. Simmons, Proc. Natl. Acad. Sci. U. S. A., 99, 13926 (2002).
73. M. T. Donnelly and C. J. Hawkey, Aliment. Pharm. Therap., 11, 227 (1997).
74. N. Shimizu, Y. Ozawa, M. Yamaguchi, T. Goseki, K. Ohzeki, and Y. Abiko,
J. Periodontol., 69, 670 (1998).
75. J. Singh, R. Hamid, and B. S. Reddy, Cancer Res., 57, 3465 (1997).
REFERENCES 603
76. A. A. Nanji, D. Zakim, A. Rahemtulla, T. Daly, L. Miao, S. Zhao, S. Khwaja, S. R.

Tahan, and A. J. Dannenberg, Hepatology, 26, 1538 (1997).
77. H. Inoue, T. Tanabe, and K. Umesono, J. Biol. Chem., 275, 28028 (2000).
78. D. Hwang, Annu. Rev. Nutr., 20, 431 (2000).
79. M. Zeyda, A. B. Szekeres, M. D. Saemann, R. Geyeregger, H. Stockinger, G. J. Zlabinger,
W. Waldhausl, and T. M. Stulnig, J. Immunol., 170, 6033 (2003).
80. B. E. Cowing and K. E. Saker, J. Nutr., 131, 1125 (2001).
81. T. Lemberger, B. Desvergne, and W. Wahli, Annu. Rev. Cell Dev. Biol., 12, 335 (1996).
82. C. Chambrier, J. P. Bastard, J. Rieusset, E. Chevillotte, D. Bonnefont-Rousselot,
P. Therond, B. Hainque, J. P. Riou, M. Laville, and H. Vidal, Obes. Res., 10, 518 (2002).
83. S. Y. Moya-Camarena, J. P. Vanden Heuvel, S. G. Blanchard, L. A. Leesnitzer, and
M. A. Belury, J. Lipid Res., 40, 1426 (1999).
84. Y. Takahashi, M. Kushiro, K. Shinohara, and T. Ide, Biochim. Biophys. Acta 1631, 265
(2003).
85. E. Duplus and C. Forest, Biochem. Pharmacol., 64, 893 (2002).
86. J. M. Ntambi and H. Bene, J. Mol. Neurosci., 16, 273 (2001).
87. D. B. Jump and S. D. Clarke, Annu. Rev. Nutr., 19, 63 (1999).
88. G. P. Hayhurst, Y. H. Lee, G. Lambert, J. M. Ward, and F. J. Gonzalez, Mol. Cell. Biol.,
21, 1393 (2001).
89. Y. Inoue, G. P. Hayhurst, J. Inoue, M. Mori, and F. J. Gonzalez, J. Biol. Chem., 277,
25257 (2002).
90. S. A. Khan and J. P. V. Heuvel, J. Nutr. Biochem., 14, 554 (2003).
91. P. Couture, W. R. Archer, B. Lamarche, N. Landry, O. Deriaz, L. Corneau, J. Bergeron,
and N. Bergeron, Metabolism, 52, 1454 (2003).
92. P. Perez-Martinez, J. M. Ordovas, J. Lopez-Miranda, P. Gomez, C. Marin, J. Moreno,
F. Fuentes, R. A. F. de la Puebla, and F. Perez-Jimenez, Am. J. Clin. Nutr., 77, 809 (2003).
93. S. Vincent, R. Planells, C. Defoort, M. C. Bernard, M. Gerber, J. Prudhomme, P. Vague,
and D. Lairon, Proc. Nutr. Soc., 61, 427 (2002).
94. M. Rantala, M. L. Silaste, A. Tuominen, J. Kaikkonen, J. T. Salonen, G. Alfthan, A. Aro,
and Y. A. Kesaniemi, J. Nutr., 132, 3012 (2002).
95. R. Doll and R. Peto, J. Natl. Cancer Inst., 66, 1191 (1981).
96. Healthy, well nourished populationInteraction of diet and health, 1999. The United
States Department of Agriculture. Science and Education Impact.
97. A. P. Simopoulos, Biomed. Pharmacother., 56, 365 (2002).
98. M. de Lorgeril, P. Salen, J. L. Martin, I. Monjaud, J. Delaye, and N. Mamelle,
Circulation, 99, 779 (1999).
99. W. S. Harris, Am. J. Clin. Nutr., 65, 1645S (1997).
100. A. Leaf and P. C. Weber, N. Engl. J. Med., 318, 549 (1988).
101. D. P. Rose, Proc. Soc. Exp. Biol. Med., 216, 224 (1997).
102. D. P. Rose and J. M. Connolly, Pharmacol. Therap., 83, 217 (1999).
103. A. P. Simopoulos, Can. J. Physiol. Pharm., 75, 234 (1997).
104. Y. Goto, H. Tamachi, and E. H. Moriguchi, Prostaglandins Leukot. Essent. Fatty Acids,
48, 337 (1993).
105. C. E. Hock, M. A. Holahan, and D. K. Reibel, Am. J. Physiol., 252, H554H560

(1987).
106. G. E. Billman, J. X. Kang, and A. Leaf, Lipids, 32, 1161 (1997).
107. L. M. Braden and K. K. Carroll, Lipids, 21, 285 (1986).
108. T. Minoura, T. Takata, M. Sakaguchi, H. Takada, M. Yamamura, K. Hioki, and
M. Yamamoto, Cancer Res., 48, 4790 (1988).
109. B. S. Reddy and H. Maruyama, Cancer Res., 46, 3367 (1986).
110. B. S. Reddy and S. Sugie, Cancer Res., 48, 6642 (1988).
111. T.P.OConnor, B. D. Roebuck, F. J. Peterson, B. Lokesh, J. E. Kinsella, and T. C.
Campbell, J. Natl. Cancer Inst., 81, 858 (1989).
112. W. E. Lands, T. Hamazaki, K. Yamazaki, H. Okuyama, K. Sakai, Y. Goto, and V. S.
Hubbard, Am. J. Clin. Nutr., 51, 991 (1990).
113. H. Okuyama, T. Kobayashi, and S. Watanabe, Prog. Lipid Res., 35, 409 (1996).
114. T. Kuroishi and S. Tominaga, Jpn. J. Cancer Chemoth., 28, 168 (2001).
115. K. Shichikawa, Y. Takenaka, A. Maeda, R. Yoshino, M. Tsujimoto, H. Ota,
T. Kashiwade, and I. Hongo, Ryumachi, 21, 35 (1981).
116. W. E. Lands, Ann. N. Y. Acad. Sci., 676, 46 (1993).
117. Y. Goto, Nutr. Rev., 50, 398 (1992).
118. D. Hwang and S. H. Rhee, Am. J. Clin. Nutr., 70, 545 (1999).
119. D. B. Jump, J. Biol. Chem., 277, 8755 (2002).
120. O. Adam, Eur. J. Clin. Nutr., 49, 703 (1995).
121. J. W. Alexander, Nutrition, 14, 627 (1998).
122. D. R. Robinson, S. Tateno, B. Patel, and A. Hirai, J. Parenter. Enteral Nutr., 12, 37S
(1988).
123. T. H. Lee, T. Sethi, A. E. Crea, W. Peters, J. P. Arm, C. E. Horton, M. J. Walport, and
B. W. Spur, Clin. Sci., 74, 467 (1988).
124. P. Needleman, A. Raz, M. S. Minkes, J. A. Ferrendelli, and H. Sprecher, Proc. Natl.
Acad. Sci., U. S. A., 76, 944 (1979).
125. B. A. Watkins, Y. Li, K. G. D. Allen, W. E. Hoffmann, and M. F. Seifert, J. Nutr., 130,
2274 (2000).
126. L. G. Raisz, C. B. Alander, and H. A. Simmons, Prostaglandins, 37, 615 (1989).
127. H. R. Knapp, I. A. Reilly, P. Alessandrini, and G. A. FitzGerald, N. Engl. J. Med., 314,
937 (1986).
128. J. M. Kremer, Lipids, 31, S243-S247 (1996).
129. C. L. Curtis, C. E. Hughes, C. R. Flannery, C. B. Little, J. L. Harwood, and B. Caterson,
J. Biol. Chem., 275, 721 (2000).
130. L. C. Hamilton, J. A. Mitchell, A. M. Tomlinson, and T. D. Warner, FASEB J., 13, 245
(1999).
131. A. F. Badawi, A. El Sohemy, L. L. Stephen, A. K. Ghoshal, and M. C. Archer,
Carcinogenesis, 19, 905 (1998).
132. R. Graber, C. Sumida, and E. A. Nunez, J. Lipid Mediat. Cell Signal, 9, 91 (1994).
133. J. A. Ross, A. G. Moses, and K. C. Fearon, Curr. Opin. Clin. Nutr. Metab. Care, 2, 219
(1999).
REFERENCES 605
134. S. Camandola, G. Leonarduzzi, T. Musso, L. Varesio, R. Carini, A. Scavazza,

E. Chiarpotto, P. A. Baeuerle, and G. Poli, Biochem. Biophys. Res. Commun., 229,
643 (1996).
135. R. Ross, N. Engl. J. Med., 340, 115 (1999).
136. T. Collins, Lab. Invest., 68, 499 (1993).
137. F. De Nigris, L. O. Lerman, M. Condorelli, A. Lerman, and C. Napoli, Antioxid. Redox
Sign, 3, 1119 (2001).
138. A. Keys, J. T. Anderson, and F. Grande, Lancet, 2, 959 (1957).
139. D. M. Hegsted, R. B. McGandy, M. L. Myers, and F. J. Stare, Am. J. Clin. Nutr., 17, 281
(1965).
140. R. P. Mensink, Am. J. Clin. Nutr., 57, 711S (1993).
141. P. Kris-Etherton, R. H. Eckel, B. V. Howard, S. St Jeor, and T. L. Bazzarre, Nutrition
Committee Population Science Committee and Clinical Science and Committee of the
American Heart Association, Circulation, 103, 1823 (2001).
142. F. Perez-Jimenez, J. Lopez-Miranda, and P. Mata, Atherosclerosis, 163, 385 (2002).
143. F. M. Sacks and M. Katan, Am. J. Med., 113, 13S (2002).
144. P. N. Hopkins, Am. J. Clin. Nutr., 55, 1060 (1992).
145. M. Laidlaw and B. J. Holub, Am. J. Clin. Nutr., 77, 37 (2003).
146. R. N. Lemaitre, I. B. King, D. Mozaffarian, L. H. Kuller, R. P. Tracy, and D. S. Siscovick,
Am. J. Clin. Nutr., 77, 319 (2003).
147. W. E. Connor and S. L. Connor, Adv. Internal Med., 35, 139 (1990).
148. M. J. Malloy and J. P. Kane, Adv. Internal Med., 47, 111 (2001).
149. M. A. Austin, B. McKnight, K. L. Edwards, C. M. Bradley, M. J. McNeely, B. M. Psaty,
J. D. Brunzell, and A. G. Motulsky, Circulation, 101, 2777 (2000).
150. L. Djousse, S. C. Hunt, D. K. Arnett, M. A. Province, J. H. Eckfeldt, and R. C. Ellison,
Am. J. Clin. Nutr., 78, 1098 (2003).
151. D. R. Janero, Free Radical Biol. Med., 11, 129 (1991).
152. K. Esposito, F. Nappo, F. Giugliano, G. Giugliano, R. Marfella, and D. Giugliano, Am. J.
Clin. Nutr., 77, 139 (2003).
153. J. M. Hodgson, M. L. Wahlqvist, J. A. Boxall, and N. D. Balazs, Am. J. Clin. Nutr., 58,
228 (1993).
154. R. J. Nicolosi, T. A. Wilson, G. Handelman, T. Foxall, and J. F. J. Keaney, J. Nutr.
Biochem., 13, 392 (2002).
155. E. Vos and S. C. Cunnane, Am. J. Clin. Nutr., 77, 521 (2003).
156. S. M. Grundy and M. A. Denke, J. Lipid Res., 31, 1149 (1990).
157. S. M. Grundy, N. Abate, and M. Chandalia, Am. J. Med., 113, 25S (2002).
158. F. B. Hu and W. C. Willett, J. Am. Med. Assoc., 288, 2569 (2002).
159. P. M. Kris-Etherton, P. M. Kris-Etherton, A. E. Binkoski, G. Zhao, S. M. Coval, K. F.
Clemmer, K. D. Hecker, H. Jacques, and T. D. Etherton, Proc. Nutr. Soc., 61, 287 (2002).
160. P. O. Bonetti, L. O. Lerman, and A. Lerman, Arterioscl. Throm. Vas. Biol., 23, 168
(2003).
161. M. Toborek, Y. W. Lee, R. Garrido, S. Kaiser, and B. Hennig, Am. J. Clin. Nutr., 75, 119
(2002).
162. S. K. Peng, C. B. Taylor, J. C. Hill, and R. J. Morin, Atherosclerosis, 54, 121 (1985).
163. G. A. Boissonneault, B. Hennig, and C. M. Ouyang, Proc. Soc. Exp. Biol. Med., 196, 338
(1991).
164. B. Hennig and G. A. Boissonneault, Atherosclerosis, 68, 255 (1987).
165. P. M. Ridker, Circulation, 107, 363 (2003).
166. E. T. H. Yeh and J. T. Willerson, Circulation, 107, 370 (2003).
167. J. A. Farmer and G. Torre-Amione, Curr. Atheroscl. Rep., 4, 92 (2002).
168. B. K. Nallamothu, A. M. Fendrick, and G. S. Omenn, Pharmacoeconomics, 20, 429
(2002).
169. M. P. Mattson, I. I. Kruman, and W. Duan, Ageing Res. Rev., 1, 95 (2002).
170. E. Falk, J. Zhou, and J. Moller, Lipids, 36, S3 (2001).
171. A. Tedgui and Z. Mallat, Circ. Res., 88, 877 (2001).
172. O. Barbier, I. P. Torra, Y. Duguay, C. Blanquart, J. C. Fruchart, C. Glineur, and B. Staels,
Arterioscl. Throm. Vas. Biol., 22, 717 (2002).
173. C. H. Lee and R. M. Evans, Trends Endocrinol. Metabol., 13, 331 (2002).
BAILEYS INDUSTRIAL
OIL AND FAT
PRODUCTS
Sixth Edition
Volume 2
Edible Oils
Edited by
Fereidoon Shahidi


1
Butter
David Hettinga
1. INTRODUCTION
Buttermaking is one of the oldest forms of preserving the fat component of milk. Its
manufacture dates back to some of the earliest historical records, and reference has
been made to the use of butter in sacrificial worship, for medicinal and cosmetic
purposes, and as a human food long before the Christian era. Documents indicate
that, at least in the Old World, the taming and domestication of animals constituted
the earliest beginnings of human civilization and culture. There is good reason to
believe, therefore, that the milking of animals and the origin of buttermaking
predate the beginning of organized and permanent recording of human activities.
The evolution of the art of buttermaking has been intimately associated with the
development and use of equipment. With the close of the eighteenth century, the
construction and use of creaming and buttermaking equipment (other than that
made of wood) began to receive consideration, and the barrel churn made its
appearance.
By the middle of the nineteenth century, attention was given to improvement in
methods of creaming. These efforts gave birth to the deep-setting system. Up to that
time, creaming was done by a method called shallow pan. The deep-setting system
shortened the time for creaming and produced a better quality cream. An inventive
Bavarian brewer, in 1864, conceived the idea of adapting the principle of the
laboratory centrifuge. In 1877, a German engineer succeeded in designing a
machine that, although primitive, was usable as a batch-type apparatus. In 1879,
1
2 BUTTER
engineers in Sweden, Denmark, and Germany succeeded in the construction of

cream separators for fully continuous operation (1).
In 1870, the year before the introduction of factory buttermaking, butter produc-
tion in the United States totaled 514 million lbs, practically all farm made. Authen-
tic records concerning the beginning of factory buttermaking are meager. It appears
that the first butter factory was built in Iowa in 1871. This also introduced the
pooling system of milk for creamery operation (1).
Other inventions that assisted in the development of the butter industry included
the Babcock test (1890), which accurately determines the percentage of fat in milk
and cream; the use of pasteurization to maintain milk and cream quality; the use of pure
cultures of lactic acid bacteria; and refrigeration to help preserve cream quality.
Multiple butter fat products, including butter oils, anhydrous butter fat, butter
fatvegetable oil blends, and fractionated butter fats, are manufactured around
the world today. In the past, butter fat in the form of butter was the primary pre-
servation technique. Today, the preferred preservation method involves the proces-
sing of butter fat to the anhydrous butter oil state, then hermetically packaging
under nitrogen to substantially increase the shelf life and reduce the incidence of
degradation.
Historically, milkfat has been held in the highest esteem, whether in liquid milk,
as cream, or as butter. Its consumption was associated with a higher standard of
living. In recent times, with the prosperity of the Western world, per capita con-
sumption has been decreasing. Ironically, this phenomenon contradicts all historical
patterns for butter fat consumption and use. Several reasons exist for this decline.
This chapter explores the chemical composition, marketing, technology, processing,
quality, legal restrictions, and uses for butter and butter fat.
2. CHEMICAL COMPOSITION
Some of the information in this chapter comes directly from the fourth edition of
Baileys (2). Jensen and Clark (3) have provided a complete review of the lipid
composition, and data have been selected for inclusion in this review.
The composition of milkfat is somewhat complex. Although dominated by tri-
glycerides, which constitute some 98% of milkfat (with small amounts of diglycer-
ides, monoglycerides, and free fatty acids), various other lipid classes are also
present in measurable amounts. It is estimated that about 500 separate fatty acids
have been detected in milk lipids; it is probable that additional fatty acids remain to
be identified. Of these, about 20 are major components; the remainder are minor
and occur in small or trace quantities (4, 5). The other components include phos-
pholipids, cerebrosides, and sterols (cholesterol and cholesterol esters). Small
amounts of fat-soluble vitamins (mainly A, D, and E), antioxidants (tocopherol),
pigments (carotene), and flavor components (lactones, aldehydes, and ketones)
are also present.
The composition of the lipids of whole bovine milk is given in Table 1 (4, 5).
The structure and composition of the typical milkfat globule is exceedingly
CHEMICAL COMPOSITION 3
TABLE 1. Composition of Lipids in Whole

Bovine Milk (4, 5).
Lipid Weight Percent
Hydrocarbons Trace
Sterol esters Trace
Triglycerides 9798
Diglycerides 0.280.59
Monoglycerides 0.0160.038
Free fatty acids 0.100.44
Free sterols 0.220.41
Phospholipids 0.21.0
complex. The globule is probably 23 m in diameter with a 90-A-thick membrane

surrounding a 9899% triglyceride core. The composition of the milkfat membrane is
quite different from milkfat itself in that approximately 60% triglycerides are pre-
sent, much less than in the parent milkfat (Table 2) (6, 7).
It has been generally recognized that butter fat consists of about 15 major fatty
acids, with perhaps 12 or so minor (trace quantity) acids. Triglycerides are normally
defined with respect to their carbon number (CN), i.e., the number of fatty acid car-
bon atoms present in the molecule; the three carbon atoms of the glycerol moiety
are ignored. As the fatty acid spectrum of milkfat is dominated by acids containing
an even number of carbon atoms, so is the triglyceride spectrum. However, the pro-
portion of triglycerides with an odd carbon number is about three times greater than
the proportion of odd-numbered fatty acids.
Although obvious correlations exist between fatty acid composition and trigly-
ceride distribution, detailed information is lacking that would enable the triglycer-
ide distribution to be predicted from the fatty acid composition. Much more needs
to be understood of the strategy used in the bovine mammary gland in assembling a
TABLE 2. Composition of Lipids from Milkfat Globule

Membrane (6, 7).
Lipid Component Percent of Membrane Lipids
Carotenoids (pigment) 0.45

Squalene 0.61
Cholesterol esters 0.79
Triglycerides 53.4
Free fatty acids 6.3a
Cholesterol 5.2
Diglycerides 8.1
Monoglycerides 4.7
Phospholipids 20.4
a
Contained some triglycerides.
4 BUTTER
TABLE 3. Characteristics and Composition of Butter Fats.
Characteristic Valuea Range of Valuesb GLCc
Iodine number 32.9

Saponification equivalent 236.3
Reichert-Meissle value 32.5
Polenske value
Kirschner value
Fatty acid, wt. %
Butyric 3.5 2.84.0 3
Caproic 1.4 1.43.0 1
Caprylic 1.7 0.51.7 1
Capric 2.6 1.73.2 3
Lauric 4.5 2.24.5 4
Myristic 14.6 5.414.6 12
Palmitic 30.2 2641 29
Stearic 10.5 6.111.2 11
Above C18 1.6 2
Total saturated 70.6 66
Decenoic 0.3 0.10.3
Dodecenoic 0.2 0.10.6
Tetradecenoic 1.5 0.61.6 2
Hexadecenoic 5.7 2.85.7 4
Octadecenoic (oleic, etc.) 18.7 18.733.4 25
Octadecdienoic 2.1 0.93.7 2
C20 and C22 unsaturated 0.9 1
Total unsaturated 29.4 34
a
From (8) and (9).
b
From (10) and (11).
c
From (12).
complex array of fatty acids into triglycerides. This is not an arcane study; it is
necessary if processes such as fractionation are to yield products with consistent
qualities throughout the year. In effect, the detailed structure of milkfat is not yet
understood. Perhaps this is not surprising if we consider only the 15 major fatty
acids; there are 153 (3375) possible triglyceride structures using a purely random
model.
The data in Table 3 represent general characteristics and composition of butter
fat as reported by several sources (812). Note the range in values. Precise and
repeatable values are not highly correlated due to such variables as stage of lacta-
tion, feed source, cattle breed, etc. Although 16 categories of fatty acids are out-
lined, it was generally appreciated that many other fatty acids are present in
small or trace quantities. For nutritional and dairy science purposes, these data
are of value, but from a detailed scientific point of view, they afford only a vague,
broad generalization of the actual state of fatty acid composition of butter fat. A
more complete view of composition is provided in Table 4 (13, 14).
From 1956 to 1983 a great volume of information became available on the
occurrence of many minor constituents in butter fat. Somewhat less intensity
TABLE 4. Fatty Acid Composition of Milk and Butter Fat.a
Fatty Acidb Junec Decemberd Averagee Moore and Co-workersf
4:0 4.22 3.51 3.57 3.98

6:0 2.53 2.24 2.22 2.36
8:0 2.34 1.07 1.17 1.36
9:0 0.05 0.05 0.03
10 : 0 2.24 2.57 2.54 2.76
10 : 1 0.32
11 : 0 0.34 0.29 0.33
12 : 0 2.40 2.77 2.81 3.14
13 : 0 (12 : 1) 0.29 0.29 0.33 0.14
14 (br)g 0.23 0.14 0.17 0.12
14 : 0 9.01 10.58 10.06 8.39
14 : 1 (15 br) 1.54 1.61 1.63 1.84
15 : 0 1.29 1.11 1.09 1.34
16 (br) 0.42 0.39 0.38 0.35
16 : 0 22.05 25.98 24.97 30.05
16 : 1 (17 br) 2.29 2.98 2.55 2.80
17 : 0 0.69 1.08 0.91 1.00
17 : 1 (18 br) 0.37
18 : 0 (br) 0.31 0.40 0.38
18 : 0 14.27 11.58 12.07 11.74
18 : 1 30.41 24.75 27.09 24.93
18 : 8h 0.24 1.56 1.26
18 : 2 1.23 2.75 2.39 1.78
18 : 3 (20 : 0) 2.61 2.30 2.06 1.23
a
In weight percent.
b
Structural assignments are not necessarily authentic, but represent, in almost all instances, the most likely
structure for the fraction.
c
Data from the Department of Animal Industries, Storrs (Conn.) Agricultural Experiment Station; 408 samples
of milk plant production from June 1960 to June 1961.
d
Data from Storrs Agriculture Experiment Station; 48 samples.
e
For 108 samples.
f
Sec (14).
g
Branched chain.
h
Carbon number obtained by semilog plots retention time/chain length.
of interest has prevailed since then, but further information continues to appear, and
we can expect more data on butter fat as a consequence of research on the
relationship between dairy cow feeding studies and resulting butter fat fatty acid
composition.
The great variety of fatty acids in butter fat cannot be treated in detail here; refer-
ence will be made to only a few of the many available reports. Octadecadienoic
acids are present in significant amounts; there are traces of hexadecadienoic acid,
octadecatrienoic acids, and highly unsaturated C20 and C22 acids. Traces of dihy-
droxystearic acid and hydroxypalmitic acid have been detected (8, 9). A small pro-
portion of the octadecenoic acid consists, not of oleic acid, but of trans-11,12
isomer, vaccenic acid (8, 9). One report states that about 66% of one octadecienoic
6 BUTTER
TABLE 5. Positional and Geometric Isomers of Bovine Milk Lipid Fatty Acids (wt. %) (16).
Cis-Isomers Trans-Isomers
Position of
Double Bond 14:1 16:1 17:1 18:1 16:1 18:1
5 1.0 Trace 2.2

6 0.8 1.3 3.4 7.8 1.0
7 0.9 5.6 2.1 6.7 0.8
8 0.6 Trace 20.1 1.7 5.0 3.2
9 96.6 88.7 71.3 95.8 32.8 10.2
10 Trace Trace Trace 1.7 10.5
11 2.6 2.9 2.5 10.6 35.7
12 Trace Trace 12.9 4.1
13 10.6 10.5
14 9.0
15 6.8
16 7.5
acid content is normal linoleic acid, and the remainder consists of the cis-9,
trans-12 or the trans-9, cis-12 isomers (15); but other positional and geometric
isomers are undoubtedly also present (4). The positional and geometric isomers
of bovine milk lipid fatty acids are presented in Table 5 (16).
TABLE 6. Fatty Acid Distributions of 82 Acids in Butter Fat.a
Saturatedb Branchedc Monoenes

Acid Weight Percent Acid Weight Percent Acid Weight Percent
12:0 i 0.01 10:1 0.48

8:0 0.69 13:0 i Trace 12:1 0.05
10:0 1.88 14:0 i 0.03 13:1 0.003
11:0 0.12 15:0 i 0.14 14:1 0.75
12:0 2.96 15:0 2 0.23 15:1 0.02
13:0 0.10 16:0 i 0.2 16:1 1.84
14:0 11.2 17:0 i 0.36 17:1 0.2
15:0 1.52 18:0 i 0.02 18:1 30.3
16:0 27.8 19:0 br 0.01 19:1 0.14
17:0 0.71 20:0 br 0.01
18:0 12.1 21:0 br 0.01
19:0 0.05 22:0 br 0.02
20:0 0.02 23:0 br 0.01
21:0 0.06 24:0 br 0.02
22:0 0.04 25:0 br 0.0004
23:0 0.01 26:0 br 0.0004
24:0 0.02 20:1 0.52
25:0 0.02 21:1 0.01
26:0 0.02 22:1 0.02
27:0 0.00004 23:1 0.05
28:0 0.00004 24:1 0.0008
25:1 0.0008
26:1 0.0008
TABLE 6. Fatty Acid Distributions of 82 Acids in Butter Fat.a (Continued )
Dienes Polyenes Multibranchede

Acid Weight Percent Acid Weight Percent Acid Weight Percent
14:2 0.04 18:3 1.03 16:0 br3 0.01

16:2 0.02 18:4 0.10 17:0 br3 0.01
18:2 2.22 20:3 0.05 18:0 br3 0.16
20:2 0.12 20:4 0.07
22:2 0.14 20:5 0.02
24:2 0.02 22:3 0.03
26:2 0.0004 22:4 0.04
22:5 0.02
19:0 br4d 0.02
20:0 br4 0.14
21:0 br4 0.02
22:0 br4 0.02
23:0 br4 0.01
24:0 br4 0.10
25:0 br4 0.10
26:0 br3 0.01
27:0 br4 0.04
28:0 br3 0.02
28:0 br4 0.12
28:0 br5 0.01
a
Detected by urea fractionation and gasliquid chromatography in 1965 (17).
b
Acid below 8:0 were not determined (totally or partially lost during removal of solvent); also did not measure
trans-isomers, conjugated dienes and trienes, and keloacids.
c
i, iso; br, iso and/or anti-iso. Last number indicates number of methyl branches for multibranched acids.
d
The number following br indicates the number of methyl branches for multibranched acids.
e
Tentatively identified in appropriate urea fractions by semilogarithmic plots of GLC retention times.
Few compilations of the extensive fatty acid distributions in butter fat have been
made since Iverson et al. (17) reported quantitative data on 82 fatty acids that were
detected by means of urea fractionation and gasliquid chromatography (GLC)
(Table 6). Table 7 provides the fatty acid composition of bovine milk lipids.
The advent of new techniques of gas chromatography for monoglycerides, digly-
cerides, and triglycerides (18, 19) should assist markedly in the identification of the
specific triglycerides of butter fat. It has already been possible to identify and quan-
titate about 168 molecular species of bovine milk serum triglycerides, excluding
enantiomers. Nutter and Privett (20) employed liquidliquid and argentation thin-
layer chromatography (TLC) along with pancreatic lipase hydrolysis for this pur-
pose. As a result of their high degree of saturation, ruminant milkfats do not lend
themselves readily to argentation TLC, and resolution by gas chromatography using
polyester columns is a likely recourse.
There is a pronounced seasonal change in the fatty acid composition of butter
fat. It is normally several iodine number units higher in the summer than in the
winter, with corresponding variation in the relative proportions of unsaturated
8 BUTTER
TABLE 7. Fatty Acid Composition of Bovine Milk Lipids, August 1983 (3).
Number Type Identity
27 Normal saturate 228

25 Monobranched saturate 24; 13, 15, 17, 18 three or more positional
isomers
16 Multibranched 1628
62 Cis monoene 1026, except for 11 : 1, positional isomers of
12 : 1, 14 : 1, 16 : 118 : 1, and 23 : 125 : 1
58 Trans monoene 1214,1624; positional isomers of 14 : 1,
16 : 118 : 1, and 23 : 125 : 1
45 Diene 1426 evens only; cis, cis; cis, trans; or trans,
cis and trans; trans, geometric isomers;
unconjugated and conjugated and positional
isomers
10 Tripolyene 18, 20, 22; geometric positional, conjugated
and unconjugated isomers
5 Tetrapolyene 18, 20, 22; positional isomers
2 Pentapolyene 20, 22
1 Hexapolyene 22
38 Keto (oxo) saturated 10, 12, 14, 1520, 22, 24; positional isomers
21 Keto (oxo) unsaturated 14, 16, 18; positional isomers of carbonyl and
double bond
16 Hydroxy, 2-position 14 : 0, 16 : 026 : 0, 16 : 1, 18 : 1, 21 : 1, 24 : 1, 25 : 1
Hydroxy, 4- and 5-position 10 : 016 : 0, 12 : 6 and 12 : 1 9
60 Other positions
1 Cyclic, hexyl 11; terminal cyclohexyl
and saturated fatty acids. In colder climates, the difference appears to be slightly
larger. The change is usually associated with the difference in the feed of the ani-
mals in different seasons, but not completely so: cows put on green pasturage pro-
duce softer butter fat even if their feed has previously consisted of hay or silage
comparable in solid composition with the green feed.
There are also differences in the butter fat of different cows on identical rations,
and the age of the animal and duration of lactation have some influence on butter fat
composition. Much of the dairy literature provides information relating dairy ani-
mal species and the composition of the butter fat from them.
When corn and peanut oils are protected (entrapped in formaldehyde-treated
casein), significant changes in the fatty acid composition of milkfat occur
(Table 8) (21).
Protected oils are hydrolyzed in the abomasum, and the fatty acids are absorbed
in the small intestine, thereby avoiding hydrogenation. The 18:2 content in the
milkfat was increased about five-fold, and the 14:0, 16:0, and 18:0 were decreased
accordingly. Plasma and depot fats were also increased in 18:2 content by this
program (21).
Results at the USDA are similar: cows milk can be increased in 18:2 acid from
3% to 35% by feeding protected safflower oil (22, 23). However, at high 18:2 levels,
TABLE 8. Effect of Feeding Protected Corn and Peanut Oils

on Fatty Acid Composition of Bovine Milkfat (4, 21).
Fatty Acid Composition of Milk Lipids (wt. %)

Fatty Acids Corn Oil Peanut Oil Control
14 : 0 7.9 9.7 11.9

16 : 0 20.5 22.1 31.1
18 : 0 9.8 11.0 13.5
18 : 1 28.8 25.3 29.5
18 : 2 20.1 20.5 4.2
18 : 3 1.8 2.9 2.7
Others 11.1 8.5 7.1
milk develops an oxidized off-flavor, usually after about 24 h, and creams require a
longer aging time for satisfactory churning. As expected, butter that contains more
than 16% linoleic acid is soft and sticky (5).
Extensive data have been published on the Reichert-Meissl, Polenske, and
Kirschner values of mixtures of butter fat, coconut, and palm kernel oils (Table 9)
(2426). Other average characteristics of butter fat are approximately as follows:
density at 60 C, 0.887; melting point, 38 C; titer, 34 C; and unsaponifiable matter,
0.4%. The optical properties of butter fat are misleading and are in part contributed
by the nonglyceride components.
A significant variation in milkfat composition can occur in colostrum milk.
Ahren et al. (27) analyzed the content of glycerol ethers in neutral lipids and phos-
pholipids isolated from bovine colostrum and milk (Table 10). Lactone content of
butter fat has also been determined (Table 11).
Odd-numbered methyl ketones containing from 3 to 15 carbon atoms are found
in small quantities in butter fat. These compounds, along with microtraces of acet-
one, acetaldehyde, methyl sulfide, C4C10 free fatty acids, and the various lactones
already mentioned, generally are considered to be the substances that comprise the
pleasant, bland, olfactory, nonoxidative flavor and odor of milkfat. Representative
concentrations of homologous methyl ketones have been well documented (3032).
TABLE 9. Distinctive Characteristics of Butter Fat Compared with Other Fats (23).
Soy and Corn

Characteristic Butter Fat Coconut Oil Palm Kernel Oil Fats and Oils
Saponification number 210250 245260 240250 200

Refractive index, 60 C 1.4465 1.4410 1.4430 >1.4465a
Reichert-Meissl value 22.34 6.8 5.7 <1
Polenske value 224 1418 1012 <1
Kirschner value 2026 12 0.51 <0.5
a
Unless the iodine number is very nearly zero.
10 BUTTER
TABLE 10. Content of Glycerol Ethers in Neutral Lipids and Phospholipids Isolated
from Bovine Colostrum and Milk (27).
Characteristic Colostrum (% wt/wt) Milk (% wt/wt)
Total lipids 5.6 3.9

Neutral lipids in total lipids 99.0 99.3
Phospholipids in total lipids 1.0 0.7
Glycerol ethers in total lipids 0.061 0.009
Glycerol ethers in natural lipids 0.06 0.007
Glycerol ethers in phospholipids 0.16 0.25
Glycerol ethers in natural lipids of 97.4 80
total glycerol ethers
Glycerol ethers in phospholipids of 2.6 20
total glycerol ethers
The phospholipids of milkfat are found in the fat globule membrane in associa-
tion with proteins and cerebrosides. Phospholipids are amphipolar in nature and are
strongly surface active. These properties enable them to stabilize both oil-in-water
and water-in-oil emulsions (Table 12) (46).
The sterols found in the unsaponifiable fraction of milk lipids are mostly choles-
terol esters, small quantities of lanosterol, and even smaller quantities of two new
constituents: dihydrolanosterol and b-sitosterol (33).
From the standpoint of nutritional value, the Vitamin A content of butter is
important. As the source of Vitamin A in butter is b-carotene or other carotenoid
pigments in the feed of the cows, the content of this vitamin varies considerably,
being highest in the summer when the dairy herds are in pasture and lowest in
winter when there are no green feedstuffs in their rations. A portion of the carotene
TABLE 11. Amounts of y- and s-Aliphatic Lactones Isolated

from Butter Fat (ppm) (2, 28, 29).
Carbon Number s-Lactones y-Lactones
6 2.0 Trace
7 0.2a
8 2.6 0.5
9 0.4a 0.2
10 15.0 1.2
11 0.7 0.5
12 35.0 1.6
13 1.5 0.5
14 34.0 1.4
15 6.4 1.3
16 23.2 1.3
18 2.3
2,3-Dimethyl-2,4-nonadien-4-olide 0.5
a
Semiquantitative.
TABLE 12. Phospholipid Content of Bovine Milk

(46).
Phospholipid Mole Percent
Phosphatidylcholine 34.5
Phosphatidylethanolamine 31.8
Phosphatidylserine 3.1
Phosphatidylinositol 4.7
Sphingomyelin 25.2
Lysophosphatidylcholine Trace
Lysophosphatidylethanolamine Trace
Total choline phospholipids 59.7
Plasmalogens 3
Diphosphatidyl glycerol Trace
Ceramides Trace
Cerebrosides Trace
in the feed is transferred to the butter fat without change. The amount of carotene
transferred by the cow into the butter fat varies with the feeding regimen parallel to
variations in the production of Vitamin A, so that the intensity of the yellow color of
butter, to some extent, serves to indicate its Vitamin A content.
The Vitamin A potency of butter is in part due to Vitamin A as such and in part
to carotene, which is partially converted to the vitamin in the human body. The
Vitamin A content of butter is usually within the range of 612 mg/g, and the car-
otene content is in the range of 210 mg/g (33); 1 IU of Vitamin A is defined as the
amount possessing the biological activity of 0.6 mg of pure b-carotene.
The Vitamin D content of butter is much less significant than that of Vitamin A,
but it is nevertheless appreciable. It varies from about 0.1 IU/g to 1.0 IU/g, being
highest in the summer and lowest in the winter (33).
The composition of milkfat is the most important factor affecting the firmness of
butter and, therefore, its spreadability. The composition of milkfat changes primar-
ily according to the feed; therefore, the entire problem is connected to the animals
diet. The fatty acid composition of milkfat produced in various countries has been
rather accurately determined, as have the seasonal variations. In Europe, the amount
of saturated fatty acids is generally highest in winter and lowest in summer or fall
(see Table 11) (34). Green fodder decreases the amount of saturated fatty acids and
correspondingly increases the amount of unsaturated fatty acids. The differences
between the maximum and minimum values can be fairly large. For palmitic and
oleic acids, the quantitatively most important fatty acids, a difference of more than
10% between the maximum and minimum values was found in some cases. This
makes it understandable that there are also significant differences in the physical
characteristics of the butter. The structure of the triglycerides in the milkfat, along
with the fatty acid composition, is important in determining the physical character-
istics of the fat, because the softening point of fat has been found to rise as the result
of interesterification (35).
12 BUTTER
TABLE 13. Compositional Characteristics of Summer and Winter Milkfat (36).a
Fatty Acids
Iodine
Samples Volatile Saturated Monounsaturated Polyunsaturated Number
Average of total 10.98 56.50 29.81 2.50 32.2

Summer 9.49 58.82 33.53 3.14 36.8
Winter 12.45 59.15 26.15 1.86 27.7
a
N 140.
Textural characteristics of butter significantly depend on milkfat composition

and the method of manufacture. If the chemical composition of the milkfat is
known, it is possible to select the appropriate technological parameters of the but-
termaking to improve its texture. To obtain butter with constant rheological char-
acteristics and to control the parameters of the buttermaking process, it is necessary
to take into account the difference in the chemical composition and the properties
of the milkfat in various seasons. Table 13 shows various compositional changes of
milkfat derived from summer and winter milk (36).
3. MODIFICATION OF MILKFAT
3.1. Melting and Crystallization of Milkfat Triglycerides

The complex fatty acid composition of milkfat is reflected in its melting behavior.
Melting begins at 30 C and is complete only at 37 C. At any intermediate tem-
perature, milkfat is a mixture of solid and liquid. To a large extent, the solid: liquid
ratio determines the rheological properties of the fat. For example, at refrigeration
temperature, butter has a higher solids content than does a tub margarine. Hence,
the latter product is more easily spread (37).
As crystallization proceeds, the growing crystals impinge to form aggregates.
A network results, in which both the solid and liquid phases may be regarded as
continuous. Formation of the network greatly increases the firmness of the fat.
As a liquid fat is cooled, crystallization begins. There are two parts to the crys-
tallization: (1) nucleation and (2) growth. In a bulk fat, nucleation occurs at the
surfaces of impurities, a phenomenon described as heterogeneous nucleation. A
considerable degree of supercooling is necessary to initiate nucleation. Subsequent
growth of the nuclei tends to be slow in natural fats because of competitive inhibi-
tion. In materials of low molecular weight, impurities are rejected at the face of the
growing crystal. In fats, however, the various triglyceride species are so closely
related that the term impurity tends to lose its meaning (38).
3.2. Hydrogenation
Hydrogenation of various fats and oils is used extensively in industry but is not gen-
erally applied to butter fat (the high cost of the raw material argues against its use as
MODIFICATION OF MILKFAT 13
a feedstock). The process reduces the degree of unsaturation of the fat and increases
its melting point.
Given the criticism directed at milkfat because of its saturated nature, there
appears to be little future in increasing the degree of saturation by means of hydro-
genation. The reverse procedure, desaturation or dehydrogenation, offers more
attractive prospects.
Flavor deterioration in fat-rich milk and dairy products is mainly due to autoxi-
dative degradation of lipids. This degradation may be retarded by partial hydroge-
nation. The objective of partial hydrogenation or trace hydrogenation is a selective
saturation of the polyunsaturated fatty acids without saturation of the monounsatu-
rated fatty acids to improve the oxidative stability. Selective hydrogenation has
been studied for years in the vegetable oil industry with some success. This process
has been applied by some researchers to milkfat (Figure 1) (39).
Figure 1. Changes in the composition of fatty acids during the hydrogenation of milkfat (38).
14 BUTTER
3.3. Interesterification
Interesterification (also called ester interchange, randomization, and trans-
esterification) involves the exchange and redistribution of acyl groups among
triglycerides. This technology was initially developed as high-temperature interes-
terification in Germany during 19201930. Since about 19501960, the process has
been developed still further in the United States and Europe (39). The resultant
product exhibits the same total fatty acid composition as the starting material,
but the triglyceride composition and the physical properties are changed. Interester-
ification catalyzed by chemical catalysts or by lipases is used in the fat industry for
the manufacture of margarines, shortenings, and confectionery fat (40).
The fatty acid composition is not modified by interesterification, but there is a
significant modification of the glyceride composition (Table 14). In the untreated
milkfat, the triglycerides may be divided into two groups: triglycerides with lower
molecular weights (lower than C42) and the triglycerides with higher molecular
weights (C44C54) (39).
Interesterification offers opportunities for modifying the glyceride composition
of milkfat and recombined butter. The technique confers some positive nutritive
value to milkfat. One disadvantage of the interesterification reaction is the loss
of flavor during neutralization, interesterification, and subsequent deodorization.
Most manufacturers would support that little or no commercial interest will result
from using this process.
TABLE 14. Glyceride Composition of Nat-

ural and Interesterified Milkfat (mol %) (39).
Triglyceride Natural Interesterified
C22 0.1
C24 0.3 0.8
C26 0.1 1.4
C28 0.6 1.5
C30 0.9 1.3
C32 1.9 2.0
C34 4.4 3.0
C36 9.5 5.9
C38 13.1 9.1
C40 12.1 9.9
C42 7.7 7.6
C44 6.8 8.3
C46 7.5 10.7
C48 8.8 12.8
C50 11.2 14.2
C52 10.8 10.9
C54 4.6 0.4
Ratio C38:C50 1.17 0.64
3.4. Reduction of Cholesterol in Milkfat

One obvious area for development is in the modification of dairy products to satisfy
the changing dietary habits of consumers. The mounting health concerns are related
to intake of calories, cholesterol, and saturated fats. Concern about cholesterol in the
diet originates from the fact that high-serum cholesterol, especially the low-density
lipoproteins, is one of the risk factors associated with atherosclerosis. Dietary
intake of cholesterol may be one of the factors contributing to the elevation of ser-
um cholesterol; other dietary factors are high total fat, high saturated fat, and low
dietary fiber intake.
There have been many cholesterol-reduction technologies developed all over the
world because of high interest by the dairy industry. However, there are only a few
technologies available for technology transfer. Fractionation by thermal crystalliza-
tion, steam stripping, short-path molecular distillation, supercritical fluid extraction,
selective absorption, and crystallization using solvents or enzymatic modification
can achieve fat alterations of significance to the dairy industry.
Vacuum Steam Distillation. There has been direct application of cholesterol
removal by vacuum steam distillation, an old technology. This process is widely
used in the fats and oils industry for deodorization.
Cholesterol is a low-volatile compound, but it is more volatile than the major
triglycerides of milkfat. Superheated steam can be bubbled through the oil, heating
it indirectly, which provides for the latent heat of vaporization of the distilling com-
pounds and prevents steam condensation. Thus, the temperature and pressure can be
varied independently. When the sum of the partial vapor pressures of water vapor
and the distillates is equal to the total pressure, water vapor and the low-volatile
components, such as cholesterol and free fatty acids, distill over.
The process for cholesterol removal from anhydrous milkfat was patented by
General Mills (41). Fractionment Tirtiaux also disclosed the development of a
vacuum steam distillation system called the LAN cylinder (38). The steam distilla-
tion process (Figure 2) was commercialized, producing a 9095% cholesterol
reduction in anhydrous milkfat with a 95% yield that was reconstituted into 2%
fat fluid milk (42). The major disadvantage to the process is that it strips or removes
most all volatile flavor components from the fat. These flavor components must be
captured (i.e., vacreation) before the distillation process to attempt to reproduce the
delicate flavors so desired for reconstitution into a butter product.
Short-path Molecular Distillation. Short-path molecular distillation offered great
promise for the selective removal of cholesterol from milkfat. The process achieves
the simplest separation of desired substance from a mixture of components of high
molecular weight.
The rate of distillation at any given temperature is a function of the ratio P:M1/2,
where P is the partial pressure of the compound and M its molecular weight. Owing
to the temperature dependence of the rate of distillation, a fractionation of compo-
nents with different molecular weights can be carried out by holding the tempera-
ture constant until the more volatile constituents are removed. Figure 3 illustrates
potential capability of the process (43).
16 BUTTER
Figure 2. Schematic for the decholesterolization of milkfat (41).
The process has been applied to strip oil-soluble vitamins, sterols, and fatty acids
from fats and oils. Cholesterol has been successfully removed from anhydrous
milkfat in the range of 7090% (44, 45). Extensive studies were performed and
various temperatures and pressures were used to fractionate milkfat (46). Unfortu-
nately, the process has not proved to be economically feasible due to the low butter
fat yield when significant cholesterol was removed (Land OLakes research).
Absorption. One of the most promising technologies has been the use of cyclo-
dextrins to complex cholesterol from a mixture and then selectively separate the
cholesterolcyclodextrin complex. European researchers have pioneered almost
all of the research in this area, and patents have been issued (47, 48).
The process is based on the fact that b-cyclodextrin specifically forms an inso-
luble inclusion complex with cholesterol. b-Cyclodextrin is a cyclic oligosacchar-
ide of seven glucose units. It consists of 1,4-a-D-linked glucopyranose residues, as
shown in Figure 4. As a consequence of the C1 conformation of the glucopyranose
units, the secondary OH groups are located on the edge of the torus-like cyclodex-
trin molecule, whereas all the primary OH groups are on the other side (Figure 5)
(48). The central cavity is, therefore, hydrophobic, giving the molecule its affinity
for nonpolar molecules such as cholesterol. The radius of the cavity can accommo-
date a cholesterol molecule almost exactly, explaining the highly specific nature of
b-cyclodextrins ability to form an inclusion complex with cholesterol.
Figure 3. Distillation under vacuum (43).
17
18 BUTTER
Figure 4. b-Cyclodextrin
This process appears to provide considerable economic and practical advantages

over alternative cholesterol reduction technologies, such as steam distillation and
supercritical carbon dioxide fluid extraction. For instance, there is no absorption
of vitamins, it is a low-temperature operation, and it has a low capital cost. The
only economic concern is that the ratio of the addition of b-cyclodextrin to the cho-
lesterol removed is high, creating the potential for a high-cost process. Even so, the
Europeans have commercialized the process, and reduced-cholesterol butter and
cheese products have been introduced into the marketplace (49).
Multiple absorbants have been researched, and they include digitonin, tomatine,
sodium cholate (bile salts), and active carbon (45, 47, 50). Most will face severe
food regulatory problems. New Zealand (48) performed extensive research on
active carbon, carbon impregnated with different metal salts, and inert supports
impregnated with selected organic compounds. This technique (active carbon) is
Figure 5. Schematic of the b-cyclodextrin molecule, showing the hydrophobic cavity (48).
not promising, because it does not retain the delicate butter flavors or the carotenoid
pigments. An off-flavor develops in the milkfat. The University of California has
evaluated the use of food-grade saponins as absorbents (51). This process showed
promise, but most work has been discontinued due to U.S. regulatory prohibitions.
Solvents. The use of organic solvents, such as acetone, in the laboratory has
proven to be an effective method for removal of milkfat components, including
cholesterol. Unfortunately, this method creates regulatory and negative consumer
perceptions due to the potential of solvent residues in the natural butter/butter
fat-containing products.
Supercritical Fluid Extraction. The supercritical fluid extraction process created
extensive excitement in the mid-1980s in the research community as a preferred
process for cholesterol removal. Extensive research at various universities was
initiated to evaluate its potential, and significant publicity was generated within
the dairy industry (45, 46, 5256).
Liquid-like densities of supercritical gases result in liquid-like solvent powers;
this property and faster diffusion characteristics due to low-gas viscosity make
supercritical fluids attractive extraction agents. Solubility of substances in supercri-
tical gases derives from van der Waals molecular attractive forces and increases
with increasing pressure at a constant temperature. The temperature influences
the solution equilibria in a more complicated way than does the pressure. Com-
pounds can be selectively dissolved by changing the density of the gas, i.e., pressure
and temperature conditions.
The extraction of cholesterol into the mobile gas phase is determined by the
balance of a tripartite interaction: triglycerideCO2, triglyceridecholesterol, and
20 BUTTER
cholesterolCO2. As a minor constituent of the milkfat, cholesterol is probably

associated with the triglycerides for which it has higher affinity, namely, short-
and medium-chain triglycerides and, to some extent, long-chain unsaturated tri-
glyerides. As CO2 affinity is low for cholesterol at 200 bar and 80 C (low-gas
concentration), there would be less competition between triglycerides and CO2
for cholesterol. Those cholesterol molecules associated with short- and medium-
chain triglycerides would be eluted into the gas phase along with them at low-
gas concentrations. However, the cholesterol molecules associated with long-chain
triglycerides are not eluted at low-gas concentrations because of their larger size.
Cholesterol esters, being large molecules, would not be eluted at low-gas concen-
trations.
By the late 1980s, technologies for the removal of cholesterol with supercritical
carbon dioxide were offered by a number of companies (38). Commercialization
was never attempted by any major food company for removal of cholesterol.
Successful scale-up and commercialization was achieved by the General Foods
Corporation for removal of caffeine from coffee (45). The primary disadvantages
for the dairy industry were the low yields, low cholesterol removal, and the very
high capital and operating costs of the equipment.
Enzymatic. Biological procedures for cholesterol removal make use of micro-
organisms that produce enzymes to convert cholesterol into innocuous compounds.
Several enzymatic systems are being investigated in different countries of the
world. Most systems use a cholesterol reductase that converts the cholesterol into
coprostanol and coprosterol (52, 57). These converted compounds are very poorly
absorbed by the digestive system and pass through intact. Several investigators have
isolated and characterized Eubacteria able to convert cholesterol into coprostanol
from rat, baboon, and human feces. Leaves of cucumber, soybeans, corn, and beans
are known to contain similar enzymes (45, 57, 58). Lactobacillus acidophilus has
also been reported to metabolize cholesterol (51).
Once suitable enzyme systems are identified, the next step involves transfer of
the gene that codes for the enzyme into suitable micro-organisms, such as Lacto-
bacillus and Streptomyces species, for large-scale production and purification of the
enzyme. Further steps may involve attaching the enzyme onto a solid support or
adding purified enzyme in soluble form to the food systems. It is equally envisioned
that cholesterol-degrading enzymes, such as cholesterol reductase, can be geneti-
cally transplanted from one group of bacteria into lactic bacteria, which are the tra-
ditional dairy starter cultures. The cholesterol-reducing cultures could then be used
in cultured dairy products such as cheese.
The industrial scale-up of enzymatic technology is both highly complicated and
expensive. Moreover, a primary regulatory hurdle will involve demonstrating that
the end products of the cholesterolenzyme reaction and the novel compounds
formed through genetic engineering are harmless.
RegulatoryNutritional. Many of the processes for cholesterol removal will meet
with regulatory hurdles because of residues, unapproved additives, or byproduct
formation. But the real burden for the U.S. dairy industry was the 1993 Nutritional
Labeling and Education Act (54). This act created commercial prohibitions,
QUALITY CONTROL 21
because it requires that cholesterol-reduction claims cannot be applied to products

that contain more than 2 mg/g of saturated fat; butter fat is approximately 65% satu-
rated. The industry has not yet developed a cost-effective means to reduce butter fat
saturation.
For years, the nutritional community has claimed that the effect of dietary cho-
lesterol intake on serum cholesterol level was much less significant than the ratio of
total fat to saturated fat in the diet (59). The general public is becoming aware of
this and has consequently reduced its demand for low-cholesterol foods.
4. QUALITY CONTROL
In the production of all butter-fat-containing foods, quality control is essential to

ensure shelf life, safety, and the foods appearance, flavor, and texture. The key
to the final product is the quality of the raw materials used. The handling of raw
milk is of particular importance. Careful attention is paid to temperature control
in raw-milk-handling systems and, naturally, to the cleaning of all equipment
used for storage and transport. Raw milk is held refrigerated to less than 5 C on
the farms, and although this has eliminated the growth of a number of organisms,
others (notably psychrotrophs) can still multiply under these conditions. Unfortu-
nately, some of the lipolytic and proteolytic enzymes produced by these classes
of organisms are heat stable, even surviving ultrahigh temperature (UHT) treat-
ment. Today, a shelf life of many months is expected of a number of UHT-treated
products, thus the presence of lipolytic and proteolytic enzymes can be disastrous.
The only way to avoid this problem is to ensure that the numbers of organisms are
kept to an absolute minimum during all stages of the collection and manufacturing
process.
Some quality problems can be eliminated. For example, certain cattle feeds pro-
duce undesirable volatile flavors, which tend to concentrate in the butter fat portion
of the milk. Historically, the industry adopted steam stripping of cream for butter-
making to reduce the intensity of these flavors. The equipment universally used is
the vacreator (Figure 6), which is designed to pass as much as 0.30 kg steam to each
kilogram of cream. The design of the vacreator evolved during the 1930s and
1960s, culminating in the development of the Vac 25 and 16 models (59).
Vacreation accomplishes a number of tasks in one operation. Pasteurization of
the cream for buttermaking is perhaps the most important of these, followed by
steam stripping of off odors. Other functions include destruction of natural milk
lipases and development of a slightly nutty or cooked flavor in the resulting
butter.
4.1. Hazard Analysis and Critical Control Points (HACCP)

The World Health Organization (WHO) recognized HACCP as an effective rational
means to ensure food safety from the farm to consumer. The Food and Drug
Administration (FDA) and the U.S. Department of Agriculture (USDA) are moving
22 BUTTER
Figure 6. Diagram of the Vac 25/16 vacreator (59).
toward the adoption of HACCP systems to ensure the safety of foods sold in the
United States.
The HACCP program is a management tool that provides a logical and cost-
effective basis for better decision making with respect to dairy product safety.
One of the key advantages of the HACCP concept is that it enables a dairy food
manufacturing company to move away from a philosophy of control based on test-
ing to a preventive approach that identifies and controls potential hazards in the
manufacturing environment.
4.2. Composition Control

Composition control of the butter fat content has received much attention in the
ever present drive to maximize returns and create highly consistent products. As
noted, multiple technologies are under development to standardize and simplify
the processes.
4.3. Grading, Standards, and Definition

Standards for butter in the United States were established by an act of Congress and
are supported by USDA standards for grades of butter. In the revised standards, the
following definitions apply: butter refers to the food product usually known as but-
ter, which is made exclusively from milk, cream, or both, with or without common
salt, and with or without additional coloring matter. The milkfat content of butter is
not less than 80% by weight, allowing for all tolerances. Cream refers to the cream
QUALITY CONTROL 23
separated from milk produced by healthy cows. Cream is pasteurized at a tempera-

ture of not less than 73.9 C for not less than 30 min, or it can be pasteurized at a
temperature of not less than 89 C for not less than 15 s. There are other approved
methods of pasteurization that give equivalent results (60).
The flavor of the cream may be enhanced by culturing, adding food-grade lactic
acid bacteria, or adding natural flavors obtained by distilling a fermented milk;
cream may also be added to the finished butter. In addition, color, derived from
an FDA-approved source, may be used (61).
Legal requirements for butter vary considerably in different countries. For exam-
ple, in Europe, butter must contain 82% fat, and in France, it may contain a max-
imum of 16% moisture (62). In some tropical parts of the world, milkfat is used in
nearly anhydrous form, because it is less susceptible to bacterial spoilage. This pro-
duct is known as ghee. In the Middle East and India, ghee is prepared from heated
cow or buffalo milk.
The new nutrition labeling regulations, promulgated under the Nutrition Label-
ing and Education Act of 1993 (54), mandate that only strictly defined terms be
used to make nutrient content claims. For example, the term light may only be
used on products that have been specifically formulated or altered to meet one of
two conditions: (1) if the product derives 50% or more of its calories from fat,
reduce the fat level by 50% (as compared with a reference product), and (2) if
the product derives less than 50% of its calories from fat, reduce the calorie level
by one-third (compared with a reference product). Generally, butter products derive
more than 50% of their calories from fat and, therefore, must achieve a minimum
50% fat reduction to use the term light. The term reduced when used as a nutrient
descriptor requires a formulation alteration that achieves a minimum 25% reduction
in the nutrient from a reference product (63).
When considering products to be labeled with a cholesterol claim, the following
applies:
Reduced cholesterol: the product has an allowable maximum of 2 g saturated fat

and a minimum 25% cholesterol reduction per serving.
Low cholesterol: maximum of 2 g saturated fat and a maximum of 20 mg
cholesterol per serving are allowed.
Cholesterol free: maximum of 2 g saturated fat and less than 2 mg cholesterol
are allowed per serving.
When considering a fat free or no fat claim, the new regulations require the product
to have less than 0.5 g fat per serving and no added fat unless noted (i.e., trivial
fat) (54).
There are three U.S. grades of butter: AA, A, and B. Butter is graded by first
classifying its flavor organoleptically. In addition to the overall quality of the butter
flavor itself, the standards list 17 flavor defects and the degree to which they may be
present for each grade. This grade is then lowered by defects in the workmanship
and the degree to which they are apparent. Deratings are characterized by negative
body, flavor, or salt attributes, which are fully described in the standards. Butter
24 BUTTER
TABLE 15. Standards for Anhydrous Milkfat, Anhydrous Butter Oil, and Butter Oil (65).a
Composition and Quality Anhydrous Milk Fat Anhydrous Butter Oil Butter Oil
Milk fat, minimum 99.8% m/m 99.8% m/m 99.6% m/m

Water 0.1% m/m 0.1% m/m 0.3% m/m
Free fatty acids, as oleic acid, 0.3% m/m 0.3% m/m 0.4% m/m
maximum
Peroxide value, m Eq oxygen/kg 0.3 0.3 0.6
fat, maximum
Copper, mg/kg, maximum 0.05 0.05 0.05
Iron, mg/kg, maximum 0.2 0.2 0.2
a
Taste and odor at 4045 C should be acceptable for market requirements. Texture, depending on
temperature, should be smooth and fine granules to liquid.
that does not meet the requirements for U.S. Grade B is not graded. To bear the
USDA seal, the finished product must fall within the following microbiological
specifications:
Proteolytic count not more than 100 per gram.

Yeast and mold count not more than 20 per gram.
Coliform count not more than 10 per gram.
Butter should be stored at 4.4 C or lower or at less than 17.8 C, if it is to be held

for more than 30 days (62). The International Dairy Federation (IDF) has produced
specifications for milkfat (64), which include reference to the feedstock. (These
specifications relate to the time of manufacture but are often used as purchase stan-
dards.) The highest grade, anhydrous milkfat (AMF), must be produced from fresh
milk, cream, or butter, to which no neutralizing substances have been added. It
should have a clean, bland flavor when tasted at 2025 C and a peroxide value
(PV) of less than 0.2 meq oxygen/1 kg fat. Anhydrous butter oil may be produced
from butter or cream of different ages and has no pronounced, unclean, or other
objectionable taste or flavor. The term butter oil should be used where there is
no pronounced unclean or other objectionable taste or odor. The FAO/WHO Codex
standard for milkfat is shown in Table 15 (65).
4.4. Specialized Analytical Methods

The dairy industry produces a valuable fat that has a desirable flavor and positive
consumer awareness; these attributes must be protected. Significant development
activity has occurred for rapid, simple methods to detect adulteration (66, 67).
As lack of spreadability was determined to be a major butter negative, a flurry of
research was initiated creating the need for measurement techniques (6870). Many
procedures have been used or proposed to assess the microbiological quality of
the milk or cream. Generally, microbiological tests are performed to determine
the hygiene of production and storage conditions or for safety reasons. Tests include
QUALITY CONTROL 25
total counts and counts for specific classes of micro-organisms, such as yeasts and
molds, coliforms, psychotrophs, and pathogens such as Salmonella. Rapid-screen-
ing tests based on dye reduction or direct observation using a microscope or auto-
matic total counters are also in use.
4.5. Lipase Activity

An increasing problem is lipolysis in butter fat after manufacturing, which is caused
by thermoresistant lipase enzymes that are created in the milk or cream by psycho-
trophic bacteria or by residual native lipases that survive pasteurization. Based on a
determination of the lipase activity in cream, the keeping quality of manufactured
butter in regard to lipolysis can be predicted with reasonable accuracy. A similar
prediction for sweet cream butter can be based on lipase activity in the serum phase
(71). The characteristic lipolytic flavors that can develop in milk products are pri-
marily associated with the short- and medium-chain fatty acids that are relatively
abundant in milkfat; they have lower flavor threshold values than the long-chain
fatty acids. As a result of improvements in the quality of raw milk and the standards
of processing, lipolytic rancidity is seldom present in the fat source before its use in
recombination (72).
4.6. Oxidation
The flavor of dairy products is largely determined by the fat component. Conse-
quently, it is particularly important to restrict the development of oxidized off-flavors
in the fat source before use. Oxidation is the chief mode of deterioration of fats and
a major factor in determining the shelf life of fat-containing foods (72). Unsaturated
fatty acid esters react with oxygen to form peroxides. Although flavorless them-
selves, peroxides are unstable and readily decompose to yield flavorful carbonyl
compounds. The latter are the source of the characteristic oxidized flavors that
are detectable at low concentrations. The rate of oxidation depends on the concen-
tration of dissolved oxygen, the temperature, the presence of pro-oxidants such as
copper and iron, the degree of unsaturation of the fat, and the presence of antiox-
idants that may retard the onset of oxidation. Compared with many fats, milkfat has
a good oxidative stability, because it is high in total saturates, low in polyunsatu-
rates, and contains natural antioxidants, principally a-tocopherol.
The development of oxidative rancidity in milkfat is the major determinant of
the stability of the fat on storage. Dissolved air in the milkfat can give dissolved
oxygen levels of up to 40 ppm at 30 C. In practice, the dissolved oxygen level in
the freshly processed milkfat would be about 5 ppm at 45 C, a level sufficient to
permit the development of oxidative rancidity, but if the milkfat were allowed to
equilibrate with the air, then this level could increase to 33 ppm with a consequent
increase in the rate of development of oxidative rancidity. The solubility curve for
oxygen in milkfat is a compound of the solubility curves for the liquid and solid
phases (Figure 7). Though the solubility decreases with increasing temperature
for both phases, the solubility of oxygen in the liquid phase is much higher than
for the closely packed solid phase (73).
26 BUTTER
Figure 7. The effect of temperature on the solid fat content and solubility of oxygen in milkfat
(73).
The oxygen level in the milkfat may be limited by either active or passive
actions. For passive control, processing procedures and plant design are established
to minimize air exposure. Deaeration devices (74), vacreation (60), the use of anti-
oxidants (72), effective destruction of lipases (75), and nitrogen spanning of con-
tainer headspace are examples of active control of product quality.
5. BUTTER MANUFACTURE
5.1. Milk and Cream Separation

The most basic and oldest processing method is cream separation. Ancient people
are known to have used milk freely. It is probable that they used the cream that rose
to the top of milk that had been held for some time in containers, although there is
little in ancient literature to suggest that such use was common. It is well estab-
lished that, in early times, butter was produced by churning milk.
The principal questions concerning separation in a butter manufacturing facility
are the choice of cream fat content and the choice of the separation technique, milk
separation before or after pasteurization, temperature of separation, and regulation
of the fat content. Separation of cream from milk is possible because of a difference
in specific gravity between the fat and the liquid portion, or serum. Whether separa-
tion is accomplished by gravity or centrifugal methods, the result depends on this
difference (48).
A modern dairy separator will separate virtually all fat globules larger than
0.8 mm, and because most of the milkfat is present in the form of such globules,
BUTTER MANUFACTURE 27
it is relatively easy to separate the bulk of the milkfat. Fat globules below 0.8 mm
are generally referred to as nonseparable globules (76).
The percentage of fat in the cream must be known and controlled. It influences
fat losses during churning. Knowledge of the fat content assists in yield estimations
for operational conditions in continuous manufacture. A number of satisfactory
analytical procedures are available, with the Babcock test being the most
common.
The chemical composition of the triglycerides, which make up milkfat, varies
throughout the year, depending on the stage of lactation and the cows diet. The
seasonal variation causes a cyclic change in the melting properties of the fat. In
the control of the buttermaking process and the physical properties of the finished
butter, this factor must be monitored. The term melting property is used rather than
softness or hardness, because these more correctly refer to altogether different attri-
butes of solids. A number of procedures have been used to follow the seasonal
change in the melting properties. The iodine number, refractive index, differential
scanning calorimetry, or pulsed nuclear magnetic resonance spectroscopy can be
used to prepare a melting curve. However, the expense and complexity of these
melting curve techniques precludes this approach in most quality-control situations
(77). The traditional chemical determinations for fat, saponification value, and
Polenske value are of limited value. They are scarcely relevant for quality control,
and the information they provide can be more usefully quantified by the determina-
tion of the fatty acid profile using gas chromatography.
Today, the use of stainless steel has essentially eliminated the exposure of the fat
to copper and iron. The presence of copper and, to a lesser extent, iron can catalyze
oxidative deterioration of butter during storage, particularly in the presence of salt
and a low pH.
5.2. Crystallization
The crystal structure of fat and the resulting physical properties of butter made by
both conventional and alternative processes have received considerable study.
When churned conventionally or by the continuous Fritz process for butter manu-
facture, most of the milkfat is contained within the fat globule in cream during the
cooling and crystallization process. The fat globule provides a natural limit to the
growth of fat crystals. Cooling and holding of cream is normally carried out over-
night, and thus, sufficient time exists for the crystallization process to approach
equilibrium (78).
The principles of crystallization of plastic fats in the type of equipment used for
margarine manufacture have been described (79). It is important for the butter to
develop small fat crystals that remain substantially discrete and do not form a
strong interlocking structure. Small crystals (e.g., 5 mm diameter) have a greater
total surface area than large crystals and will bind water and free liquid fat by
adsorption more effectively (78). Large crystals impart a gritty texture to the pro-
duct. When fats are cooled rapidly in a scraped-surface heat exchanger, fat crystal-
lization commences, but the fat is substantially supercooled on exiting. If
28 BUTTER
crystallization is then permitted to continue under quiescent conditions, crystals

will grow together and form a lattice structure. The product will thus be hard
and brittle and may tend to leak moisture. If, however, crystallization is permitted
to occur under agitated conditions (e.g., for 13 min in a pin worker), the formation
of small independent crystals will be favored and the product will have a fine,
smooth texture. If crystallization under agitated conditions is permitted to continue
for too long, the product will be too soft for most patting or bulk filling operations,
and it is likely to be too soft and greasy at warm room temperatures.
5.3. Neutralization
When lactic acid has developed in the raw, unpasteurized cream by microbial activ-
ity to a degree considered excessive, neutralizer may be added to return the cream
acidity to a desirable level. Sodium carbonates have been found suitable in practice
for batch neutralization. For continuous neutralization by pH control, sodium
hydroxide is more suitable. These chemicals must be food grade.
5.4. Heat Treatment

The heat treatment of cream plays a decisive role in the butter-manufacturing pro-
cess and the eventual quality of the butter. It is important that milk and cream be
handled in the gentlest possible way to avoid mechanical damage to the fat, a ser-
ious problem in continuous manufacture (Fritz process) of butter (80). Cream is
pasteurized or heat treated for the following reasons: to destroy pathogenic
micro-organisms and reduce the number of bacteria, to deactivate enzymes, to
liquify the fat for subsequent control of crystallization, and to provide partial elim-
ination of undesirable volatile flavors.
5.5. Batch Butter Manufacture

Today, batch processing is not used to any extent for the production of large quan-
tities of butter. Batch systems are still encountered in small butter plants, primarily
in less industrially developed countries. Continuous systems are more efficient and
cost-effective for large outputs; batch systems have low capital cost.
The processing of cream by a batch churn requires filling to approximately
3050% capacity at a cream temperature of 4.412.8 C. Cream temperature varies,
depending on season, the butter characteristics desired, and the desired rate of fat
inversion (Figure 8). Churning is accomplished by rotation of the churn at approxi-
mately 35 rpm until small butter granules appear. The process usually requires
about 45 min for coalescence of the fat globules and clean separation of the butter-
milk so that it can be drained (82). The granules may be washed with cold water to
remove surface buttermilk. Salt is then added with water if standardization of the
fat content is required. The butter is worked to ensure uniformity and desirable
body and texture characteristics and the rate of fat inversion.
Figure 8. Energy consumption and fat loss in buttermilk in relation to churning temperature and
heating temperature (81)., Energy consumption; , fat loss.
5.6. Continuous Butter Manufacture

Between 1930 and 1960, a number of continuous processes were developed. In the
Alfa, Alfa-Laval, New Way, and Meleshin processes, phase inversion takes place by
cooling and mechanical treatment of the concentrated cream. In the Cherry-Burrell
Goldn Flow and Creamery Package processes, phase inversion takes place during
or immediately after concentration, producing a liquid identical to melted butter,
before cooling and working. The Alfa, Alfa-Laval, and New Way processes were
unsuccessful commercially. The Meleshin process, however, was adopted success-
fully in the former U.S.S.R. The Cherry-Burrell Goldn Flow process appears to
have been the more successful of the two American processes (78).
The Fritz continuous buttermaking process, which is based on the same princi-
ples as traditional batch churning, is now the predominant process for butter man-
ufacture in most butter-producing countries. In the churning process, crystallization
of milkfat is carried out in the cream, with phase inversion and milkfat concentra-
tion taking place during the churning and draining steps. However, because of the
discovery that cream could be concentrated to a fat content equal to or greater than
that of butter, methods have been sought for converting the concentrated or plastic
cream directly into butter. Such methods would carry out the principal buttermak-
ing steps essentially in reverse order, with concentration of cream in a centrifugal
separator, followed by a phase inversion, cooling, and crystallizing of the milkfat (82).
Increased demands on the keeping qualities of butter require careful construc-
tion, operation, and cleaning of the milk- and cream-processing equipment, as
well as research to develop machines that will ensure butter production and packing
under conditions without contamination and air admixture. It has been demon-
strated that butter produced under closed conditions has a better keeping quality
than butter produced in open systems (83).
There are two classes of continuous processes in use: one using 40% cream, such
as the Fritz process (Figure 9) (81), and the other using 80% cream, such as the
30
Figure 9. Continuous butter maker (Westfalia). 1, Churning cylinder; 2, separation section (first working section); 3, squeeze-drying section; 4, second
working section; 5, injection section; 6, vacuum working section; 7, final working stage; and 8, moisture-control unit (81).
Cherry-Burrell Goldn Flow (84). As much as 85% of the butter in France is made
by the Fritz process. In this process, 40% fat cream is churned as it passes through a
cylindrical beater, all in a matter of seconds. The butter granules are fed through an
auger where the buttermilk is drained and the product is squeeze-dried to a low-
moisture content. It then passes through a second working stage where brine and
water are injected to standardize the moisture and salt content. As a result of the
efficient draining of the buttermilk, this process is suitable for the addition of lactic
acid bacteria cultures at this point. The process then becomes known as the NIZO
method when the lactic starter is injected (78). Advantages of the NIZO method
over traditional culturing are improved flavor development, improved acid values
as a result of lower pH, more flexible temperature treatment of the cream because
culturing and tempering often are accomplished concurrently, and most important,
production of sweet cream buttermilk.
The Cherry-Burrell Goldn Flow process is similar to margarine manufacture
(84). The process starts with 18.3 C cream that is pumped through a high-speed
destabilizing unit and then to a cream separator, from which a 90% fat plastic cream
is discharged. It is then vacuum pasteurized and held in agitated tanks to which col-
or, flavor, salt, and milk are added. Then this 80% fat-water emulsion, which is
maintained at 48.9 C, is cooled to 4.4 C by use of scraped surface-heat exchangers.
It then passes through a crystallizing tube and then a perforated plate that works the
butter. Before chilling, 5% nitrogen gas is injected into the emulsion. Improvements
of the processing continue to occur. It is now possible to manufacture butter from
high-fat cream (>82% milkfat) on a continuous basis (85).
Although the Meleshin process continues to be in widespread use in the former
U.S.S.R., the use of alternative continuous buttermaking processes based on high-
fat cream has declined in Western countries during the past 20 years (78). The prin-
cipal reasons for this decline appear to be economics and butter quality, particularly
when compared with the Fritz process. A Fritz manufacturing process can be
installed in existing batch churn factories with almost no modification to cream-
handling or butter-packing equipment. The churns could be retained in case the
Fritz breaks down. However, little batch plant equipment could be reused in the
alternative systems (i.e., Goldn Flow). When a completely new plant is being
bought, the alternative systems still tend to be more expensive, and operational
advantages over the Fritz system are not significant. Butter from the Fritz process
is nearly identical in its physical and flavor characteristics to batch-churned butter,
whereas butter produced by the alternative processes tends to be different (86).
These differences may be perceived as defects by the consumer, and manufacturers
have been reluctant to alter a traditional product.
There are a number of advantages that the alternative systems have over the
modern Fritz line (87). The most attractive advantage is the flexibility to produce
a wide range of products, with fat contents ranging from 30% to 95% buttervege-
table oil blends and the ability to incorporate fractionated fats (88). The alternative
processes also present the possibility of a number of operational advantages. The
use of an efficient centrifuge during the cream concentration stage can substantially
reduce fat losses in the buttermilk. The composition of the butter can be more
32 BUTTER
Figure 10. Schematic of the Pasilac-Danish IBC method (81).
accurately controlled, either by including a batch standardization step or by the use

of accurate continuous metering systems.
5.7. Cultured Butter Manufacture

There are several ways of making cultured butter from sweet cream. Pasilac-Danish
Turnkey Dairies, Ltd. developed the IBC method (Figure 10) (81). The main prin-
ciples of the IBC method are as follows. After sweet cream churning and buttermilk
drainage, a starter culture mixture is worked into the butter, which produces both
the required lowering of butter pH and, because of the diacetyl content of the starter
culture mixture, the required aroma. The starter mixture consists of two types of
starter culture: (1) Lactococcus lactis and (2) L. cremoris and L. lactis ssp. diace-
tylactis. With respect to production costs, the experience with this method shows
that, for the manufacture of mildly cultured butter, the direct costs are only about
one-third of the costs of other methods (81).
5.8. Reduced Fat Butter

Fat shortages during World War II first stimulated interest in low-fat spreads in the
United States. Oil-in-water spreads were first developed in the 1950s and 1960s.
However, they had a number of limitations, including a shelf life of only 1014
days (conventional butter shelf life is 4 months), a spongy texture, a lack of melting
properties, and an inability to withstand freezing (89). The publics interest in low-
calorie foods has motivated manufacturers to produce low-fat butter products.
Although these products can no longer be called butter (according to international
standards), they are nonetheless often called low-calorie butter, half-butter, light
butter, or a similar name. Many patents have been obtained for these products,
because emulsifying properties are needed to deal with the water content (nearly
50%) of butter-like spreads. In addition, the emulsion must often be stabilized
with additives. In some countries, a low-fat butter (40% total fat) containing vege-
table oil has been designated as Minarine, but Minarine can also be prepared using
only butter fat (62). Consumer interest in a reduction of additives in food products
and a growing awareness of the importance of proper nutrition have created a
demand for a low-fat product. It is now possible to produce a butter product based
exclusively on butter fat with a fat content of 40%, without using emulsifiers. How-
ever, it is precisely this wish for a lower fat content in spreads and in other products
that is forcing manufacturers to invest time and money in product development to
find a use for their excess butter fat.
The first reduced-fat butter (50% fat), called Light Butter, was introduced in the
United States by the Lipton Co. in the mid-1980s. The product was withdrawn due
to FDA objections of not meeting Standards of Identity for nomenclature. Also, the
product contained stabilizers not allowed for in the Standard of Identify for butter.
In the late 1980s, Ault, Inc. introduced a reduced-fat butter (39% fat) called Pure
and Simple, which contained no unusual additives (90). Unfortunately, this all-nat-
ural product had severe negatives: it had a short shelf life, experienced moisture
seepage, and lacked the highly desirable butter notes. In 1990, Land OLakes,
Inc. launched its Light Butter (52% fat), which contained emulsifiers, added
Vitamin A, and preservatives. The FDA was in the process of establishing standards
for reduced-fat products at this time and no objection was registered. The new stan-
dards were established in 1993 (63), which automatically required Land OLakes to
reformulate to a 40% butter fat content; it did so and relaunched. The product was a
success and has established dominance in the U.S. market.
Butter-like products with reduced-fat content are manufactured in several coun-
tries. Stabilizers, milk and soy proteins, sodium albumin or caseinate, fatty acids,
and other additives are used. A product is now available on a commercial scale in
the former U.S.S.R. that has the following composition: 45% milkfat, 10% nonfat
solids, and 45% moisture. It has a shelf life of 10 days at 5 C (91). Each country has
established its own standards for butter and butter fat products. Many are still devel-
oping standards for a reduced-fat butter product to meet the growing consumer
demand.
Manufacturers have experienced many problems with the production of low-fat
butter (92). Low-fat butter cannot be manufactured in conventional continuous but-
ter makers. The technology of producing low-fat butter and margarine products is
similar to that of ordinary margarine production, and it has nothing in common with
modern butter (Fritz process) production (Figure 11). The conditions are, of course,
more critical for products that contain only 40% fat. These low-calorie water-in-fat
emulsions have such a dense package of water droplets that unwanted phase inver-
sion during processing or structural weak points in the product can occur, which
may, for example, severely limit the microbiological shelf life. The scraped-surface
heat exchanger type of machine is preferred for production of low-fat products.
34 BUTTER
Figure 11. Schematic for the production of low-fat butter and spreads (81).
As a result of the close packing of the aqueous-phase droplets, the composition

of the water phase is critical. Protein concentrates, caseinate, gelling agents, and
special emulsifiers have been recommended to simplify the emulsification and to
stabilize the end product (9398). For manufacture, the basic material for produc-
tion is a mix that is chemically identical to the end product. This mix consists of
milkfat in the form of butter, butter oil, and fractionated butter oil or cream, in many
cases, it also has milk solids, milk concentrates (including dissolved milk powder
and caseinates), and emulsifiers (see Figure 10) (81). The fat mix (i.e., butter, butter
oil, etc.) is melted and pasteurized.
The required amount is metered into an emulsion tank. If emulsifiers are used,
they are melted and mixed with a small amount of fat before being transferred to the
Figure 12. Flow sheet of the APV Pasilac method (99).
emulsion tank and added to the fat. The required amount of water is apportioned,
nonfat milk solids are added, and the mix is pasteurized. The watermilk mix is
transferred to the emulsion tank and mixed into the fat mix. The emulsion is
then pumped to a specially designed scraped-surface cooler, where the emulsion,
under heavy mechanical treatment and rapid cooling, is supercooled and crystal-
lized, forming the water-in-fat emulsion.
Alternative methods have been developed. For example, the APV Pasilac meth-
od is widely used in Europe (Figure 12) (99). The continuous APV Pasilac method
is quite simple. Ordinary butter with a fat content of 8082% is mixed with an aqu-
eous phase to the desired fat content. The mixture is then subjected to vigorous
mechanical treatment in a special butter homogenizer, and the low-fat butter is
ready for packaging. As Figure 12 shows, the low-fat butter equipment includes
an aqueous phase plant. The production of the aqueous phase involves the following
processes:
Dissolution of one or more powders in watermilk.

Pasteurization of the solution.
Cooling of the solution to emulsification temperature.
36 BUTTER
Figure 13. Schematic of Land OLakess method for the manufacture of Light Butter.
The process makes it possible to manufacture a butter with a fat content as low as
28% (99).
Figure 13 shows the method used by Land OLakes to produce low-fat butter
(40%). The method is similar to margarine manufacture.
Fat substitutes and zero-calorie fats offer the potential to reduce the total fat con-
tent of foods. Nutrition and marketing experts predict that consumers will show the
same enthusiasm for fat substitutes as they exhibited for alternative sweeteners
(100).
Figure 14 shows a schematic of the manufacture of a no-fat spread using mod-
ified whey protein concentrate (WPC).
5.9. Physical and Organoleptic Characteristics

Consistency. Consistency has been defined as that property of the material by
which it resists permanent change of shape and is defined by the complete force
flow relation (34). This implies that the concept of consistency includes many
Figure 14. Low-fat spread using modified whey protein concentrate.
aspects and cannot be expressed by one parameter. Today, more importance is

attached to spreadability than anything else in the evaluation of the consistency
of butter. The general consensus is that butter should be spreadable at refrigerator
temperatures. There is no suitable method available to measure such a subjective
criterion as the spreadability of butter. For this reason, the firmness of butter, which
should correlate well with spreadability, was selected as the parameter to be mea-
sured. It was recommended that the use of the cone penetrometer, along with other
methods, could provide good results (101).
Flavor. One of the most important consumer attributes of butter is the pleasing
flavor. Butter flavor is made up of many volatile and nonvolatile compounds.
Researchers have identified more than 40 neutral volatiles, of which the most pro-
minent are lactones, ethyl esters, ketones, aldehydes, and free fatty acids (102). The
nonvolatiles, of which salt (sodium chloride) is the most prominent, contribute to a
38 BUTTER
balanced flavor profile. Diacetyl and dimethyl sulfide also contribute, especially in
cultured butter flavor (103).
Body and Texture. By means of appropriate qualifications of the terms body and
texture, butter graders describe the physical properties of butter that are noted by
the senses. The exact meanings of these terms have not been clearly outlined. Fre-
quently, they are used as if they had the same meaning. Certain properties such as
hardness and softness refer to the body of butter, whereas properties such as open-
ness refer to texture. However, some of the properties, such as leakiness or crumb-
liness, are confusing. Usually, most body and texture terms are used to describe a
defect, e.g., gritty, gummy, and sticky (86). Good butter should be of fine and close
texture; have a firm, waxy body; and be sufficiently plastic to be spreadable at cold
temperatures.
Color. The color of butter may vary from a light, creamy white to a dark, creamy
yellow or orange yellow. Differences in butter color are the result of variations in
the color of the butter fat, which is affected by the cows feed and season of the
year; variations in the size of the fat globule; presence or absence of salt; conditions
of working the butter; and the type and amount of natural coloring added.
Butter colorings are oil soluble and most often are natural annatto (an extraction
of the seeds of the tropical tree Bixa orellana) or natural carotenes (extractions from
various carotene-rich plants). Because they are oil soluble, colorings are added to
the cream to obtain the most uniform dispersion.
5.10. Texturization and Spreadability

The most common method to improve the spreadability of butter is to incorporate
air or nitrogen, generally increasing the volume by 33%. The product called
whipped butter is sold in a tub rather than in stick form.
The consistency of butter is determined by the percentage of solid fat present,
which is directly influenced by the fat composition, the thermal treatment given to
cream before churning, the mechanical treatment given to butter after manufacture,
and the temperature at which the butter is held (104). The European butter market
demands that butter be softer and more spreadable in winter and harder in summer.
With information on the changes in fat composition from gasliquid chromatogra-
phy analysis and the use of nuclear magnetic resonance to estimate solid fat, sui-
table tempering procedures can be selected to modify the fat composition and to
produce the most acceptable product for the consumer. A spreadable consistency
of butter can be achieved by either varying the fatty acid composition or varying
heat-step cream-ripening times and temperatures (105). There are a number of
recognized processing options available that influence butter spreadability. Exam-
ples include mechanical treatment (texturization) (104, 106109), temperature pro-
filing of the cream (Alnarping) (85), blending winter and summer butters (107),
fractionation of the anhydrous milkfat (37, 110113), interesterification of the fatty
acids (105, 114), the diet of the cow (38, 115117), and cream ripening (118).
When blending butters, it is important to understand milkfat melting and solidifica-
tion curves (Figure 15).
Figure 15. Typical milkfat melting and solidification curves obtained by differential scanning
calorimetry (107).
Mechanical treatment, or working, involves physically disrupting the three-

dimensional fat crystal network and breaking the bonds between the crystals. It
is essential that this mechanical treatment is applied to butter in which the crystal-
lization has been totally completed (usually 7 days after churning). The primary
crystal structure is strong, and once it is destroyed by kneading it does not reform
easily. A new, secondary structure is then formed. This structure is weak and
reforms quickly after having been destroyed by kneading.
In the Netherlands, bulk butter is often processed in a Mikrofix homogenizer to
give it more plastic structure before it is repackaged for retailing (109).
The first published cream tempering process (1937) was developed by the
Alnarp Dairy Institute in Sweden, which lent its name to this and similar processes
(107). In the original Alnarp process, pasteurized cream was cooled to 8 C for 2 h,
warmed to 19 C for 2 h, and then cooled to 16 C and held until churning. Modifi-
cations allowed for initial cooling of the cream to initiate nucleation of the crystals
at a temperature well below the main solidification temperature ranges (see
Figure 14). The cream was then warmed to a temperature several degrees above
that of the lower melting peak temperature to encourage a fractionation process
in which high-melting glycerides crystallized and low-melting glycerides melted
(Figure 16).
Another alternative to increasing butter spreadability is providing cows with spe-
cific feeds to change the fatty acid composition. The main opportunity for varying
fat composition comes in the indoor feeding period, when cows receive some type
of concentrate in addition to silage. If, for example, free soy oil is added to the diet,
its constituent 18:2 and 18:3 fatty acids are converted in the rumen largely to 18:0
and some 1118:1-trans (vaccenic acid). The gut wall and mammary gland contain
a desaturase, which converts 18:0 to 11-cis-18:1. The milkfat resulting from such a
diet may contain up to 60% of C18 acids, with 18:1 (cis and trans) assuming the
40 BUTTER
Figure 16. Effect of Alnarp-type treatment on firmness of winter butter (107).
dominant position. The milkfat contains a relatively low proportion of 16:0, 14:0,
12:0, 10:0, and 8:0, because of the depression of de novo synthesis within the gland.
The weight proportion of 4:0 remains fairly constant (which implies that the molar
proportion may increase slightly) and that of 6:0 decreases only slightly, both facts
indicating that 4:0 and 6:0 are synthesized by a route that is relatively much less
significant (38, 114, 116, 117). As a result of the high unsaturated C18 fatty acid
content, this type of milkfat has a lower melting spectrum. The outcome is a butter
that has reasonable spreadability at refrigeration temperatures (114).
Feeding protected unsaturated fat usually leads to an increase in the proportion
and yield of milkfat. However, a major cost is, in this instance, associated with the
preparation of the feedstuff, involving as it does homogenization of an oilwater
protein mixture and the subsequent removal of the water by spray drying. This type
of feeding practice has little commercial application at present (116). Thus,
although the scientific knowledge to alter the fatty acid composition of milkfat is
well established, economic considerations have prevented its exploitation.
In efforts to improve the spreading properties of butter in relation to hard butter
fat, one alternative put forward is the use of softfat fractions obtained in the frac-
tionation of anhydrous milkfat. Although several practical methods of fractionation
have been presented, the use of softfat fractions in buttermaking has not become
general practice. This is evidently because fractionation in all cases significantly
raises the cost of the butter produced. In addition, a common problem has been
to find suitable uses for the hardfat fractions. Furthermore, in fractionation methods
that use solvents or additives, fractionation should be linked to fat refining, and, in
this process, butter also loses its natural food classification. In studies that have used
soft butter fat fractions, a substantial softening of the butter has been obtained; how-
ever, this butter, like normal butter, hardens as the temperature increases and again
decreases (118). As milkfat exhibits a wide melting range from about 30 C, it
may be possible to use a dry fractionation process. Suitable sizes of crystals are
developed by controlled cooling of the melt, and the crystals are separated from the
liquid phase by filtration or centrifugation. The fractionation of milkfat by melt
crystallization has been extensively studied (37). The general conclusions from
these investigations were as follows: the short-chain triglycerides and the short-
chain and unsaturated long-chain fatty acids were enriched in the liquid fraction;
the efficiency of molecular size separation in the melt crystallization process was
poor; and the flavoring compounds, pigments, Vitamin A, and cholesterol were
slightly concentrated in the liquid fraction.
Currently, dry fractionation of anhydrous milkfat is performed by two conven-
tional systemsTirtiaux and De Smet (both from Belgium)which are bulk crys-
tallization processes. The widely used Tirtiaux dry fractionation process enables
one-step or up to five-step fractionation of anhydrous butter oil at any temperature,
ranging from 50 C to 2 C (37, 110113). The milkfat fractions thus obtained can
be used as such or the fractions can be blended in various proportions for use as
ingredients in various food-fat formulations. The major shortcoming inherent in
this system is the long residence time (812 h) for nucleation and crystal growth.
Butter samples made from low-melting liquid fractions and from a combination
of primarily low-melting liquid fractions and a small amount of high-melting solid
fractions exhibited good spreadability at refrigerator temperature (4 C) but were
almost melted at room temperature (21 C). Butters made with a high proportion
of low-melting liquid fraction, a small proportion of high-melting solid, and a small
proportion of very high-melting solid fractions were spreadable at refrigerator tem-
perature and maintained their physical form at room temperature (Figure 17).
Figure 17. Solid fat content profiles of the control and anhydrous milkfat ( ); the low-
melting, high-melting, and 20% very high-melting milkfat ( ); and the milkfat fractions used in
the low-melting, high-melting, and 20% very high-melting butter: 30S fraction (), 13L fraction
(, bottom line), and 15S fraction (, top line). The fraction number includes the fractionation
temperature ( C) and its physical form (solid or liquid) (110).
42 BUTTER
A pourable butter, formulation, nonfractionated at room temperature has been

reported in the patent literature (119).
5.11. Anhydrous Milkfat Manufacture

The quality assurance program for manufacture of butter oil, or anhydrous milkfat
(AMF), also focuses on the quality of the raw materials. Naturally, many of the
same considerations apply to handling raw cream for AMF manufacture that apply
to butter, except that vacreation is not used. As it is stored under ambient condi-
tions, care against oxidation is essential. Oxidation is perhaps the most important
mechanism by which milkfat deteriorates in quality. As the oxidation reaction is
autocatalytic (i.e., the products of the reaction act as catalysts to promote further
reaction), the normal quality-control tests, peroxide value and free fat acidity, could
give misleading results when applied to stored butter. Methods of deaeration have
been developed that could reduce potential oxidation (74).
Milkfat is present in milk or cream as part of a stable oil-in-water emulsion. The
emulsion is stabilized as a result of the protein and phospholipid-rich milkfat glo-
bule membrane (MFGM) surrounding the milkfat. During AMF manufacture, the
aim is to break the emulsion and to separate out all of the nonfat solids and water
(Figure 18). To achieve this, the MFGM must first be disrupted mechanically or
Figure 18. Anhydrous milkfat manufacture (120).

chemically. Homogenization, an example of mechanical treatment, disrupts the

membrane, destroying the membrane layer. For chemical destruction of the mem-
brane layer, an acid such as citric acid can be added to lower the pH of milk or
cream to about 4.5 (120). The protein will precipitate, removing a component to
maintain an intact fat globule. Direct-from-cream AMF plants usually have three
separators. The first concentrates cream from 40% to about 75% fat before phase
inversion in a homogenizing device. The oil separator then separates the liberated
butter oil to about 99% purity. The oil is washed with water before the third (pol-
ishing) separator, and the final traces of moisture are removed in a dehydrator at
95 C and under a vacuum of 3550 torr. The dehydrator is usually a simple vessel,
and the butter oil is introduced either as a thin film on to the walls or as a spray, to
maximize the surface area exposed to the vacuum. Such a device will not remove
significant off-flavors, because the vapor flows, temperatures, and pressures are
inappropriate for flavor stripping. When producing AMF from butter, fresh or block
butter is softened just to a pumpable stage (approximately 50 C) and transferred
to a plate heat exchanger to increase the temperature (7080 C). The oil phase is
concentrated through separators and dried under vacuum. Some washing is possible
before the final separator removes the last traces of nonfat components (see
Figure 18).
In terms of the preparation of products and their appearance and texture, AMF
has several advantages over traditional butter. The latter is, in fact, subject to sea-
sonal variations, which affect its physical properties. The advantages of AMF are
linked to the possibility of standardizing its physical properties (by the selection
and mixture of the raw materials used in production) and the possibility of adapting
its properties using the fractionation technique. This is particularly important for
its use on an industrial level where, given the automation of production and the
standardization of the texturization stage (temperature), it is necessary to maintain
constant physical (rheological) and organoleptic properties (121).
5.12. Packaging
The objective of any packaging system is to protect the product from deleterious
environmental conditions. Many packaging systems have been developed to protect
the milkfat from biological, chemical, and mechanical deterioration. Bulk contain-
ers, such as 5-L cans and 20- and 25-L drums, are popular for packaging industrial
materials such as butter oil and AMF. These containers usually have a welded side
seam and are plain internally. Unlike edible oils, milkfat is corrosive to tinplate, so
an internal gold epoxy phenolic lacquer, such as International Paints IP 180, is
required. If the containers have a welded side seam, the internal raw steel edge of
the weld needs to be protected by applying a side-strip lacquer (i.e., a two-part
epoxy polyamine). All lacquers used on food cans should have FDA status (122).
Larger drums (i.e., 210-L nominal and 218-L maximum capacity) need more rigid
walls to withstand the greater mechanical stresses in handling. When chilled storage
is available, less rigid forms of packaging may be used. Concentrated butter is now
being retailed for domestic cooking, aided by a European Community subsidy. This
44 BUTTER
concentrated butter is normally packaged in either parchment or a foilparchment

laminate similar to that used for butter (123).
For packaging in foil, parchment, or other flexible film, the milkfat needs to be in
a plastic state, similar to that of freshly churned or reworked butter. The optimum
softness of the fat depends on the packaging system being used. Traditionally, the
United States has packaged consumer portion sizes of individually wrapped four
0.25-lb sticks in 1-lb units. Multiple sizes from 5 g to 25-g units of foil-wrapped
continentals or polyethylene molded cups, have been used in the food service
sector. For bulk handling, the plastic fat or butter can be packed into polyethy-
lene-lined fiberboard boxes. In all cases, the aim is to fill the desired quantity of
fat with the minimum giveaway. Aeration of the milkfat should also be avoided
as oxidative rancidity is the principal factor limiting the shelf life.
5.13. Storage and Transport

Storage conditions depend on the end use for the product, the packaging system,
and the desired storage time. Milkfat in hermetically sealed cans and drums is
the least affected by its storage environment. Ambient storage is commonly used
and must be to the same standards as used for other food stuffs (122). In the
European community and the United States, temperature is not a major factor,
but it can be in tropical countries, where the temperature may rise to 3540 C.
At temperatures in excess of 30 C, the milkfat deteriorates significantly more
rapidly, and there is an increased risk that the stored fat will have a stale, oxidized
off-flavor.
Transport of drums and cans may be at ambient temperature, though the storage
life may benefit from refrigeration. High humidity and wet conditions should be
avoided to minimize the risk of corrosion or mold growth on the packaging that
would entail an additional cleanup operation. The drums should be stowed away
from excessive heat and noxious chemicals. For most journeys, standard freight
containers may be used.
Milkfat in polyethylene-lined fiberboard boxes is at a far higher risk from its sto-
rage environment. As the packaging is permeable to oxygen, the product is more
prone to oxidation. Chilled storage (10 C, preferably 5 C) must be used both to
reduce the rate of oxidation and to maintain the rigidity of the pack. The humidity
of the storage area must also be controlled to prevent mold growth on the fiber-
board. As the pack is also permeable to odors, the storage area should not be shared
with other food stuffs with strong odors that could cause off-flavor absorption (e.g.,
fish, onion, garlic).
Flavor transmission and oxidation are less of a problem at temperatures below
20 C, which is preferable for long-term storage. When conserving and preserving
stocks of butter for extended periods (5 years or more), a process has been devel-
oped by which butter is placed in the refrigerated chamber or warehouse, which has
been sealed airtight. The air is evacuated and replaced with nitrogen or other inert
gas mixture so that the pressure in the chamber is equal to the exterior atmospheric
BUTTER FAT PRODUCTS 45
pressure. This process allows for extended storage without mold growth and devel-
opment of rancidity (124).
When butter has been frozen, textural characteristics may have been deleter-
iously affected. An invention to improve texture has been described in which large,
deep-frozen blocks of butter with the desired moisture content are reworked
by chipping them in a butter chipper while adding measured quantities of water.
The butter chips are then fed through a vacuum chamber into a butter churn
designed as a continuous kneading mill. The butter chips are continuously conveyed
under pressure through the kneading mill by means of a high-pressure butter pump
(125).
It is not possible to set rigid standards for the shelf life of milkfat. Shelf life
depends primarily on the acceptance quality criteria of the user and will be affected
by (1) the quality of the feedstock, (2) the packaging system, and (3) temperature.
With increasing storage time, the flavor defects are more likely to become notice-
able. Flavor does not correlate easily with peroxide value (126). At a peroxide value
of <0.6, oxidized flavors are unlikely, but if the peroxide value is >1, then some
oxidized flavor may be expected. It must be pointed out that these figures are based
on the International Dairy Federation (IDF) method (127) for measuring peroxide
value and that other methods are likely to give different results. Other grades of
milkfat defined in the IDF standard are anhydrous butter oil and butter oil. Anhy-
drous butter oil is the product obtained from butter for cream; it may be of different
ages and should have no pronounced, unclean, or other objectionable taste or odor.
Butter oil is the product obtained from butter or cream; it may also be of different
ages and should have no pronounced, unclean, or other objectionable taste or odor.
6. BUTTER FAT PRODUCTS
6.1. Butter FatVegetable Oil Blends

The first commercial development of a spread made from a combination of butter
fat and vegetable oil was in Sweden in 1963. The product, Bregott, contains 80%
fat, of which 80% is milkfat and 20% is soybean oil (128). Bregott is a margarine
according to Swedish and American food standards. Swedish scientists also devel-
oped and successfully commercialized the first reduced-fat spread in Europe in
1974. The product, Latt and Lagom, contains 3941% fat, of which 60% is milkfat
and 40% is soybean oil. It is considered to be a low-calorie margarine. Bregott is
exported to Australia, and Latt and Lagom to Japan and France (89). In Finland,
where Bregott is popular, oil from rapeseed is used (83).
Other products (under license from the Bregott patent) are Voimariini in Finland,
Bremykt in Norway, Smjorvi in Iceland, and Dairy Soft in Australia. Similar pro-
ducts are Clover in the U.K.; and Dairy Gold, Kerry Gold, and Goldn Soft in Ire-
land. This list is, however, not complete. The latter blends are high-fat products
(7582% fat), and the amount of butter fat of the total fat is about 50% (93).
Most of these products are manufactured in a churning process in a churn or a
continuous butter machine.
46 BUTTER
Figure 19. Flow diagram of the manufacture of buttervegetable oil mixtures (93).
The first steps in the manufacture of Bregott are pasteurization of the cream, fol-
lowed by cooling and temperature treatment. The cultures are the same as those
used in buttermaking. Measured quantities of cream and soybean oil are mixed
in the churn or the oil is continuously injected before churning in a continuous
butter machine. The byproduct is sour buttermilk.
The most commonly used vegetable oil is soybean oil. Products with a low
percentage of butter fat will contain not only vegetable oil but also hydrogenated
vegetable fats to achieve a good plasticity. If the minor part of the total fat is butter
fat, as in Golden Churn from the U.K., the manufacturing process is completely
different from modern butter production. In this case, the technology is analogous
with normal margarine manufacture, where some part of the fat is replaced with
butter fat. The emulsion is cooled in scraped-surface coolers (Figure 19).
Very low-fat spreads have recently been developed. The first European commer-
cialized product was made by St. Ivel and is called St. Ivels Lowest. It contains
25% butter fat and has a lower saturated fat content than sunflower margarine (129).
In the early 1980s, blends of butter and vegetable oil products appeared in the
U.S. market. The U.S. market leader was Country Morning Blend made by Land
OLakes. These blends generally were 40% butter and 60% vegetable oil for a total
fat content of 80%, within the margarine Standard of Identity and designation. With
the increasing popularity of reduced-fat (less than 80%), spreads, starting in the
mid-1980s, other blends with butter fat contents of 225% were introduced. The
80%-fat margarine blends are losing market shares to lower fat spreads and blends
(130).
A number of processes have been developed using continuous churns (97) and
alternative systems similar to the Cherry-Burrell Goldn Flow process (98). The
major disadvantage to churning, either batch or continuous, is that the resultant
buttermilk is adulterated with some vegetable fat and is less valuable than standard
buttermilk. An advantage of alternative processing systems is their ability to
accommodate easily the manufacture of reduced-fat spread blends. All butter fat
vegetable oil blends provided alternatives to butter for the consumer when concerns
of health (e.g., fat saturation) and spreadability are desired.
6.2. Ghee
By definition, ghee is a product obtained exclusively from milk or fat-enriched milk
products of various animal species by means of processes that result in the near
total removal of water and nonfat solids (similar to anhydrous milkfat) and in the
development of a characteristic flavor and texture. Even so, most ghee contains
some nonfat solids to enhance the flavor.
Typically, ghee is manufactured by heating butter to temperatures well above
those used during AMF manufacture. The high-temperature treatment of the nonfat
milk solids and milkfat leads to the development of a strong buttery flavor. How-
ever, traditional ghee, as produced in the Middle East and Asia, has a more rancid
taste due to less sophisticated methods of preparation and storage. Manufacturers in
the European Community are also producing ghee by adding ethyl butyrate to anhy-
drous milkfat (119, 131). Alternative synthetic flavors have been developed to add
ghee flavor notes to butter oil.
A synthetic mixture consists of d-G10 lactone (3 ppm), d-G12 lactone (15 ppm),
decanoic acid (5 ppm), and kenanone-2 (10 ppm). This technique is simpler, less
time-consuming, and more economical that the technique that uses powders. The
shelf life of this flavored butter oil is 2.5 months. However, the addition of synthetic
antioxidants, butylated hydroxyanisole (BHA) at the 0.02% level, enhances its shelf
life so it can compete well with conventional ghee (131).
6.3. Butter Fat as an Ingredient

Recombined Products. For recombination of milk and dairy products, the two pri-
mary ingredients required are AMF and nonfat milk solids. A range of fat sources is
available for the recombining industry, but only a few of these are in widespread
use. In most countries, anhydrous milkfat is usually the sole fat source. Numerous
products can be made by putting together the correct proportions of water, flavor,
and other ingredients as desired (e.g., sugar, emulsifier, and stabilizers) to make
sweetened condensed milk, ice cream, recombined butter, or milk (1%, 2%, or
full-fat). For these applications, AMF of the highest quality should be used to avoid
off-flavor development. The recombination of butter has been reported in detail
(132).
In response to the problems associated with the handling of unsalted butter, a
new milkfat product was developed in New Zealand to combine the superior flavor
of butter with the ease of handling of AMF. Initial shipments of this product, fresh
frozen milkfat for recombining (FFMR) were favorably received, and FFMR
quickly became established as the preferred alternative to unsalted butter (72).
48 BUTTER
Pastry, Cake, and Biscuit Products. In general, fats play several essential nutri-
tional, technological, functional, and organoleptic roles in most all-bakery applica-
tions. As a result of its physical properties, fat plays a major part in the production
of the majority of items in the pastry, cake, biscuit, and chocolate confectionery
sector; for example, in the preparation of pastry cream and in the desired appear-
ance and texture of the end product. These physical properties include, above all,
the rheological properties (consistency, plasticity, texture, etc.), and the properties
of fusion and crystallization depend on the type of fat, the temperature, and the
working conditions of the product.
The fats used in pastry and biscuit confectionery have different functions, which
are determined by their rheological properties (plasticity and texture). In pastry,
these principal functions are (1) an increase in the plasticity of the pastry (e.g.,
hard pastry with a low level of hydration) and (2) a break in the body of the pastry
(i.e., the fat makes the gluten structure discontinuous, which gives the desired
crumbliness in, for instance, biscuits) (121).
ConfectioneryLiquors and Liqueur. In chocolate confectionery and for pastry
creams, it is the physical properties linked to the fusion and the crystallization of
the fat that are essential. For milk chocolate, for coating or in bars, AMF can be
used in proportions that depend on its compatibility with cocoa butter, whose prop-
erties of hardness and rapid fusion at 35 C cannot be altered. Thus it is currently
accepted that AMF with high fusion levels obtained by the fractionation technique
can be used. In general, milkfat has an interesting characteristic: it inhibits the
appearance of fat bloom (133).
For pastry creams, the ideal is an AMF, which causes rapid melting in the mouth.
Depending on the type of pastry cream, a wider choice of AMF can also be offered,
thanks to the fractionation technique.
Due to the low level of milkfat in dark chocolate, fat bloom is a problem with
this product. The hard fraction of milkfat (milkfat stearin) has been reported to act
as an antibloom in dark chocolate, giving the chocolate an increased shelf life.
However, the use of hardened milkfat is limited in several major chocolate produ-
cing countries (133).
Regulations for the amount of milkfat, nonfat milk solids, whole milk solids, and
total milk solid allowed in milk chocolate vary among countries. Fats other than
milkfat are allowed in milk chocolate in some countries, although different flavors
and textures may result in the chocolate.
Although the preferred source of milkfat in cream liqueurs and associated bev-
erages is undoubtedly double cream, its use may lead to problems. In particular,
cream contains calcium and other ionic materials. One solution is to wash the cream
to remove all ionic materials, but this approach is cumbersome in practice. The pre-
ferred approach is to use anhydrous butter fat as the starting material.
Other Uses. The use of butter or anhydrous milkfat requires more added emul-
sifiers in ice creams and ice milks, because the naturally occurring milkfat emulsion
will have been destroyed in the manufacturing process. Milkfat is also used in fresh
cream, frozen cream, dry cream, and plastic cream. Ice creams contain a high level
of milkfat, and its manufacture uses substantial quantities of milkfat worldwide.
6.4. Butter Fat Powders

The main purposes of transforming fats into powdered forms are to increase
their microbial stability and to enhance their handling and functional properties.
Powdered fats are available in two principal forms. The first form is feasible
only if the fat contains sufficiently large amounts of high-melting compounds. If
a fine jet of molten fat is sprayed into an ambient atmosphere, it will set as fine
discrete particles. The second form embodies the use of a carrier. This form is
necessary for oils and fats, such as milkfat, that contain appreciable amounts
of low-melting triglycerides. The carrier can be added to the fat before or after
spraying.
Dry creams are commonly produced as an ingredient for many applications.
They consist of at least 40% butter fat, but can range up to 70% fat, 2257% nonfat
milk solids, and 0.55% moisture.
The Commission for Dried and Condensed Milk of the International Dairy Fed-
eration (1962) proposed to designate cream powders of 60% or more milkfat as
butter powder. Later, it was suggested that a minimum of 80% milkfat be
used for butter powder (134). Butter powder composition can vary in the amounts
and types of the nonfat constituents, depending on the application. Powders can
include nonfat milk solids, sodium caseinate, sodium citrate, sodium aluminum sili-
cate, sucrose, and lactose (134). Other possible minor constituents are antioxidants,
emulsifiers, flow agents, and stabilizers. The objective is to create a stable, fluffy,
free-flowing, nongreasy, and loose creamy powder. Butter powder containing 80
85% fat could not possibly meet the legal and organoleptic texture requirements of
butter when reconstituted with water; therefore, it is not in market competition for
conventional butter use. The potential applications for high-fat powdered products
include butter-like spreads, coffee or beverage whiteners, soup, sauce and dessert
creamers, convenience ice cream, scrambled eggs and omelettes, pudding and pan-
cake mixes, and aerating fats.
6.5. Specialty Butter Fat Products

Flavored Butter. Flavored butters (garlic, onion, pepper, lemon, etc.) have been
successfully mass marketed in Europe (135), but this success has not translated to
the U.S. market. There exists a limited specialty market in upscale stores and
delicatessens and certain food service applications. Manufacture is quite simple.
Creating the desired flavor blends remains an art and skill.
Hypoallergenic Butter. A U.S. patent was granted in 1992 for the manufacture of
hypoallergenic butter (136). The patent has limited claims. The product is a sterile
butter-like product made from anhydrous milkfat; it contains no nonfat solids
(99.9% free of NFS).
Butter Flavors. Technologies for the hydrolysis of butter fat to produce and con-
centrate the free fatty acids to enhance the butter flavor of products have been avail-
able for decades. More recently, biotechnologists have developed methods for
producing a variety of fairly pure enzymes, economically and in large quantities.
The increased availability of lipases (glycerol ester hydrolases) from microbial
50 BUTTER
sources has made it possible for researchers to employ the catalytic properties of
these enzymes in innovative ways. One application in which the use of lipases
has become well established is the production of lipolyzed flavors from feedstocks
of natural origin.
Immobilization of lipases on hydrophobic supports has the potential to (1) pre-
serve, and in some cases enhance, the activity of lipases over their free counter-
parts; (2) increase their thermal stability; (3) avoid contamination of the lipase-
modified product with residual activity; (4) increase system productivity per unit
of lipase employed; and (5) permit the development of continuous processes. As
the affinity of lipases for hydrophobic interfaces constitutes an essential element
of the mechanism by which these enzymes act, a promising reactor configuration
for the use of immobilized lipases consists of a bundle of hollow fibers made from a
microporous hydrophobic polymer (137).
Extended-life Creams. Extended-life creams are produced using normal separa-
tion techniques but involve a high-temperature, single- or double-heat treatment
(95135 C). The temperatures employed render the product almost sterile. Any sur-
viving bacteria tend to be spore forming types. Packaging is usually carried out on
aseptic machines or nonaseptic machines modified with, for example, H2O2 spray
and ultraviolet lights (76).
Short-life Creams. For short-life creams, the shelf life depends on a low-
bacteriological-count milk with good plant hygiene. Heat treatment tends to be
in the region of 7590 C with 330 s hold, followed by cooling to below 10 C.
Final cooling to below 5 C is normally carried out in aging tanks or in the retail
container in the cold store. Shelf life can be up to 12 days (76).
Ultrahigh Temperature Creams. Heat treatment for ultrahigh temperature (UHT)
creams is produced by indirect or direct heating to 135140 C with 1 4 s hold
before cooling to ambient temperature. Aseptic packaging is essential. As this pro-
duct is designed for long shelf life (34 months), formation of a cream plug or fat
rise in the container must be avoided. Hence, all UHT creams, including whipping
cream, must be homogenized. Homogenization can be carried out either upstream
or downstream. If carried out downstream, an aseptic homogenizer must be
employed (76).
Decholesterolized Milkfat. In the 1980s, there was significant research and mar-
ket activity in developing decholesterolized milkfat. All this activity was for
naught, for the hypothesis of creating a healthier fat (for butter or milk or other
dairy product) was not sound. The nutrition community had long recognized that
the link between dietary cholesterol and serum cholesterol was weak and that the
ratio of total fat-saturated fat had a greater impact on health. In addition, the FDA
issued new standards in 1993 (63) that effectively negated the value of decholester-
ing milkfat. The new law required that, to be called low cholesterol, the fat must
contain no more than 2 g of saturation fat per serving. Butter fat is approximately
65% saturated. As the technology to desaturate milkfat is not cost-effective, decho-
lesterization has no economic value.
Desaturated Milkfat. In addition to chemical and enzymatic means of desatura-
tion, there have been extensive studies on feeding cows specific diets to change
butter fat saturation as well as increasing the ratio of potentially desirable

fatty acids (38, 115, 116). In general, good progress in the understanding of
rumen physiology, digestion, and function has occurred, but economic
potential remains unacceptable. The most promising technologies are the use of
protected fats in a feeding regimen. These fats are protected in a way that they
pass through the rumen (point of fat hydrogenation) into the remaining digestive
system for absorption and subsequently into the mammary glands. Unsaturated
fats that are fed to cows have a great opportunity to remain unsaturated as
they are synthesized into milkfat. Biotechnology may offer alternatives in the
modification of edible fats and milkfat. Research has led to new methods of
lipolysis and esterification, but the developments are still at the laboratory level.
Nevertheless, commercial application may emerge from these interesting areas of
research.
Nutraceauticals and Healthy Fats. Almost all dairy products tend to be excellent
carriers of specialized nutrients (vitamins, minerals, specialized cultures, and
micronutrients), thus there is potential for fortification to enhance the natural nutri-
tive properties of dairy products, creating nutraceauticals or functional foods. The
use of specialized cultures such as Lactobacillus acidophilus and Lactobacillus bifi-
dus, which have generally recognized nutritive characteristics, is ideal for cultured
butter. A butter spread product using these cultures (Fittisport) has been launched in
the French market; another with lower fat content has been introduced in the Ger-
man market (138). In addition, it has been shown that the free fatty acids of milkfat
have inhibitory effects against certain pathogens (e.g., Listeria monocytogenes)
(139).
A structured, lipid-containing dairy fat is covered by a U.S. patent (95). The
invention relates to a trans-esterification product of a mixture of fatty acids and tri-
glycerides, including milkfat, in the form of cream or butter as the main component.
The product has nutritional applications and may also be used as an enteral or par-
enteral supplement.
Nonfood Applications. The nonfood use of milkfat has been insignificant. Milk-
fat and milkfat fractions could, however, have some potential possibilities for profit-
able use, for example, in manufacture of pharmaceutical or technochemical
products. Land OLakes, in association with Amerchol, has pioneered the use of
milkfat fractions in cosmetics. The first product, called Cremoral, was launched
into the marketplace in 1993 (140). The major factor that has stimulated renewed
interest in using milkfat for technochemicals and other nonfood applications in
the United States has been the significant decline in price, especially relative to
alternative fats.
It is doubtful if any single market for milkfat can be found that will compensate
for the decrease in butter sales. Rather, it will be necessary to look for a large
number of relatively small outlets. If this is to be done, the dairy industry must
understand the detailed structure of milkfat and establish the functional properties
of its constituent fractions. This approach has been applied with considerable
success to the protein fraction of milk. It may be just as rewarding when applied
to milkfat.
52 BUTTER
7. ECONOMICS
Production and consumption of butter continues a long-declining trend. U.S. pro-

duction of butter for 20012004 is given in Figure 20 (141). A dramatic shift
occurred, starting in 1985, to the table spreads category of products (less than
80% fat) from full-fat butter, margarine, and blends (88). The spreads category
encompasses all nonStandard of Identity table spreads (i.e., 079.9% fat).
In a 1984 survey, the most important barriers to increased butter sales were listed
in the following order (88):
Price (opinion of an overwhelming majority when butter is compared with

margarine).
Health (negative consumer attitudes toward cholesterol and saturated fats are
increasing).
Poor spreadability.
Inadequate promotional spending.
Product innovation in margarine and spreads.
Legislation and regulatory restrictions.
Butter manufacture continues to serve as the safety valve for the dairy industry. It
absorbs surplus milk supply above market requirements for other dairy products.
Milk not required by the demand for these products overflows into the creamery,
is skimmed, and the cream is converted to butter. When the milk supply for other
products runs short of their demand, milk normally intended for buttermaking is
diverted into the channels where needed. Even though consumption patterns have
Million Pounds
75.2
73.4
61.2
58.9
56.7
kg 106 (lb 106)
54.4
52.2
49.8
47.5
45.3
43.1
40.8
38.5
36.3
34.0
31.7
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Graph USDA, DMN; Source USDA, NASS 2001 2002 2003 2004
Figure 20. Per capita consumption of butter and margarine in the United States, 19681993.
ECONOMICS 53
dramatically changed over the years, the butter industry never fails to take up the
slack in the relationship of supply and demand for all other dairy products.
Butter is both an intervention and a market product. To counteract growing
stocks, special uses are created within the scope of the milk market organization,
which contribute to not having too much butter in common storage. For some years
now, between 300,000 and 400,000 tons of butter have been sold annually at
reduced prices to the cheese pastry, ice cream, and chocolate industries in competi-
tion with vegetable fats (142).
Despite this and other measures, it was not possible after the milk market orga-
nization took effect to prevent an imbalance between butter production and con-
sumption. Even the U.K.s accession to the EC in 1973 did not bring about a
turn in the overall development. Although imports from New Zealand into the
new member country were cut from nearly 165,000 to 55,000 tons and although
the U.K. turned mainly to France, the Netherlands, and Germany as new supplier
countries, the rise of butter prices caused restricted consumption on the English
market. In the course of a few years, per capita consumption dropped from
8.8 kg in 1970 to 3.3 kg in 1991 (143).
Between 1977 and 1987, there was an overall decline in the per capita consump-
tion of butter of 16% in 14 European countries, a similar trend was noted in the
United States. By 1993, the per capita butter consumption increased in the United
States. Price is a strong purchase determinant, and the price of butter has signifi-
cantly decreased in the United States due to USDA price support policy shifts (89).
A peak in production surplus in the EC was reached in 1986. This was due not
only to increasing supplies but also to a notable drop in the consumption of milkfat.
The consumer turned to products with a reduced fat content. This trend applies to
almost all milk products and has substantially increased the availability of milkfat
for butter production (67, 144). Table 16 gives production data for the EU-15 and
other countries for butter, dairy spreads, and margarine blends (145).
U.S. prices received for butter by manufacturers, primary receivers, and others at
the wholesale level are based primarily on activities on the Chicago Mercantile
Exchange (CME). The Dairy Market News of the U.S. Department of Agriculture
reports Chicago Mercantile Exchange prices, which serve as reference prices for
formula pricing of butter (146).
A weekly average price for grade AA butter in July 2004, per the CME was
$1.7408 (Carlot) (147).
Spot prices on the exchange, less freight charges to Chicago, are the almost-
exclusive basis for prices received at the manufacturing plants for bulk butter. In
addition, a manufacturer may receive a premium for uniformity, size of shipment,
a special flavor characteristic, or some other characteristic. Manufacturers who sell
only bulk butter are generally pricetakers, not pricemakers.
Manufacturers who soft-print and package butter sell it to primary receivers,
grocery chains, dairies, and restaurants. Such manufacturers may, depending on
competitive conditions, receive a better return than those who sell only bulk.
Primary receivers buy butter from manufacturers at spot prices (plus possible
premiums) and sell to several types of customers. Print butter (packaged in pound
54 BUTTER
TABLE 16. Production of Butter, Dairy Spreads, Margarine, and Blended Spreads.
Country Years
EU-15 1990 1991 1994 1995 1996 1997 1998

a
Austria 35,283 36,131 36,609 36,700 38,600 39,416 39,901
Belgiumb 55,050 55,050 55,050 33,524 26,130 29,837 30,225
Denmark 28,400 20,100 80,000 80,000 56,080 53,000 53,000
Finlandc 55,700 51,800 45,300 44,700 46,628 50,000 50,000
Franced 453,934 411,410 374,813 375,250 406,350 396,477 392,759
Germany 411,300 546,500 558,000 570,000 560,000 577,200 555,000
Greeced 2,400 2,400 2,140 1,826 1,514 2,587 2,975
Irelande,f 494 249 254 5 3 3 3
Italy 79,500 80,300 80,300 80,300 80,300 80,300 80,300
Netherlands 177,794 163,304 127,991 131,954 126,094 134,500 149,039
Portugalf,g 15,000 15,986 16,300 19,400 19,400 19,400 19,400
Spain 37,500 37,500 27,000 27,000 27,000 27,000 27,000
Swedeng,h 20,649 21,211 19,700 16,400 14,500 14,500 14,500
UK 111,000 111,712 115,000 115,000 115,000 115,000 115,000
Totalg,h (13) 1,484,004 1,553,653 1,538,457 1,532,059 1,517,599 1,539,220 1,529,102
Australia 116,470 118,732 142,800 131,417 NA NA NA
Brazila 74 70 NA NA NA NA NA
Czech Republic 119,167 102,870 75,088 77,113 74,515 61,880 65,353
Hungaryj 38,500 28,911 NA 22,000 NA NA NA
Israele 4,000 4,000 NA NA NA NA NA
Japank 90,900 85,750 NA NA NA NA NA
Moroccoe 13,000 13,000 NA NA 13,000 500 NA
Neth. Antilles 0 0 0 76 0 0 0
Norway 24,069 23,265 NA 13,791 24,433 1,971 NA
Switzerland 42,826 43,442 40,050 41,250 39,290 39,021 40,540
Turkey
NA not available.
a
Revised figures for 1995.
b
Revised figures.
c
Revised figures for 1994.
d
FAO estimate for 1992, 1993.
e
Estimate for 1993.
f
Estimate for 1997 and 1998.
g
Estimate for 1990, FAO figures for 1991, 1992, 1993.
h
Estimate for 199419951996.
i
Estimate for 1990, 1991, 1992.
j
FAO figures for 1991, 1992, 1993.
k
Estimate for 1992, 1993.
cartons, usually 4 quarter pounds) is sold to grocery chains and wholesalers who
supply retail food stors. Bulk butter is sold to other receivers, butter wholesalers,
food processors, and cold storage firms. These sales are based on spot prices plus
markup to cover handling, overhead, and profit.
Primary receivers of butter are both pricetakers and pricemakers. Prices they
receive (and pay) are based on spot market prices. As many of the primary receivers
REFERENCES 55
are members of the Chicago Mercantile Exchange, they can influence the spot
market price by buying and selling butter there. Nonmembers can also buy and
sell on the exchange through brokers.
All products sell on a combination of price, perception, and performance. Unfor-
tunately, butter is easily the most expensive of the yellow fats. In terms of percep-
tion, all fats are under pressure because of their caloric density. Butter suffers
further because it was labeled saturated and has a high-cholesterol content; both
properties have been the subjects of adverse comments by the nutritional and
medical community. The rise in concern for fat and cholesterol in the U.S. market
has overshadowed the concern for chemicals and preservatives.
The flavor and mouth feel of butter are greatly superior to any other yellow fat,
but its physical and rheological properties, particularly its poor spreadability at
refrigerated temperature, make butter less attractive to many consumers. The mar-
garine and spread industry can tailor its product in terms of spreadability. As noted
earlier, many advances in the ability to alter the texture and rheology of butter have
been made, but costs tend to deter manufacturers from applying the technologies
for marketplace consumption. Apparently, the consumer demand for spreadability
characteristics is inhibited by an unwillingness to pay for the convenience. To those
who demand a natural food and appreciate its delicate flavor, butter will remain the
preferred product.
REFERENCES
1. O. F. Hunziker, The Butter Industry, 3rd ed., Printing Products Corp., Chicago, Illinois,
1940, pp. 123.
2. D. Swern, ed., Baileys Industrial Oil and Fat Products, vol. 1, 4th ed., John Wiley &
3. R. G. Jensen and R. W. Clark, in N. P. Wong, ed., Fundamentals of Dairy Chemistry, 3rd
ed., Van Nostrand Reinhold Co., Inc., New York, 1988, pp. 171213.
4. W. R. Morrison, in F. D. Gunstone, ed., Topics in Liquid Chemistry, Logos Press, London,
1970, p. 51.
5. R. G. Jensen, J. Amer. Oil Chem. Soc., 50, 186192 (1973).
6. J. R. Brunner, in E. Tria and A. M. Scann, eds., Structural and Functional Aspects of
Lipoproteins in Living Systems, Academic Press, Inc., London, 1969, p. 545.
7. U. Bracco, J. Hidalgo, and H. Bohren, J. Dairy Sci., 55, 165 (1972).
8. J. L. Henderson and E. L. Jack, Oil Soap, 21, 9092 (1944).
9. E. L. Jack and J. L. Henderson, J. Dairy Sci., 28, 6578 (1945).
10. E. L. Jack and L. M. Smith, J. Dairy Sci., 39, 125 (1956).
11. G. A. Gartan and W. R. Duncan, J. Dairy Sci. Food Agri., 7, 734739 (1956).
12. Anonymous, Composition and Constants of Natural Fats and Oil by Gas-Liquid
Chromatography, Archer-Daniels-Midland, Minneapolis, Minnesota, 1961.
13. R. G. Jensen, G. W. Gander, and J. Sampugna, J. Dairy Sci., 45, 329331 (1962).
14. J. L. Moore, T. Richardson, and C. H. Amundson, J. Amer. Oil Chem. Soc., 42, 796799
(1965).
56 BUTTER
15. M. F. White and J. B. Brown, J. Amer. Oil Chem. Soc., 26, 385388 (1949).
16. J. D. Hay and W. R. Morrison, Biochem. Biophys. Acta, 202, 237 (1970).
17. J. L. Iverson, J. Eisner, and D. Firestone, J. Amer. Oil Chem. Soc., 42, 10631068 (1965).
18. A. Kuksis, Can. J. Biochem., 49, 1245 (1972).
19. A. Kuksis, J. Chromatog. Sci., 10, 53 (1972).
20. L. J. Nutter and O. S. Privett, J. Dairy Sci., 50, 1194 (1967).
21. T. W. Scott, L. J. Cook, and S. C. Mills, J. Amer. Oil Chem. Soc., 48, 358364 (1972).
22. L. F. Edmondson, R. A. Yoncoskie, N. H. Rainey, F. W. Douglas Jr., and J. Bitman,
23. R. D. Plowman et al. J. Dairy Sci., 55, 204 (1972).
24. G. D. Elsdon and P. Smith, Analyst, 52, 6366 (1927).
25. G. D. Elsdon and P. Smith, Oil Fat Ind., 4, 103105, 111 (1927).
26. K. A. Williams, Analyst, 74, 504510 (1949).
27. L. Ahren, L. Bjorck, T. Raxnikiewictz, and O. Claesson, J. Dairy Sci., 63, 741745
(1980).
28. B. W. Tharp and S. Patton, J. Dairy Sci., 43, 475 (1960).
29. J. E. Kinesella, S. Patton, and P. S. Dimick, J. Amer. Oil Chem. Soc., 44, 202205 (1967).
30. J. E. Kinsella, S. Patton, and P. S. Dimick, J. Amer. Oil Chem. Soc., 44, 449454 (1967).
31. R. C. Lawrence, J. Dairy Sci., 30, 161 (1963).
32. J. E. Langler and E. A. Day, J. Dairy Sci., 47, 1291 (1964).
33. C. R. Brewington, E. A. Caress, and D. P. Schwartz, J. Lipid Res., 11, 355 (1970).
34. V. Antila, in Proceedings of the XXI International Dairy Congress, Moscow 2, 1982,
p. 159.
35. P. W. Parodi, J. Dairy Res., 48, 131 (1981).
36. J. Kisza, K. Batura, B. Staniewski, and W. Zatorski, Brief Commun., 1(1), 352 (1982).
37. A. Boudreau and J. Arul, Int. Dairy Fed. Bull., 260, 710, 1316 (1991).
38. W. Banks, Int. Dairy Fed. Bull., 260, 36, 13, 22 (1991).
39. A. Huyghebaert, H. DeMoor, and J. Decatelle, in K. K. Rajah and K. J. Burgess, eds.,
Milk Fat, Society of Dairy Technology, Huntingdon, U.K., 1991, pp. 4461.
40. E. Frede, Int. Dairy Fed. Bull., 260, 12 (1991).
41. S. S. Marschner and J. B. Fine (to General Mills, Inc.), U.S. Patent 4,804,555, 1989.
42. Anonymous, R&D Applications, Cholesterol Reduced Fat, Prepared Foods, Chicago,
Illinois, 1989, p. 99.
43. Anonymous, Short Path Molecular Distribution, CVC Products, Inc., Rochester,
New York, 1988.
44. U. Bracco, Brit. Patent 1,559,064, 1980.
45. D. H. Hettinga, Altering Fat Composition of Dairy Products, Council Agri. Sci. Technol.,
Ames, Iowa, 1991.
46. J. Arul, A. Boudreau, J. Makhlouf, R. Tardif, and B. Grenier, J. Dairy Res., 55, 361
(1988).
47. A. R. Keen, Eur. Patent 0318326A2, 1989.
48. D. Oakenfull and G. S. Sidhu (to Sidoak), Austral. Patent Applications 54768/90 and
55112/90, 1990.
REFERENCES 57
49. F. Morel, Food Process., 1056, 1921 (1990).

50. T. J. Micich, J. Agr. Food Chem., 38, 1839 (1990).
51. Anonymous, Research Review Special Report 2, Wisconsin Milk Marketing Board,
Madison, Wisconsin, 1989.
52. Anonymous, Dairy Ind. Int., 55(6), 3738 (1990).
53. Anonymous, Food Eng., 61(2), 83 (1989).
54. H. Chen, S. J. Schwartz, and G. A. Spanos, J. Dairy Sci., 75(10), 26592669
(1992).
55. D. H. Hettinga, Designing Foods, National Academy Press, Washington, D.C., 1988,
p. 292.
56. J. Arul, A. Boudreau, J. Makhlouf, R. Tardif, and T. Bellavia, J. Amer. Oil Chem. Soc., 65,
1642 (1988).
57. S. Harlander, J. Amer. Oil Chem. Soc., 65(11), 1727 (1988).
58. D. Beitz, personal communication, 1991.
59. P. A. E. Cant, in Ref. 39, pp. 122135.
60. Code of Federal Regulations, Title 7, Part, 58, Subpart P, U.S. Government Printing
Office, Washington, D.C., 1983.
61. D. H. Hettinga, in Y. H. Hui, ed., Encyclopedia of Food Technology, vol. 1, John Wiley &
62. M. M. Chrysam, in T. H. Applewhite, ed., Baileys Industrial Oil and Fat Products, vol. 3,
John Wiley & Sons, Inc., New York, 1985.
63. R. Wood, K. Kubena, B. OBrien, S. Tseng, and G. Martin, J. Lipid Res., 34(1), 111
(1993).
64. International Dairy Federation Standard 68A, IDF, Brussels, Belgium, 1977.
65. FAO/WHO Codex Standard A2, Codex Alimentarius, Rome, Italy, 1986.
66. M. Collomb and M. Spahni, Mitteil. Gebiete Lebensmit. Untersuch. Hyg., 82(6), 615
662 (1991).
67. T. Sato, S. Kawano, and M. Iwanoto, J. Dairy Sci., 73(12), 34083413 (1990).
68. M. Lucisano, E. Casiraghi, and C. Pompei, J. Texture Stud., 20(3), 301315 (1989).
69. C. J. Strugnell, J. Food Sci. Technol., 14(2), 128129 (1990).
70. L. Forman, P. Stern, and E. Matouskova, Milchwissenschaft, 44, 761764 (1989).
71. B. Schaffer and S. Szakaly, Brief Commun., 1(1), 337 (1982).
72. R. Norris, Int. Dairy Fed. Bull., 260(5), 2325 (1991).
73. T. Richardson and M. Korycka-Dahl, in P. F. Fox, ed., Developments in Dairy Chemistry,
vol. 2, Elsevier Applied Science, London, 1983.
74. M. Schroder, Ger. Patent DF 36 23 313 A1, 1988.
75. F. Hulsemeyer, Kieler Milchwirtschaft. Forschung., 42(2), 115117 (1990).
76. J. Bird, in Ref. 39, pp. 3036.
77. B. D. Dixon, Int. Dairy Fed. Bull., 204, 35, (1986).
78. M. Kimenai, Int. Dairy Fed. Bull., 204, 16 (1986).
79. L. H. Wiedermann, J. Amer. Oil Chem. Soc., 55(11), 823 (1978).
80. M. Schweizer, Int. Dairy Fed. Bull., 204, 10 (1986).
81. P. Bjerre, in Ref. 39, pp. 6373.
58 BUTTER
82. F. H. McDowall, The Butter Makers Manual, New Zealand University Press,
Wellington, New Zealand, 1953.
83. P. Friis, Brief Commun., 1(1), 326 (1982).
84. Anonymous, Goldn Flow Continuous Buttermaking Method, Bulletin G-493, Cherry-
Burrell Corp., Chicago, Illinois, 1954.
85. M. Kawanari, K. Okamoto, Y. Honda, and M. Jukushima, J. Jpn. Soc. Food Sci. Technol.,
39(3), 227232 (1992).
86. D. H. Kleyn, Food Technol., 46(1), 118121 (1992).
87. J. Toma, J. D. Curtis, and C. Sobotor, Food Technol., 42, 9395 (1988).
88. Anonymous, Int. Dairy Fed. Bull., 170 (1984).
89. W. P. Charteris and M. K. Keogh, J. Soc. Dairy Technol., 44(1), 38 (1991).
90. I. Macnab, Modern Dairy, 68(3), 23 (1989).
91. S. Gulyayev-Zaitsev, V. Dobronos, J. Berezansky, and N. Zhirnaya, Brief Commun., 1(1),
(1982).
92. O. Gerstenberg, Dairy Ind. Int., 53(11), 28 (1988).
93. K. Anderson, Int. Dairy Fed. Bull., 260, 1718 (1991).
94. A. Pedersen, Scand. Dairy Inform., 4(3), 7475 (1990).
95. V. K. Babayan, G. L. Blackbrun, and B. R. Bistrian (to N.E. Decaness Hospital), U.S.
Patent 4,952,606, 1990.
96. G. Rosenbaum, Dairy Foods, 93(7), 13 (1992).
97. F. Graves (to Land OLakes, Inc.), U.S. Patent 4,425,370, January 10, 1984.
98. D. Antenore, D. Schmadeke, and R. Steward (to Land OLakes, Inc.), U.S. Patent
4,447,463, May 8, 1984.
99. B. Schaffer and S. Szakaly, Brief Commun., 1(1), 363 (1982).
100. A. Giles, J. Amer. Oil Chem. Soc., 65, 17081712 (1988).
101. G. Van den Berg, in Proceedings of the XXI International Dairy Congress, Moscow 2,
IDF, Brussels, Belgium, 1982, p. 153.
102. T. Siek and J. Lindsay, J. Dairy Sci., 53, 700 (1970).
103. J. E. Kinsella, Food Technol., 29(5), 8298 (1975).
104. N. Cullinane and D. Eason, Brief Commun., 1(1), 349 (1982).
105. B. K. Mortensen and K. Jansen, Brief Commun., 1(1), 334 (1982).
106. S. Berntsen, Int. Patent Application WO 92/19111 A1, 1992.
107. A. M. Fearson and D. E. Johnston, Dairy Ind. Int., 53(10), 2529 (1988).
108. A. Petersen, Int. Dairy Fed. Bull., 260, 12 (1991).
109. E. DeJong, G. Van den Berg, Voedingsmiddelenttechnologie, 19(5), 3538 (1986).
110. K. E. Kaylegian and R. C. Lindsay, J. Dairy Sci., 75(12), 33073317 (1992).
111. J. Bumbalough (to Wisconsin Milk Marketing Board, Inc.), U.S. Patent 4,839,190,
June 13, 1989.
112. F. M. Fouad, F. R. Vandevoort, W. D. Marshall, and P. G. Farrell, J. Amer. Oil Chem. Soc.,
67(12), 981988 (1990).
113. J. Makhlouf, J. Arul, A. Bordreau, P. Verret, and M. R. Sahasrabudhe, J. Can. Inst. Food
Sci. Technol., 20(4), 263245 (1987).
REFERENCES 59
114. J. B. Mickle, Tech. Bull. T-83, Oklahoma Agriculural Experiment Station, Stillwater,
Oklahoma, 1960.
115. W. Banks and W. W. Christie, Outlook Agr., 19(1), 4347 (1990).
116. R. P. Middaugh, R. J. Baer, D. P. Casper, D. J. Schingoethe, and S. W. Seas, J. Dairy Sci.,
71(12), 31793187 (1988).
117. G. A. Stegeman, R. J. Baer, D. J. Schingoethe, and D. P. Casper, J. Dairy Sci., 75(4), 962
970 (1992).
118. H. Rohm and S. Raaber, J. Food Sci., 57(3), 647650 (1992).
119. T. Landon, U.S. Patent application 20040033303, February 19, 2004.
120. K. K. Rajah and K. J. Burgess, in Ref. 39, pp. 3743.
121. R. Barts, Int. Dairy Fed. Bull., 260, 1920 (1991).
122. R. A. Wibley, in Ref. 39, pp. 136147.
123. Anonymous, Technical Guide for the Packaging of Milk and Milk Products, Document
143, 2nd ed., IDF, Brussels, Belgium, 1982.
124. R. G. Mansour, Int. Patent WO 9202 144, February 20, 1992.
125. R. Egli (to Egli AG), Int. Patent WO 89/10689 A1, 1989.
126. Proceedings of IDF Seminar, Singapore, Int. Dairy Fed. Doc., 142 (1982).
127. Anonymous, IDF Study 74, Brussels, Belgium, 1974.
128. M. Zillen, Document 116, IDF, Brussels, Belgium, 1977, pp. 14.
129. W. P. Carteris and M. K. Keogh, J. Soc. Dairy Technol., 44, 3 (1991).
130. G. Behrens, Document 224, IDF, Brussels, Belgium, 1988, pp. 89.
131. B. K. Wadhwa and M. K. Jain, Indian J. Dairy Sci., 44(6), 372374 (1992).
132. R. Spieler, Document 149, IDF, Brussels, Belgium, 1980, pp. 132135.
133. M. Gordon, Int. Dairy Fed. Bull., 260, 20 (1991).
134. P. Sitaram and S. K. Gupta, Agr. Rev., 9(2), 81104 (1988).
135. B. Guenault, Dairy Ind. Int., 57(6), 32 (1992).
136. L. S. Girsh (to Immunopath Profile, Inc.), U.S. Patent 5,112,636, May 12, 1992.
137. F. X. Malcata, C. G. Hill, and C. H. Amundson, Biotechnol. Bioeng., 39(11), 10971111
(1992).
138. G. Comi, S. DAubert, and M. Valenti, Ind. Aliment., 29(281), 358361 (1990).
139. E. J. Mann, Dairy Ind. Int., 2, 19 (1992).
140. L. Kintish, J. Soap Cosmetics Chem. Special., 69(2), 51 (1993).
141. Dairy Market News, 71, Rept 28 (July 1216, 2004).
142. U.S. Department of Commerce, Bureau of the Census, Statistical Abstracts of the United
States, 113th ed., U.S. Government Printing Office, Washington, D.C., 1993.
143. E. Hetzner and P. Jacknik, Int. Dairy Fed. Bull., 293, 19 (1993).
144. Anonymous, Trends in the U.S. Consumer Attitudes and the Supermarket, Food
Marketing Institute, Washington, D.C., 1993.
145. International Margarine association of the Countries of Europe. http://www.imace.org.
146. Economic Research Service, USDA. (2004). The Structure of Dairy Markets: Past,
Present, Future AER-757. Available: USDA.gov.
147. Chicago Mercantile Exchange, Cash Markets, July 16, 2004.
2
Canola Oil
R. Przybylski, T. Mag, N.A.M. Eskin, and B.E. McDonald
University of Manitoba,
1. INTRODUCTION
During the past two decades, the production of Brassica oilseeds has increased,
making it one of the major sources of vegetable oil in the world. The development
of canola oil (low erucic acid and low glucosinolate rapeseed oil) from the original
high erucic acid rapeseed oil (HEAR) by Canadian plant breeders produced a
premium quality oil. This improvement in the oil resulted from a reduction in erucic
acid (C22:1) to levels below 2%. In fact, this acid now contributes less than 2% of
the total composition of the fatty acids in canola oil varieties. In addition, the level
of glucosinolates in the seed has been lowered to less than 30 mmol/g, resulting in
better meal quality. In this chapter, the origin, composition, properties, and utiliza-
tion of canola and HEAR oils are discussed. In addition, utilization of both oils as
edible oils, food ingredients, and industrial applications are discussed. The term
canola is used as the name for rapeseed with substantially lowered amounts of
erucic acid and glucosinolates. The term canola is used mainly in the American
continent and in Australia, and rapeseed is used commonly in Europe and other
countries.
61
62 CANOLA OIL
2. ORIGIN
Oilseed rape species used to produce canola oil and meal are from the Brassica
genus in the Cruciferae family. They were first cultivated in India almost 4000
years ago. Large-scale planting of rapeseed was first reported in Europe in the thir-
teenth century. The Brassica species probably evolved from the same common
ancestor as wild mustard (Sinapis), radish (Raphanus), and arrugula (Eruca).
3. DEVELOPMENT OF CANOLA
Early rapeseed cultivars had high levels of erucic acid in the oil and glucosinolates
in the meal. The presence of these components in rapeseeds was a health concern.
The high levels of erucic acid were blamed for producing fatty deposits in the heart,
skeletal muscles, and adrenals of rodents as well as impairing growth. Plant breed-
ing programs initiated in Canada resulted in the identification in 1959 of Liho, a
rapeseed line containing low levels of erucic acid. A program of backcrossing
and selection was conducted to transfer the low erucic acid trait into agronomically
adapted cultivars. This led to the first low erucic acid cultivar of B. napus, Oro, in
1968 and the first low erucic acid B. rapa cultivar, Span, in 1971. Because of the
health concerns related to high levels of erucic acid, over 95% of the rapeseed
grown in Canada in 1974 were low erucic acid varieties.
Glucosinolates were also considered detrimental in rapeseed meal fed to poultry,
swine, and ruminants. The hydrolyzed products of glucosinolates, namely, isothio-
cyanates and other sulphur-containing compounds, were shown to interfere with the
uptake of iodine by the thyroid gland, contribute to liver disease, and reduce growth
and weight gain in animals. Consequently, plant breeders realized that if rapeseed
meal was to be used in animal feed, the glucosinolate content had to be reduced. A
Polish line with a low-glucosinolate trait, Bronowski, was identified by Dr. Krzy-
manski in the late 1950s. Breeding efforts to introduce this trait into low erucic acid
lines by Dr. Baldur Stefansson at the University of Manitoba resulted in the release
of the worlds first low erucic acid, low glucosinolate cultivar of B. napus, often
called the double-zero rapeseed. This was followed in 1977 by the release of the
first low erucic acid, low glucosinolate cultivar of B. rapa, Candle, by Dr. Keith
Downey of the National Research Council of Canada in Saskatoon. Approximately
80% of the total Canadian rapeseed acreage in 1980 consisted of the double-zero
cultivars. The detailed history of the development of canola is described in a
booklet titled The Story of Rapeseed in Western Canada (1).
The name canola was registered by the Western Canadian Oilseed Crushers in
1978 and subsequently transferred to the Canola Council of Canada in 1980. It
included those cultivars containing less than 5% erucic acid in the oil and 3 mg/g
aliphatic glucosinolates in the meal. In 1986, the definition of canola was amended
to B. napus and B. rapa lines with less than 2% erucic acid in the oil and less than
30 mmol/g glucosinolates in the air-dried, oil-free meal. The oil was added to the
Generally Recognized as Safe (GRAS) list of food products in the United States.
COMPOSITION 63
It was much more difficult to introduce the low erucic acid trait into the
European rapeseed lines because they were primarily of the winter type. This
extended the time required to produce each generation and crosses between spring
low erucic acid rapeseed (LEAR) cultivars and winter lines resulted in undesirable
segregates. Nevertheless, the development of European LEAR varieties was accom-
plished within 15 years. European acreage of rapeseed declined during the 1970s as
result of health concerns. In 1977, the low erucic acid trait was made mandatory in
Europe. Initially the new LEAR cultivars produced lower yields and lower oil
content compared with the traditional rapeseed cultivars. Subsequent plant breeding
overcame these problems with European production of LEAR increasing substan-
tially by 1984. The other rapeseed growing areas of the world, notably, India and
China, did not take part in the development and conversion to canola type rapeseed.
In these areas, HEAR still predominates.
Canola oil produced in Canada is obtained from genetically modified seeds of
Brassica napus and Brassica rapa (campestris). These cultivars, low in erucic
acid and glucosinolates, are different in chemical, physical, and nutritional charac-
teristics from high erucic acid rapeseed oil. Current Canadian plant breeding pro-
grams continue to focus on the development of oils with specific characteristics to
meet the consumer demands and food manufacturing practices, such as lowering
the content of saturated fatty acids and designer oils for specific applications.
4. COMPOSITION
4.1. Nature of Edible Oils and Fats

Edible oils and fats are composed primarily of triacylglycerols (TAG), ester of one
molecule of glycerol, and three molecules of fatty acids. Analysis of canola oils
showed that TAGs constituted 94.4% to 99.1% of the total lipid (2). The typical
composition of canola, rapeseed, and soybean oils is presented in Table 1.
TABLE 1. Constituents of Canola, Rapeseed, and Soybean Oils.
Component Canola Rapeseed Soybean
Triacylglycerols (%) 94.499.1 91.899.0 93.099.2

Phospholipids (%)
Crude Oil up to 2.5 up to 3.5 up to 4.0
Water-degummed up to 0.6 up to 0.8 up to 0.4
Acid-degummed up to 0.1 up to 0.2
Free Fatty Acids (%) 0.41.2 0.51.8 0.31.0
Unsaponifiables (%) 0.51.2 0.51.2 0.51.6
Tocopherols (mg/kg) 7001200 7001000 17002200
Chlorophylls (mg/kg) 550 555 Trace
Sulfur (mg/kg) 325 535 Nil
Iron (mg/kg) <2 <2 <2
Adapted from Mag (2) and Ying et al. (3).

64 CANOLA OIL
4.2. Fatty Acid Composition of Canola Oil

The stigma of the erucic acid (C22:1n - 9) in rapeseed oil has lingered despite firm
evidence that this fatty acid was more of a threat to rats than to humans. It is suffi-
cient to say that the discovery of chain shortening of erucic acid to oleic acid by
peroxisomes was one of the most fundamental breakthroughs in understanding fatty
acid metabolism in the last few decades. Once in the oleic acid form, the erucic acid
residue is as readily catabolized by mitochondria, as are palmitic and other fatty
acids (4). The reduction of erucic acid in rapeseed oil resulted in a marked increase
in octadecanoic acids, and their contribution in canola oil is around 95% of all fatty
acids present (Table 2).
Plant breeders have also developed canola oil with the linolenic acid content
reduced to 2% (5) (Table 2). The storage stability of this oil showed improvement
as compared with regular canola oil (7). Frying performance of this oil was
improved together with better storage stability of fried products such as french fries
and potato chips (8, 9). Canola has been further genetically modified to produce oil
with oleic acid content ranging from 60% to 85% (10). Field production of this oil
showed that the very high content of oleic acid was hard to reproduce. The fatty
acid composition of the field-produced oil is presented in Table 2. High oleic
TABLE 2. Comparison of Major Fatty Acids in Some Vegetable Oils (w/w %).
Fatty Acid Canola HEAR LLCAN HOCAN LTCAN GLCO LLFlax Soybean SUN
C10:0 0.1
C12:0 38.8
C14:0 0.1 0.1 0.1 4.1 0.1 0.1 0.1
C16:0 3.6 4.0 3.9 3.4 2.7 4.2 6.4 10.8 6.2
C18:0 1.5 1.0 1.3 2.5 1.6 3.7 4.1 4.0 4.7
C20:0 0.6 1.0 0.6 0.9 0.4 1.0 0.1 0.2
C22:0 0.3 0.8 0.4 0.5 0.2 0.5 0.1 0.1
C24:0 0.2 0.3 0.3 0.3 0.2 0.2 0.1 0.1
Saturated 6.3 7.1 6.6 7.7 48.1 9.9 10.9 14.9 11.3
C16:1 0.2 0.3 0.2 0.2 0.2 0.2 0.1 0.3 0.2
C18:1 61.6 14.8 61.4 77.8 32.8 24.4 16.9 23.8 20.4
C20:1 1.4 10.0 1.5 1.6 0.8 0.8 0.1 0.2
C22:1 0.2 45.1 0.1 0.1 0.5 0.1
MUFA 62.4 69.7 63.1 79.9 34.3 25.5 17.2 24.3 20.2
C18:2n6 21.7 14.1 28.1 9.8 11.3 26.1 70.1 53.3 68.8
C18:3n3 9.6 9.1 2.1 2.6 6.3 1.3 1.8 7.6
C18:3n6 1.0 37.2
PUFA 31.3 23.2 30.2 12.4 17.6 64.6 71.9 60.8 68.8
Abbreviations: LLCANLow linolenic acid canola oil; HOCANHigh oleic acid canola oil; GLCOCanola oil
with gamma linolenic acid; LLFlaxFlaxseed oil with reduced content of linolenic acid; LTCANCanola oil
with high content of lauric acid; SUNSunflower oil; MUFAMonounsaturated fatty acids; PUFA
Polyunsaturated fatty acids.
Adapted from Ackman (4), Vecchia (6), and Tso et al. (11).
COMPOSITION 65
acid canola oil resembles the composition of olive oil more closely than that of the
regular canola oil. This oil showed improved frying stability and produced better
quality fried potato chips than regular canola oil (8). Warner and Mounts (9) found
that up to 2% of linolenic acid is required in frying oils to produce positive
characteristic flavour in fried foods. This is caused by the formation of oxidation
products from linolenic acid, which are the main contributors to the fried food
flavour.
Recently, canola oil with a high content of lauric acid (39%) was developed to be
used in confectionery coatings, coffee whiteners, whipped toppings, and center
filling fats (Table 2). Further, canola oil with a content of stearic acid as high as
40% is available to be used as a replacement for hydrogenated fats in bread
and bakery markets (6). Another canola oil containing approximately 10% of
palmitic acid with better crystallization properties was developed. Canola oil for
the health food market containing up to 40% of gamma linolenic acid is also
now available (11).
4.3. Minor Fatty Acids

Minor fatty acids present in oils often differ from their acid family members by the
location of the double bond. Most of these acids are present in canola oil in the
0.010.1% range, except for C16:1n -7, which is around 0.3%. Most of these minor
fatty acids are from the n-7 series, in varying proportions to the n-9 isomers (4).
A similar series of minor fatty acids were found in B. rapa variety Candle (12).
Conjugated C18:2 fatty acids were also found in canola oils. A number of these
acids are artifacts of refining and deodorization, although some were also observed
as natural components in some oil seeds. The refining process is a source of artifact
fatty acids caused by the isomerization of one or more of the double bonds of cis
linolenic acid (13). These trans-isomers can be found in any oil containing linolenic
acid after refining and deodorization accounting for 1% or more of the parent
acid (4).
Canola oil is the only known edible oil containing one or more fatty acid, with
sulphur atom as the integral part of the molecule. The structure of the proposed
molecule of this fatty acid suggests the possible formation or presence of many
isomers (14).
In the sediment from industrial winterization, additional minor fatty acids and
alcohols with 26 to 32 carbon atoms in the chain have been found in waxes and
triacylglycerols (15). Most of these compounds are extracted from the seed coat
and can initiate sediment formation in canola oil (16).
4.4. Triacylglycerols
Triacylglycerols are the most abundant lipid class found in canola oil. The combi-
nation of fatty acids on the glycerol moiety represent the most complex mixture
with n3 possible molecular species, where n is the number of fatty acids present
in the oil.
66 CANOLA OIL
TABLE 3. Composition of Major Triglycerides of Canola Oils (%).
Triacylglyceride Canola (CO) LLCO HOCO
LnLO 7.6 1.7 1.5

LLO 8.6 11.0 1.1
LnOO 10.4 2.6 8.6
LnOP 2.1 0.5 1.1
LOO 22.5 28.4 12.7
LOP 5.7 4.2 2.2
OOO 22.4 32.8 49.5
POO 4.6 4.8 7.7
SOO 2.6 2.4 5.0
PPP 0.1 1.4 2.8
LLP 1.4 1.1 0.8
LOS 1.6 1.9 1.0
LLL 1.3 1.6 0.2
LnLL 1.4 0.0 0.3
LnLnO 1.7 0.4 0.1
Others 6.0 5.2 5.4
Abbreviations: LLCOLow linolenic canola oil; HOCOHigh oleic canola oil;

LnLinolenic; LLinoleic; OOleic; SStearic; PPalmitic; OthersGroup of
15 triacylglycerols with contribution below 1% each.
Adapted from Kallio and Currie (18).
The TAG molecular species profile represents a key to understanding the physi-
cal characteristics of the oil and is a unique method of identification (17). The posi-
tion of fatty acids on the glycerol molecule was originally found for HEAR oil to be
based on saturation. Long-chain (C20:0 -C24:0) saturated fatty acids are found
mostly in the sn-1 and sn-3 positions, whereas the octadecenoic (C18) acids, espe-
cially linoleic and linolenic, are primarily in the sn-2 position (18, 19). The
composition of triacylglycerols in modified canola oils is presented in Table 3.
As can easily be predicted, in high oleic acid canola oil, the most abundant triacyl-
glycerols was triolein, whereas in regular canola oil, four triacylglycerols were
detected in almost equal amounts, namely, olein-dilinolein, linolenin-dilinolein,
triolein, and linoleic-diolein.
Jaky and Kurnik (20) investigated the concentration of linoleic acid in the sn-1,3
and sn-2 positions. They found that in HEAR oil, at least 95% of linoleic acid was
concentrated in the sn-2 position, whereas in canola oil, only 54% was in this posi-
tion. The increased amounts of linoleic acid in canola oil were incorporated into the
sn -1,3 position to replace erucic acid. Kallio and Currie (18) found that triacylgly-
cerols with acyl carbon number (ACN) 54 and two double bonds consisted of
acylglycerols where stearic acid was present predominantly in the sn-2 position.
Acylglycerols with saturated acids in this position usually have a higher melting
point, poor solubility, and can cause problems with digestibility. Additionally,
high melting acylglycerols can stimulate/initiate sediment formation and affect
clarity of the oil (21).
COMPOSITION 67
TABLE 4. Composition of Phospholipids in Canola Oil During Processing (%).
Oil Sample Phosphorus (mg/kg) PC PE PI PA PS
Solvent 529.0 31.2 18.8 19.8 21.6 3.1

Expeller 242.3 34.3 16.1 18.7 20.3 4.5
Degummed 12.2 2.8 10.8 28.9 38.4 14.6
Abbreviations: PCPhosphatidyl choline; PEPhosphatidyl ethanolamine; PIPhosphatidyl inositol;

PAPhosphatidic acid; PSPhosphatidyl serine.
Adapted from Przybylski and Eskin (23).
4.5. Polar Lipids

Sosulski et al. (22) examined the polar lipids (PL), phospholipids, and glycolipids
in several rapeseed cultivars, including a low erucic acid winter cultivar grown in
Poland and found that phospholipids accounted for the major portion (3.6%) of total
polar lipids, whereas glycolipids contributed only 0.9%. A more recent study by
Przybylski and Eskin (23) reported changes in phospholipids during canola oil pro-
cessing, as shown in Table 4.
Significant amounts of phosphatidic acid (PA) formed during processing indicate
hydrolysis of other phospholipids. This can be attributed to the effect of phospho-
lipases and hydrothermal treatment during the conditioning of flaked seeds. Cmolik
et al. (24) observed an increase in phospholipid amounts from 0.5% to 15% during
conditioning of seed flakes. It was reported that hydratable phospholipids such as
phosphatidylcholine (PC) and phosphatidylethanolamine (PE) assist the removal of
nonhydratable phospholipids. Phosphatidylinositol (PI) and PA are considered non-
hydratable phospholipids and are difficult to remove during degumming. Phospho-
ric acid was found to be the most effective degumming agent for reducing the levels
of lysophosphatidylethanolamine. Application of acids for degumming also
removes the majority of iron from the oil. In practice, canola oil phosphatides
are removed to concentrations below 0.1% using an aqueous solution of citric
acid and water (25). Lecithins obtained by degumming of canola oil using only
water formed more stable oil in water emulsions compared with lecithins obtained
from acid degumming (26).
Sosulski et al. (22) and Smiles et al. (27) examined the fatty acid composition of
the individual phospholipids in the LEAR variety from winter rapeseed cultivars
(Table 5).
TABLE 5. Fatty Acid Composition of Phospholipids (w/w%).
Phospholipid 16:0 16:1 18:0 18:1 18:2 18:3 20:1
Phosphatidyl Choline 8.7 0.8 1.2 55.8 30.9 1.9 0.5

Phosphatidyl Inositol 21.8 0.8 1.9 33.6 38.1 3.6
Phosphatidyl Ethanolamine 17.7 1.8 2.0 47.7 27.3 2.7 0.5
Adapted from Sosulski et al. (22) and Smiles et al. (27).

68 CANOLA OIL
Phosphatidylcholine contained the highest amount of unsaturated fatty acids,

mostly oleic and linoleic acids. The other two principal phospholipids were rich
in palmitic, linoleic, and linolenic acids. The presence of highly unsaturated fatty
acids in phospholipids is important as they are prone to oxidation and can cause
accelerated deterioration of the oil. It was also reported that phospholipids have
a tendency to complex heavy metals, and these complexes are a stable form of
catalyst, which can initiate and stimulate oxidation (28).
4.6. Tocopherols
The main components of unsaponifiables in vegetable oils are tocopherols and
sterols, present in different amounts. Tocopherols are recognized as very efficient
natural antioxidants and their amount in the plant is probably governed by the content
of unsaturated fatty acids. Canola oil contains relatively high amounts of tocopher-
ols. Isomers of tocopherols have different antioxidative activity in vitro and in vivo.
In the food system, the antioxidant activity of the tocopherol isomers decreases in
the following order: g > d > b > a (29). Tocopherols are about 250 times more
effective than butylated hydroxytoluene (BHT) (30). Lipid peroxy radicals react
with tocopherols several magnitudes faster than with other lipids. A single molecule
of tocopherol can protect about 103 to 106 molecules of polyunsaturated fatty acids
in the plant and animal cells. This explains why the ratio of tocopherols to PUFA in
the cells is usually 1:500 to provide sufficient protection against oxidation (31).
These components are also effective as singlet oxygen quenchers, but they are
less potent than carotenoids. A single molecule of tocopherol can react with up
to 120 molecules of singlet oxygen (32). The high potency of tocopherols as anti-
oxidants and quenchers of singlet oxygen is based on their ability to be transformed
back from the oxidized form into the active structure by other molecules such as
ascorbic acid and glutathione (33). Meanwhile, tocotrienols are not present in
canola oil. However, plastochromanol-8, which is a derivative of g-tocotrienol,
TABLE 6. Tocopherols Content in Selected Vegetable Oils (mg/kg).
Oil a b g d P-8 Total
HEAR 268 426 97 790

Canola 272 423 75 770
LLCanola 150 314 7 47 517
HOCanola 226 202 3 42 473
HOLLCanola 286 607 8 83 983
Soybean 116 34 737 275 1162
Sunflower 613 17 19 649
Corn 134 18 412 39 603
LLFlax 26 213 9 130 377
Abbreviations: HEARHigh erucic acid rapeseed; LLCanolaCanola oil with low

content of linolenic acid; HOCanolaCanola oil with high content of oleic acid;
LLFlaxFlax oil with low content of linolenic acid; P-8Plastochromanol-8.
Adapted from Zambiazi (34) and Normand et al. (35).
COMPOSITION 69
but having a longer side chain, was detected in canola and flax oils, and its anti-
oxidative activity was found to be similar to a-tocopherol (34).
The composition of tocopherols in some common vegetable oils compared with
canola oil is summarized in Table 6. Canola oil contains mostly a- and g-tocopher-
ols, usually in a 1:2 ratio. The content of tocopherols in refined, bleached, and deo-
dorized (RBD) oils is affected by processing, mainly extraction, refining, and
deodorization. The lowest content of tocopherols was found in cold-pressed canola
oil. However, when the temperature of pressing was increased, the amount of toco-
pherols in oil doubled. Solvent-extracted oils contain higher amounts of tocopherols
than cold-pressed oils, and the content of tocopherols was similar to oils from hot
pressing. The largest portion of these compounds is removed during the deodoriza-
tion process (36). However, tocopherols in the distillate may be recovered and used
in nutraceutical applications.
4.7. Sterols
Sterols are present in canola oil in two forms in equal amounts, free sterols and
esterified sterols (19, 37). The fatty acid composition of the esterified sterol fraction
in canola oil is shown in Table 7.
The fatty acid distribution in esterified sterols differs from that found for canola
oil. In the sterol esters, higher levels of palmitic and stearic acids were observed.
All three major sterols were equally distributed in esterified and free sterol fractions
in canola oil. Twice the amount of brassicasterol was found in free sterols than in
esterified sterols. The total amount of sterols in rapeseed and canola oils ranges
from 0.7% to 1.0%. The composition of major sterols in common vegetable oils
is presented in Table 8.
Brassicasterol is the major and unique sterol present in rapeseed and canola oils.
This sterol is often used to determine adulteration of other oils with rapeseed/canola
oil (4, 39). Sterols are also affected by processing, with about 40% of these
TABLE 7. Fatty Acid Composition of Esterified Sterols

in Canola Oil.
Contribution (%)
Fatty Acid Sterol Esters Canola Oil
C14:0 3.1 0.1

C16:0 17.5 3.6
C18:0 18.4 1.5
C18:1 30.9 60.2
C18:2 20.5 21.6
C18:3 7.6 9.6
C20:0 0.8 0.4
C22:1 1.2 0.2
Adapted from Gordon and Miller (38).

70 CANOLA OIL
TABLE 8. Proportions of Major Sterols in Selected Vegetable Oils (%).
Sterol HEAR CAN LLCAN HOCAN HOLLCAN SOY SUN Corn
Cholesterol 0.4 0.1 0.1 0.1 0.1 0.3 0.1 0.1

Brassicasterol 13.2 13.8 12.2 10.8 16.2
Campesterol 34.4 27.6 31.2 33.9 28.8 18.1 7.5 17.2
Stigmasterol 0.3 0.5 0.2 0.8 0.1 15.2 7.5 6.3
bSitosterol 47.9 52.3 51.3 48.7 50.9 54.1 58.2 60.3
5-Avenasterol 2.1 1.9 1.9 1.8 2.1 2.5 4.0 10.5
7-Avenasterol 1.6 1.1 1.1 1.9 0.8 2.0 4.0 1.1
7-Stigmasterol 2.1 2.3 2.1 2.1 2.3 1.4 7.1 1.8
Total (g/kg) 8.8 6.9 6.3 7.1 6.9 4.6 4.1 9.7
Esterified (g/kg) 4.5 4.2 4.0 4.4 4.2 5.8 2.1 5.6
Abbreviations: HEARHigh erucic rapeseed oil; CANCanola oil; LLCANLow linolenic canola oil;
HOCANHigh oleic canola oil; HOLLCANHigh oleic low linolenic canola oil; SOYSoybean oil;
SUNSunflower oil.
Adapted from Ackman (4), Strocchi (39), Zambiazi (34), and Gordon and Miller (38).
components removed from the oil during deodorization. Refining changes the content
of sterols where isomerization of these compounds is one of the processes (40, 41).
The chemical structure of phytosterols is similar to that of cholesterol so that
these compounds may be involved in oxidative reactions. Przybylski and Eskin
(42) found some oxidation products formed from plant sterols during storage of
fried food products. Similar oxidation products were found in soybean oil and
wheat flour (43). In light of health concerns associated with cholesterol oxidation
products, potential health risks of phytosterol oxidation products are now receiving
serious attention.
4.8. Pigments
Pigments present in canola impart undesirable color to the oil. They also promote
photo-oxidation as well as inhibit catalysts used for hydrogenation. Chlorophylls
without phytol such as chlorophyll derivatives and pheophorbides may have health
effects because of their photo-toxicity, which may be followed by photosensitive
dermatitis (44). A bleaching step is necessary during oil processing to remove
chlorophyll, chlorophyll derivatives, and other color bodies. Changes in chloro-
phylls during canola oil processing are summarized in Table 9.
During oil processing, chlorophyll completely decomposes to derivatives, which
are more difficult to remove during bleaching. This necessitates much higher
amounts of activated bleaching earth to be used to achieve complete removal
of all chlorophyll derivatives (45).
In addition to chlorophyll pigments, carotenoids are also present in canola oil. In
crude canola oil, carotenoids were present at around 130 mg/kg, composed mainly
of xanthophylls (90%) and carotenes (10%). During refining and bleaching, the
amount of carotenoids was reduced to 10 mg/kg (46).
COMPOSITION 71
TABLE 9. Chlorophyll Pigments in Canola Oil During Processing (mg/kg).
Oil After Chlor a Pheo a Pheo b Pyro a Pyro b
Expeller 6.27 4.48 1.79 5.37 0.67

Extraction 1.88 3.31 1.34 16.57 3.13
Expeller Extraction 1.79 5.55 1.34 9.76 1.43
Degumming 0.27 7.16 1.07 9.40 1.84
Alkaline Refining 0.22 6.27 1.12 9.13 1.79
Bleaching 0.56 0.32 0.21 0.25
Abbreviations: Chlor aChlorophyll a; Pheo aPheophytin a; Pheo bPheophytin b; Pyro

aPyropheophytin a; Pyro bPyropheophytin b.
Adapted from Suzuki and Nishioka (45).
The type and content of chlorophylls and their derivatives in the seed define the
quality of extracted and processed canola oil, which has an impact on the quality of
the processed oil. Composition and content of these pigments is related to the
maturity of the seed (Table 10).
In the fully matured seed, only 4 mg/kg of chlorophylls was observed, whereas
in physiologically matured seed, 35 days before maturity, 1239 mg/kg was found.
At maturity, only chlorophyll a and b were present, and all possible isomers/
derivatives were observed at other stages of maturation (Table 10).
4.9. Trace Elements

The proposed Codex standard for edible low erucic acid rapeseed gives the maxi-
mum levels permitted for iron, copper, lead, and arsenic. Although these metals are
found in other edible oils and are present naturally in the seed, nevertheless, they
can also be introduced during handling and processing. Diosady et al. (48) and
Elson et al. (49) examined the effect of processing on trace elements in canola
TABLE 10. Changes in Composition and Content of Chlorophylls During Canola Seed
Maturation (mg/kg).
Time to Maturity (Days)

Chlorophylls 35 27 20 14 6 0
Chlorophyll a 19.5 23.4 27.2 58.7 41.9 82.4

Chlorophyll b 22.2 22.1 15.8 27.3 54.1 17.1
Pheophytin a 43.1 39.8 40.9 10.1 1.1 0.0
Pheophytin b 8.5 7.4 11.3 2.0 0.0 0.0
Pheophorbides a 2.2 1.7 0.6 1.5 0.0 0.0
Pyropheophytin 1.2 2.3 2.5 0.0 0.0 0.0
Methylpheophorbide 3.2 3.5 1.8 0.5 3.0 0.5
Total Amount (mg/kg) 1239 906 463 48 8 4
Adapted from Ward et al. (47).

72 CANOLA OIL
TABLE 11. Mineral Elements Content in Canola Oils (mg/kg).
Oil Sample Phosphorus Iron Calcium Sulfur Zinc Lead
Crude Oil 1190.0 3.52 296.0 6.5 2.4 0.24

Degummed with
Water (WDG) 222.0 1.32 169.0 1.2 2.1
Phosphoric Acid (PDG) 117.2 0.63 34.8 1.5
Bleached
WDG 0.21 0.23 5.6
PDG 0.19 0.59 4.1 0.87
Deodorized
WDG 0.25 0.25 0.07
PDG 0.22 0.38
Adapted from Diosady et al. (48) and Elson et al. (49).
oils. It is evident from the data in Table 11 that processing reduces the amount of toxic
and damaging trace elements, particularly lead, iron, and sulfur. Phosphorus and
calcium form salts, which are insoluble in the oil and can be readily removed during
the degumming process.
Sulfur in canola oil is in the form of organic compounds, which are decomposi-
tion products of glucosinolates. Although these sulfur components occur in trace
quantities, they are known to inhibit catalysts used for hydrogenation as well as giv-
ing a characteristic odour to the oil. Recent developments in analytical methods for
sulfur content evaluation revealed that soybean, sunflower, and even coconut oils all
contain sulfur at a level of 210 mg/kg. Only Brassica oils contain significant quan-
tities of divalent sulfur components. Crude canola oils may contain 15 to 35 mg/kg
of sulfur, whereas in RBD canola oils, the amount of sulfur compounds is reduced
to 9 mg/kg or lower (50). Sulfur components may also improve the stability of the
oil. Some of these components can act as antioxidants and protect the oil from auto-
xidation by complexing hydroperoxy radicals with the sulfur to form stable com-
pounds (51). Other positive action of these compounds consists of inactivating
catalysts involved in oxidative processes, such as metals (51).
4.10. Commercial Crude Oil, Refined and Deodorized Oil

A typical chemical composition of crude, refined, and deodorized canola oils is pre-
sented in Table 12. The deodorized oil data represents the oil quality used as a food
ingredient.
The values for the crude oil compare closely with those of other commercial oils,
such as soybean oil, when produced according to good extraction practices. Chloro-
phylls and sulfur compounds levels are higher in canola oil compared with most
other commodity oils. The deodorized oil data reflect good refining practice and
are similar to the data obtained with other deodorized commodity oils processed
for food applications.
COMPOSITION 73
TABLE 12. Typical Chemical Analysis Data of Crude and Refined,

Bleached, and Deodorized (RBD) Canola Oil.
Parameter Crude oil RBD
Free fatty acids, % 0.31.2 0.03

Phosphorus, mg/kg 300500 <2
Water degummed 120200
Acid-water degummed. 1040
Chlorophyll, mg/kg 430 <0.025
Sulfur, mg/kg 215 <1
Iron, mg/kg 0.51.5 <0.2
Copper, mg/kg <0.2 <0.02
Nickel, mg/kg <0.3
Peroxide value, me/kg 0.53.0 0 (freshly deodorized)
Anisidine value 13 <2
Color, Lovibond <1.5 Red/10 Yellow
Moisture, % <0.3
Flavour bland
Adapted from T. Mag (Unpublished data).
4.11. Oxidative Stability

The stability of canola oil is limited mostly by the presence of linolenic acid, chlor-
ophyll, and its decomposition products and other minor components with high che-
mical reactivity, such as trace amounts of fatty acids with more than three double
bonds. These highly unsaturated fatty acids can possibly be formed during refining
and bleaching (52). The presence of 7% to 11% of linolenic acid in the acylglyce-
rols of canola oil places it in a similar category with soybean oil with respect to
flavor and oxidative stability. The deterioration of flavor as the result of auto -
and photo-oxidation of unsaturated fatty acids in oils and fats is referred to as oxi-
dative rancidity.
The solubility of oxygen in oil is about 35 times greater than in water. The
amount of oxygen present in oil, dissolved during manipulation, is sufficient to
oxidize the oil to a peroxide value of around 10 (53, 54).
The rate of oxidation of fats and oils is affected by the oxygen partial pressure,
access of oxygen, the degree of unsaturation of fatty acids, the presence of light,
heat, antioxidants, and pro-oxidants such as copper, iron, and pigments. The
optimal stability of oil was achieved when iron and copper were below 0.1 and
0.02 ppm, respectively (28).
The degradation of oils and fats by light exposure is primarily a photocatalyzed
oxidation. During photooxidation, singlet oxygen is generated by transformation of
energy to a sensitizer, which activates oxygen. Singlet oxygen is an extremely reactive
specie of oxygen, 1500 times more reactive than normal oxygen, and reacts with
double bonds of unsaturated fatty acids to form peroxides or free radicals. Typical
photosensitizers include chlorophylls and their decomposition products formed dur-
ing maturation of seed and processing, heme compounds, and polycyclic aromatic
74 CANOLA OIL
hydrocarbons (28). It has been found that chlorophyll degradation products are
more active as photosensitizers than chlorophyll (55).
5. PHYSICAL PROPERTIES
The properties of canola oil are governed by the components present in the oil and
described by the general standards for vegetable fats and oils. Selected physical
properties for canola oil in comparison with HEAR oil are shown in Table 13.
5.1. Relative Density

Typical values for the specific gravity of canola oil are shown in Table 13. Ackman
and Eaton (56) indicated that a different proportion between eicosenoic (C20:1) and
octadecanoic, polyunsaturated fatty acids could be a major factor in changing the
relative density of canola oil. Noureddini et al. (57) described relationship between
temperature and density of vegetable oils including canola. As for other liquids, the
density for vegetable oils is temperature dependent and decreases in value when the
temperature increases. The same authors also modified the Rackett equation to
calculate density, which is based on the composition of fatty acids.
5.2. Viscosity
Viscosity values estimate the relative thickness or resistance of an oil to flow. The
viscosity of RBD canola oil was higher than for soybean oil.
Lang et al. (58) and Noureddini et al. (59) found the viscosity of canola and
other vegetable oils was affected by temperature and proposed an equation to cal-
culate viscosity in the temperature range of 4 100 C. The viscosity of HEAR oil is
significantly higher than that of canola oil.
TABLE 13. Physical Properties of Canola and Rapeseed Oils.
Value
Parameter Canola HEAR
Relative density (g/cm3; 20 C/water at 20 C) 0.9140.917 0.9070.911

Refractive index (nD 40 C) 1.4651.467 1.4651.469
Crismer value 6770 8082
Viscosity (kinematic at 20 C, mm2/sec) 78.2 84.6
Cold test (15 hrs at 4 C) Passed Passed
Smoke point ( C) 220230 226234
Flash point, open cup ( C) 275290 278282
Specific heat (J/g at 20 C) 1.9101.916 1.9001.911
Saponification number 188192 168181
Iodine value 110126 97108
Abbreviations: HEARHigh erucic acid rapeseed oil.

PHYSICAL PROPERTIES 75
5.3. Smoke and Flash Point

The smoke point is the temperature at which a fat or oil produces a continuous wisp
of smoke. This provides a useful characterization of its suitability for frying, and
often 200 C is specified as the minimum by regulations (Table 13).
The flash point defines the temperature at which the decomposition products
formed from frying oils can be ignited. This temperature ranges from 275 C to
330 C for different oils and fats (Table 13). An increase in the content of unsatu-
rated fatty acids usually decreases the flash and smoke points (60).
5.4. Cold Test

The cold test measures the resistance of oil to the formation of sediment at 0 C or
4 C (Table 13). Sediment formation is usually caused by compounds with a high
melting temperature, mainly waxes and triacylglycerols with long-chain saturated
fatty acids (15). The formation of haze in canola oil is not a common occurrence,
but it may happen occasionally (2). It has been observed that canola oil produced
from seeds grown in dry/drought conditions develop sedimentation more easily.
This may be related to the higher content of saturated fatty acids formed as a
response to drought stress conditions (15).
5.5. Crismer Value

The Crismer value (CV) measures the miscibility of an oil in a standard solvent
mixture, composed of t-amyl alcohol, ethyl alcohol, and water in volume proportion
5:5:0.27 (Table 13). This parameter is one of the specification criteria used for inter-
national trade, mostly in Europe; however, today it is rarely used. The values
obtained are characteristic within a narrow limit for each kind of oil. The miscibil-
ity of oil is related to the solubility of acylglycerols and is affected mainly by the
unsaturation and chain length of the constituent fatty acids.
5.6. Saponification Value

The saponification value is defined as the weight of potassium hydroxide, in
milligrams, needed to saponify 1 g of fat. This parameter is inversely proportional
to the molecular weight of the fat. In other words, the higher the molecular weight,
the lower the saponification value. Replacement of long-chain fatty acids such as
erucic acid in canola oil by octadecenoic fatty acids increased the saponification
numbers from 168181 to 188192 because of the reduction in molecular weight
(Table 13).
5.7. Iodine Value

The iodine value (IV) is an empirical test indicating the degree of unsaturation of a
fat or oil. It is defined as the number of grams of iodine absorbed by 100 g of fat.
76 CANOLA OIL
TABLE 14. Melting Characteristics of the Octadecanoic

Fatty Acid Family.
Fatty Acid Melting Point ( C)
Palmitic 64.5
Stearic 69.6
Oleic (cis 9-octadecenoic) 13.2
Elaidic (trans 9-octadecenoic) 43.7
Octadecenoic (cis 6) 28.6
Linoleic (cis 9, 12) 5.1
Linolenic (cis 9, 12, 15) 11.2
Adapted from Mag (2).
The higher value for canola oil is caused, in part, by the replacement of erucic acid
with octadecenoic acids, mainly oleic acid, accompanied by a slight increase in
linoleic and linolenic acids (Table 13). The iodine value can also be calculated
from fatty acid composition using the specific factors for each unsaturated fatty
acid (61). The calculation method provides more accurate data than the iodine
absorption assessment.
5.8. Melting Characteristics, Polymorphism, and Crystal Properties

Canola oil has a homogeneous fatty acid composition, with 95% being contributed
by octadecanoic fatty acids (4). Reducing the erucic acid content had a marked
effect on the melting characteristics and the type of crystal formed when the oil
is hydrogenated. Hydrogenation of canola oil is used to form the products such
as shortenings and margarines. Increasing the degree of hydrogenation causes the
fatty acid composition to be more homogeneous. This results in a tendency to form
large beta crystals on solidification, which is undesirable in margarines and short-
enings. Trans-isomers formed during hydrogenation have higher melting points
than cis fatty acids (Table 14) (62). Trans-isomers introduce greater variety in
the fatty acid composition of the hydrogenated oils, thereby reducing the beta
crystallization tendency of the oil.
6. CANOLA OIL EXTRACTION AND PROCESSING
Processing methods developed over the years are designed to extract canola oil
from the seeds to produce a high-quality raw oil for further processing and a
high-quality protein meal as an animal feed.
6.1. Canola Oil Extraction

Canola seeds contain 38 44% oil (8% moisture basis) so that the most efficient
method of extracting the oil is by mechanically (pressing) expelling about 60% of
CANOLA OIL EXTRACTION AND PROCESSING 77
Figure 1. Canola seed extraction.
the oil, which is followed by solvent extraction of the remainder. A similar method
is applied to other seeds of high oil content, such as sunflower seeds. Figure 1 gives
an overview of the process sequence most commonly applied in canola seed
extraction.
6.1.1. Seed Cleaning Canola seed is generally received at the extraction plant
directly from the farm, or from a seed gathering station. Weed seeds, grain seeds,
and other foreign material must be removed from the seed before extraction of the
oil. The seed cleaning process consists of three stages: aspiration to remove dust
78 CANOLA OIL
and very light material, screening to remove oversized particles, and rescreening to
remove undersized material. The equipment usually comes as one unit, and the
removal of foreign material is to less than 2.5%. Disposal of the separated
material is used primarily in animal feed.
6.1.2. Preconditioning This is essentially a preheating step to bring the seed

temperature before flaking to about 3040 C. Under these conditions, the seeds
are more pliable and thus less likely to shatter during flaking seeds. Preheating
equipment may be of the fluid bed type using hot air or steam, direct, or indirect
heat in rotary kilns equipped with steam-heated coils.
Moisture content in the seed can be adjusted, if required, during this process step
by regulating the airflow through the equipment. The desired moisture content prior
to flaking is in the range of 7.09.5%.
6.1.3. Flaking The preheated canola seed is flaked on smooth-surface rolling

mills. Some operators advocate flaking in two stages. In the first stage, a flake
thickness of about 0.40.7 mm is produced. In a second set of rolls, a flake of
0.20.3 mm thickness is produced. The main reason for two-stage flaking is the
small size and the deformability of the seed, which usually allows some whole
seed to pass through in a single-stage flaking. However, many plants operate
single-stage flaking with satisfactory results.
Flaking ruptures the cell walls, which allows some of the oil to be separated
from the seed residue by simple pressing. Further, the oil retained in the seed resi-
due can now be more efficiently leached out in subsequent solvent extraction of the
press cake with hexane. Properly flaked material is more easily treated in the sub-
sequent cooking operation and requires less mechanical energy in pressing. It is
essential that the residual oil content in the meal be low for this operation to be
economical.
6.1.4. Cooking The flaked material is subjected to a cooking step. In this

process, the flakes are heated either indirectly on steam-heated surfaces in stack
cookers or in rotary cookers equipped with steam coils. Cooking temperatures
are usually held to 75100 C. Ambient air may be admitted to the cooking
equipment to adjust the moisture content of the material to 57% for the expeller
operation.
Cooking serves several important functions. It coalesces minute oil droplets into
larger ones, which can be easily separated as well as changes the properties of the
protein so as to make the oil more easily extractable. But temperature and heating
time must be carefully controlled to avoid negative effects on color and sulfur levels
in the oil, protein degradation, and the percolation properties of the flaked material.
Cooking also inactivates two key enzymes. The first is myrosinase, which hydro-
lyses glucosinolates into oil-soluble sulfur-containing compounds. The second is
phospholipase, which hydrolyses phospholipids into nonhydratable phosphatides.
Myrosinase activity has been extensively studied and is carefully controlled during
the cooking operation that minimizes hydrolysis of glucosinolates and reduces
the presence of sulfur in the oil (63, 64). The optimal temperature range for the
myrosinase activity is between 50 C and 70 C with moisture content in the range
of 57%.
Phospholipase activity as affected by cooking operations has also been a subject
of interest but to a lesser extent. Dahlen and Kristofferson (65) found that as cook-
ing temperatures were raised from 85 C to 100 C, nonhydratable phospholipids
content decreased in the oil, whereas total amounts of phosphorus, color and sulfur
increased. Higher color and sulfur are undesirable as increased sulfur content
impairs the catalyst in hydrogenation of the oil. Unger (66) showed that not only
total phosphorus, but also free fatty acid amounts, increased with higher cooking
temperatures. Current cooking operations use temperatures in the range of
75100 C. The sulfur content of canola oil is in the order of 215 mg/kg as deter-
mined by inductively coupled plasma (ICP) spectroscopy. It is apparent, that in
practice the selection of cooking temperature is a compromise between opposing
effects on oil quality and economics.
Process equipment for seed cooking other than the stack cookers and rotary
cookers described above are also used in the seed extraction industry. Three
processes are worth noting, even though there is no known application of these
in pretreating in canola/rapeseed processing: the Alcon process by Lurgi of
Germany, the Exergy process by Stork of Sweden, and the use of expanders.
All of these cooking processes have in common that they are designed to heat the
material very quickly in a few minutes or even seconds. This avoids the main short-
comings of stack cookers and rotary cookers, long residence times in the range of
1030 minutes. In these processes, rapid heating makes it possible to inactivate
enzymes very quickly and to expose the seed or flakes to an increased temperature
for very short time. Rapid heating depends on very good mixing of the seed mate-
rial during heating, and results in very good temperature control, uniform heat treat-
ment, and prevention of enzyme-catalyzed damage to the oil in the seed. An
unexpected bonus of these heat treatments is the formation of the oilseed with
high porosity, which facilitates improve oil yield during solvent extraction.
The Alcon process. This process is being used with soybean flakes. It was the
first commercially applied rapid cooking method, and its main purpose was to inac-
tivate phospholipases. Soybean oil from Alcon-treated soybean flakes is free of
nonhydratable phosphatides, and water degumming is sufficient to completely
remove phospholipids. In the treatment, soybean flakes are introduced into a tower,
where direct steam with intensive agitation provides rapid heating of the seeds and
the required temperature to inactivate enzymes is achieved uniformly in a few
minutes. The heated flakes are discharged into a cooling chamber before the hexane
extraction. To our knowledge, this process is not commercially used for canola seed
processing. Details of this process were described by Penk (67).
The Exergy process. In this process, whole seeds or flakes are suspended in
superheated steam at high pressure when they are moved turbulently through a
heated pipe loop. This heat treatment requires only a few seconds to achieve the
temperature needed to inactivate enzymes (68). In the case of canola, whole seed
or flakes can be treated this way. Heat treatment of whole seeds prevents enzymatic
80 CANOLA OIL
degradation of oil, which occurs after flaking when content of seed is exposed to an
enzymes action. The heated seeds are cooled before going on to flaking or extrac-
tion. The process was originally developed for efficient drying of waste sludges and
is being used in Scandinavian countries to produce high-bypass protein canola
meal. It has been proposed as pretreatment for canola seeds before extraction.
Some commercial applications may in fact exist in Scandinavia. The cost of the
equipment is an obstacle to wide application.
Extruders. Extruders, or expanders, are widely applied in the oilseed industry,
but mostly for press-cake conditioning in canola oil extraction, and for extruding
soybean flakes into expanded collets for improved extraction yield. Extruders as
cookers are used in some soybean extraction plants to achieve enzyme inactivation.
Lusas investigated inactivation of enzymes during extrusion of soybean (69). Extru-
ders could be used to pretreat canola seed to assist prepressing and to inactivate the
enzymes; however, commercial application of this process in canola industry is not
known.
6.1.5. Pressing/Expelling (Prepressing) The heat-conditioned, flaked seed is

conveyed to continuous screw-presses or expellers. The function of the expellers is
to reduce the oil content of the seed from about 42% to 1620%. Extraction of the
remaining oil is then much more efficient and economical. Some discharge of very
fine solids (715%) with the oil draining from the expeller is unavoidable. These
fines are separated from the oil by gravity in settling tanks followed by filtration and
recycled to the conditioning stage. The press-cake, which may be put through a
mechanical breaker to produce uniform-sized particles, is now ready for solvent
extraction. Unger (66) and especially Buhr (70) have described the operation of
expellers with canola seed in detail.
6.1.6. Extruders In some plants, the press-cake is subjected to mechanical extru-

sion to improve its solvent extraction properties. Extruders (expanders) used for this
purpose consist of a barrel with a rotating shaft fitted with flights. Steam is added,
and heating and mixing take place along the length of the barrel. Pressure is devel-
oped, and the material is then discharged through the small openings of a die plate
at the end of the barrel. The pressure release on discharge expands the extruded
material, making it very porous. These small-diameter, porous pieces of press-cake
(collets) have excellent solvent extraction properties. As a result, throughput of the
solvent extraction equipment is significantly enhanced.
6.1.7. Solvent Extraction The press-cake from the expellers and cake breakers,
or as collets from extruders, is conveyed to the solvent extraction stage. Some cool-
ing of the 80100 C material with ambient air is usually done during conveying to
minimize vaporization of hexane when the cake enters the extractor. A variety of
extractor designs are in use, often made by the same firms supplying expellers.
The solvent used is extraction hexane at 5060 C.
In the extractor, the solids, which are at about 80 C, are solvent-washed in stages,
first with hexane already high in oil content (miscella) and then with progressively
leaner miscella and, finally, with pure hexane. The various extractor designs
approximate countercurrent extraction to a greater or lesser extent. The oil content
in the solid material (meal) is reduced down to about 1%, depending somewhat on
equipment design and throughput rate and on how well the cooking, flaking, and
prepress operations were carried out. The flow rate of hexane to the extractor is
adjusted to provide an oil concentration of 2530% in the oil-in-hexane solution
(miscella).
6.1.8. Desolventizing of Meal and Oil The meal and the miscella are stripped
of the solvent to recover solvent-free meal and oil. The solvent-saturated meal is
conveyed to a desolventizer, which is a vertical tank equipped with steam-heated
trays and rotating sweep arms just above each tray to agitate the solid material.
Reduced pressure and some live steam sparging are used to evaporate the hexane
and to dry the meal. The meal is heated and toasted to about 105 C with
residence time of about 3040 minutes. Residual hexane in the meal is in the order
of 1500 mg/kg, and moisture in the range of 1518%. Lower hexane concentrations
are difficult to achieve. Grant et al. (71) described factors affecting canola meal des-
olventizing (68). Schneider and Schuette (72) have investigated the difficulties in
desolventizing of canola meal in relation to fiber structure. Some removal of glu-
cosinolates, their breakdown products, as well as protein denaturation occurs (73).
To achieve the best meal quality, the process must be well controlled with respect to
temperature (110 C max.) and time.
Meal cooling and drying is an integral part of the desolventizer (Desolventizer-
Toaster-Dryer-Cooler) or separate rotary kilns. The temperature is reduced to
3040 C, moisture to 811%, and hexane to about 800 mg/kg. This corresponds
to a hexane loss of approximately 1 liter per metric ton of seed processed. Cooled
meal may be ground to a uniform particle size, or pelleted, ready for storage and
marketing. The hexane and moisture vapors are vented from the meal desolventiz-
ing equipment and condensed, the water and hexane are separated, and the hexane
is reused. The miscella containing the oil is desolventized in three-stage evaporator
equipment. The hexane vapor from this operation is condensed for reuse. A prop-
erly sized and operated extraction plant looses no more than about 12 liters of
hexane/ton of seed processed.
6.1.9. Quality Assurance: Seed, Oil and Meal The first step in the quality
assurance process is the grading of seed deliveries from the farm to the extraction
plant. In Canada, for example, the Canadian Grain Commission as a third party is
responsible for setting the grading rules by which the industry conducts trades. The
rules are published in the Official Grain Grading Guide by the Office of the Chief
Grain Inspector, Inspection Division, Winnipeg, Canada (74). Three grades, from
No. 1 to No. 3, are distinguished, with No. 1 being best. Poorer grade seed, which
may sell at a discount, produces lower quality oil. This is especially evident in high-
er chlorophyll and free fatty acid content. Nonhydratable phosphatide content may
also be higher with poorer grades. All of these factors make for greater difficulty
and higher costs in oil processing.
82 CANOLA OIL
TABLE 15. Canola Meal Specifications (75).
Characteristic Specification
Protein (weight %) min. 36

Fat (oil) (weight %) max. 4
Moisture (weight %) max. 12
Crude fiber (weight %) max. 12
Glucosinolates (mmol/g) max. 30
Screen analysis 97 weight % of the meal shall pass through a 2.00-mm
(US # 10) sieve, and 90 weight % shall pass through
1.70-mm (US # 12) sieve.
Canadian quality assurance on crude oil, degummed oil, and meal is governed by
the Trading Rules of the Canadian Oilseed Processors Association (COPA) (75).
Official AOCS methods are used to analyze quality parameters (76). Glucosinolate
determinations on meal are not usually done. Typical properties of oil and meal are
given in Table 1 and Table 15, respectively.
6.2. Canola Oils Processing

Processing of crude canola oil to edible oil products is very similar to that applied
to other vegetable oils. Figure 2 gives an overview of the process steps that are
applied in the industry.
6.2.1. Degumming The crude oil from prepressing and from solvent extrac-
tion is usually blended and then degummed before being stored for sale or further
processing. Degumming removes phosphatides coextracted with the oil, as they
tend to separate from the non-degummed oil as sludge during storage. The
phosphatide content of crude oil varies, but it is usually in the order of 1.01.5%
or, measured as phosphorus, 400600 mg/kg. Two main degumming methods are in
use: (1) using water to precipitate phosphatides (water degumming) and (2) using
an acid such as citric, malic, or phosphoric and water (super-degumming, or acid-
water degumming) to achieve nearly complete degumming. After contacting the oil
with these reagents, the oil is centrifuged to separate the precipitated material.
Other degumming processes for achieving nearly complete degumming are also
applicable, such as enzymatic degumming, but so far has seen only very limited
use.
6.2.1.1. Degumming with Water Degumming with water only involves contact-
ing the oil at about 80 C with about 2% water in a gently agitated tank with a 530-
minute contact time, to keep the precipitated phosphatides in suspension and allow
time for agglomeration, and then separating in a disk centrifuge. This leaves from
100200 mg/kg of phosphorus (0.250.5% phosphatides) in the oil, depending on
the extent to which nonhydratable phosphatides are present. The oil lost with the
separated phosphatides is in the order of 3540% of the separated material.
Figure 2. Canola oil processing.
6.2.1.2. Degumming with Acid and Water In degumming with acid and water,
the nonhydratable phosphatides (NHP) are also removed. A number of process ver-
sions, which have been developed over the years, are suitable, such as the Unilever
Super-degumming (acid degumming) process (77) and various other proprietary
processes. Reaction conditions are much more critical than in water degumming.
Usually, the following sequence of steps, with some variations, are applied: Crude
oil is brought to about 6070 C and then contacted with about 0.10.5% of a 50%
solution of citric or malic acid with very intensive mixing. Contact may vary from
fractions of a second to about 15 minutes, depending on mixing intensity. The oil is
then contacted with about 2% water, either before or after cooling to 2545 C.
About 13 hours of residence time is then given in a continuous, stirred reactor.
84 CANOLA OIL
Agitation in the reactor is designed to avoid shear and keep the precipitated phos-
phatides uniformly suspended. The oil/phosphatide mixture is then heated very
rapidly, immediately ahead of a self-desludging disk centrifuge, in which the pre-
cipitated phosphatides are separated from the oil. The oil loss in this process repre-
sents from 3550% of the separated phosphatides, with the higher losses occurring
with oils that are relatively high in nonhydratable phosphatides. The separated
phosphatide phase is added to the meal in the desolventizer. This raises the residual
oil content of the meal to about 23% and raises its energy content.
Residual phosphatide concentrations are from 550 mg/kg of phosphorus
depending on process details and the level of nonhydratable phosphatides in the
crude oil. Nonhydratable phosphatides can vary widely, but are usually in the range
of 2045% of total phosphatides. With higher concentrations of nonhydratable
phosphatides, longer agglomeration time is required to achieve clean separation
in the centrifuge.
In the most recent development, acid and aqueous sodium hydroxide, rather than
acid and water are being used, especially with lower quality oils (acid-caustic
degumming). This represents an intermediate between acid-water degumming
and alkali refining. It obviates the need for cooling, similar to the practice in alkali
refining. Phosphoric acid rather than an organic acid is preferred in this case. Phos-
phatides as well as some of the other impurities are removed, and if sufficient, alkali
is applied to saponify the free fatty acids, fully refined oil is obtained.
6.2.2. Alkali and Physical (Steam) Refining Degummed oil is further purified
in a process of refining. One of two methods are used, namely, alkali refining, espe-
cially with water-degummed oil, and physical refining with acid-water-degummed,
or acid-caustic-degummed oil, that is, with the more completely degummed oil.
Alkali refining is the most common process used, even with acid-water-degummed
oil. Physical refining of canola oil on a plant scale is relatively rarely practiced. It
requires well-degummed oil of moderate chlorophyll and free fatty acid content,
but it is then economical, especially, with respect to capital investment. CanAmera
Foods (now part of Bunge, Inc) in Canada has successfully operated a refinery in
Western Canada based on physical refining since 1984.
6.2.2.1. Alkali Refining In alkali refining, the oil is first contacted with about
0.050.1% of concentrated phosphoric acid in a high-intensity mixer to help preci-
pitate phosphatides. It is then contacted with an approximately 12% concentrated
aqueous solution of sodium hydroxide to neutralize the free fatty acids present, any
excess phosphoric acid and to precipitate phosphatides. This requires about 23%
of the solution. Very intensive mixing is used. Temperatures and contact times may
vary from about 90 C and only seconds for both acid and caustic in the short-
mix process to about 40 C and about 15 minutes each in the long-mix process.
The oil/soap mixture is then heated to about 90 C, if required, and centrifuged to
separate the aqueous soap phase. This phase also contains the precipitated phospha-
tides and some triglyceride oil. Usually, self-desludging disk centrifuges are used
to avoid frequent shutdowns for cleaning. The centrifuged oil must be further
contacted with about 510% hot, soft water to reduce soap levels from about
500 mg/kg to <50 mg/kg. Solid bowl disk centrifuges are adequate for this service.
Alkali refining reduces free fatty acids to <0.05% and phosphorus to <3 mg/kg.
Iron and copper concentrations are reduced to below the detection limit. Colored
compounds (chlorophyll derivatives, carotenoids) are not affected significantly,
although chlorophyll derivatives can be reduced by 3070% of the original concen-
tration under certain process conditions. The concentration of sulfur compounds is
reduced slightly. This oil is now ready for bleaching. The soap phase from this
operation can be added to the meal, similar to the disposal of the phosphatides
from degumming, or it can be acidulated and used as a feed fat. Mag (78) has
reviewed industrial refining practice in some detail. Carlson (79) reviewed refining
methods and process costs.
6.2.2.2. Physical Refining In physical refining, acid-water-degummed oil with

phosphorus content below 50 mg/kg and preferably below 25 mg/kg is first sub-
jected to a phosphoric acid pretreatment, as in the short-mix alkali refining process.
It is then contacted with acid-activated bleaching clay in a standard bleaching pro-
cess at 95105 C. With precipitated phosphatides, chlorophyll derivatives and some
carotenoids are adsorbed on the clay and removed by filtration. Together with the
acid-water degumming of the oil, this is the most important stage of physical refin-
ing. It delivers bleached oil ready for physical (steam) refining/deodorizing. Except
for the free fatty acids in the oil, all other minor constituents are now at the same
concentrations as in alkali-refined oil, and in addition, chlorophyll derivatives are
reduced to the concentration required of bleached oil, namely, <0.050 mg/kg.
Usually, from 0.7% to 2% of acid-activated bleaching clay is used, depending pri-
marily on the concentration and type of chlorophyll derivatives and the residual
phosphatide concentration that was present in the degummed oil.
The removal of the free fatty acids in the bleached oil is done by steam distilla-
tion in a deodorizer. This, simultaneously, deodorizes the oil. Because deodoriza-
tion is, also, the last process normally carried out on edible oils, this step may be
delayed until other processes, such as hydrogenation of the oil, have been done.
6.2.3. Adsorptive Bleaching Alkali-refined oil requires bleaching. It usually

contains traces of soap, and the concentration of chlorophylls and their derivatives
is still essentially that present in the crude oil. Some of the oxidation products and
traces of heavy metals, notably, iron are also removed. But, the most important
effect to be achieved with canola oil is the removal of chlorophyll derivatives.
These compounds accelerate oil oxidation in presence of light (80) and give an
undesirable green color to the oil. Acid-activated clays are used in bleaching. Their
adsorptive properties are especially effective for the removal of green com-
pounds. With high chlorophyll concentrations, >30 mg/kg, chlorophyll specific
clays have been used. One feature of these clays is that they are especially highly
acid activated. As with other oils, the adsorption of carotenoids on bleaching clays
is not efficient. Some reduction occurs, partly because of adsorption and partly
because of heat breakdown. In usual bleaching practice, the emphasis is on
86 CANOLA OIL
chlorophyll removal. Heat bleaching of the carotenoids during deodorization is

relied on to remove most of the red color present in the oil.
6.2.3.1. Adsorptive Clay Adsorptive bleaching is carried out under vacuum with
the oil at about 100 C. About 530 minutes of contact time is given, while the oil/
clay slurry is progressively dried to about 0.1% moisture content. This appears
to give the best adsorption efficiency. As indicated earlier, about 0.72% clay
(1015% moisture content) may be required to achieve chlorophyll removal to
<0.050 mg/kg measured as chlorophyll a. This level of chlorophyll derivatives is
innocuous with respect to oxidation and color of the oil. Mag (81) has given an
overview of edible oil bleaching practice, with particular attention to canola oil.
Brimberg (82) and Henderson (83) have investigated kinetic aspects of chlorophyll
adsorption important to the process. Suzuki and Nishioka (84) investigated changes
in chlorophyll derivatives during seed and oil processing up to the bleaching stage
and determined the ease of adsorption of the various chlorophyll derivatives on
activated clay and on active carbon. With acid-activated clay, they found the ease
of adsorption to decrease in the following order: pheophytin a > pyropheophytin
a pheophytin b > pyropheophytin b. Chlorophyll a and b are converted to their
respective pheophytins early in seed processing. Thus, they do not play a significant
role in the adsorption of chlorophyll derivatives. Bleaching processes suitable for
canola oil do not differ from those used for other oils.
6.2.3.2. Active Carbon Active carbon is rarely used to remove chlorophyll com-
pounds is used very little in canola oil bleaching. It presents greater difficulty in
handling and it retains more oil than activated clays do and is much more expen-
sive. In bleaching efficiency tests, active carbon has been shown to be somewhat
more efficient than activated clay at high concentrations of chlorophyll compounds
in canola oil, but less efficient at very low concentrations. Thus, active carbon is not
a suitable adsorbent to achieve the removal of chlorophyll derivatives to the very
low concentrations mentioned earlier.
6.2.3.3. Synthetic Silicas In recent years, the use of synthetic silicas as a pre-
treatment in the bleaching of oils has been advocated. In connection with canola
oil bleaching, it is claimed that because of the ability of these silicas to adsorb
soap and phosphatides more efficiently than acid-activated clays, adsorption of
chlorophyll derivatives by the clays in subsequent bleaching is much more efficient.
Presently available silicas do not adsorb chlorophyll derivatives or carotenoids.
Parker (85) working for W. R. Grace & Co.-Conn., a manufacturer of silicas, is
advocating a synthetic silica pretreatment followed by packed bed bleaching with
acid-activated clay.
6.2.3.4. Other Methods to Remove Chlorophyll Derivatives Other approaches to

the removal of chlorophyll derivatives compounds from canola oil than adsorption
are known and are being practiced. Bergman (86) and Szemraj (87) described the
use of concentrated phosphoric acid and vacuum drying of the oil to precipitate
TABLE 16. Bleached Canola Oil After Alkali Refining or Acid Degumming.
Characteristic Alkali Refined/Bleached Acid Degummed/Bleached
Free fatty acids (%) 0.1 as for crude

Phosphorus (mg/kg) <3 <3
Chlorophyll (mg/kg) <50 <50
Iron (mg/kg) <0.1 <0.1
Peroxide value (meq/kg) Nil Nil
Anisidine value 13 13
Sulfur (mg/kg) 113 113
Soap (mg/kg) Nil Nil
impurities from rapeseed oil, especially chlorophyll derivatives compounds.

Experience has shown that the precipitation of chlorophyll is much more efficient
with super-degummed canola oil. The removal of the precipitated material from
the oil can be by filtration on a suitable filter aid, or by centrifuging. When the
precipitation step is followed by alkali refining of the oil, the acid-precipitated
chlorophyll derivatives are simply removed as part of the heavy phase in the
centrifugal soapstock separation.
Another approach is to make use of the fact that in contact with alkali, chloro-
phyll derivatives also react to form a precipitate (88). This can be used in the course
of alkali refining. The precipitate can be removed together with the aqueous soap
phase, but at this time, specific details are still proprietary. Experience so far has
shown that as much as about 70% of the chlorophyll derivatives can be removed
in the course of alkali refining.
Typical properties of alkali-refined, bleached canola oil and of acid-water-
degummed, acid pretreated, bleached canola oil ready for hydrogenation or steam
refining/deodorization are given in Table 16. With the exception of the concentra-
tion of free fatty acids, the two process routes produce the same bleached oil
quality.
6.2.4. Hydrogenation Hydrogenation changes the melting behavior of oils and

improves their oxidative stability. It is applied to many oils, including canola.
Hydrogen is added to the double bonds of unsaturated fatty acids at temperatures
of 160200 C and pressures of 100300 kPa in the presence of a nickel catalyst to
facilitate the reaction. Chlorophyll (89), phosphatides, soaps, and especially sulfur
compounds poison the catalyst, raising process costs. These are removed in
refining and bleaching, except sulfur. Sulfur compounds are only reduced somewhat
in these treatments. de Man et al. (90) investigated the role of sulfur compounds in
nickel catalyst poisoning in canola oil hydrogenation and found that as little as
1 mg/kg of sulfur in the form of allyl isothiocyanate (AITC) had a noticeable effect
on hydrogenation rate. Additions of 3 and 5 mg/kg of AITC had an even greater
effect on the rate of hydrogenation, and trans-isomer formation was very signifi-
cantly increased in the range of iodine values from 95 to 80. AITC is typical of
compound formed from the breakdown of glucosinolates.
88 CANOLA OIL
TABLE 17. Hydrogenated Canola Oils for Margarines, Shortenings, and Frying Fats.
Iodine Fatty Acid Composition (%) SFI at C
Value C16:0 C18:0 C18:1 C18:2 C18:3 Trans 10.0 21.1 26.7 33.3 40.0
115(L) 4.5 2 60 22 10
90 4.5 3 80 9 2 20 2.5
85 4.5 5 86 2 1 42 18 5
80 4.5 11 83 45 30 18 7 <1
75 4.5 14 80 50 50 31 25 11 4
The determination of sulfur compounds in the range of concentrations, which

occur in edible oils, is difficult. The literature on concentrations found in rapeseed
and canola oils can be confusing. Considering relevance to hydrogenation effects
and speed of analysis, ICP spectroscopy for determining sulfur in canola oil is pre-
sently the most appropriate method.
The process of hydrogenation of canola oil is usually carried out batch-wise.
There are various reactor designs in use. Often, in-house designs are used, but
many companies in the field of supplying the edible oil industry with process equip-
ment are able to furnish proven hydrogenation and catalyst removal equipment.
Depending on temperature, pressure, and, to some extent, the specific catalyst
properties, a different melting behavior of the hydrogenated fat can be obtained.
The oxidative stability improvement, also, is modified somewhat by choice of pro-
cess conditions. Usually, conditions are chosen that provide this selectivity toward
the desired melting behavior and oxidative stability improvement. In canola oil,
very low concentrations of linolenic and linoleic acid can be achieved with very
little hardening of the oil (Table 17). Melting behavior is most usually evaluated
by determining the proportion of solid fat in a sample over the temperature range
of interest, usually 1040 C. Dilatometry is still extensively used in North America
for this purpose, rather than the commonly used and more modern wide-line nucle-
ar magnetic resonance spectroscopy.
Because of the small concentration of sulfur compounds in canola oil, there is a
somewhat greater tendency of sulfur poisoning of the catalyst, with the result that
slightly higher trans-isomer concentrations may come about compared with, for
example, soybean oil. Hatfield (91) has focused on this aspect of canola oil
hydrogenation.
Removal of the nickel catalyst from the oil is by filtration of the oil from the
reactor after cooling to about 100 C on filter aid precoated filters. Adding filter
aid to the oil and then filtering is also satisfactory. Nickel concentrations of
315 mg/kg are achieved. This is followed by a second filtration on a filter aid pre-
coated filter, possibly after contacting the oil with a small amount of citric acid
(0.05% as a 50% aqueous solution) as a chelating agent. In stubborn cases, it is
necessary to bleach the oil with acid-activated bleaching clay. Nickel must be
removed to below 0.30.5 mg/kg, that is, to Ni-negative by the ammonium sulfide
test. Synthetic silicas have also been advocated as a treatment of hydrogenated oil
for nickel removal either before or after catalyst filtration (92).
6.2.4.1. Hydrogenated Canola Oil in Edible Oil Products Hydrogenated fats

from canola oil play an important role in the production of the wide range of fat
products in the countries in which canola oil is used. They often appear in blends
with other oils to combine the advantages of canola oil with those of other oils.
They are especially important as lightly hydrogenated oils for their exceptional
stability, This is caused, in part, by the high content of monounsaturated fatty acids
in canola oil. These products are very popular because of their good stability and
pourability and their low saturated fatty acid content. The introduction of nonhydro-
genated high-oleic acid (75%) containing canola oil in the market may eventually
reduce the importance of lightly hydrogenated oil.
Table 17 shows some typical iodine values, fatty acid compositions, and solid fat
indices for hydrogenated canola oils used in margarine, shortening, and frying fat
formulations. The nonhydrogenated, liquid oil data are given for comparison. It
should be noted that the values given might differ somewhat in practice depending
on hydrogenation conditions and the catalyst used.
The above hydrogenated canola oil stocks, or similar ones, are used extensively
in shortenings and as stable frying fats. Industry makes a wide variety of canola-
based hydrogenated oil stocks for tailor-made shortenings. Lightly hydrogenated
canola oil (IV 90), for example, has a special advantage in that even at very low
levels of solid fat, it has very low polyunsaturation, which makes it an excellent
pourable frying fat base of very good stability. Very highly hydrogenated canola
oils are rarely used in margarine and baking shortenings, because of their tendency
to form large (beta) crystals over time. This impairs eating properties and baking
performance. deMan and deMan (93) and Naguib-Mostafa (94) studied the crystal-
lization behavior of hydrogenated canola oil used for margarine and shortening
products. DSousa et al. (95) stated that at least 11% of palmitic acid is required
when palm oil is used in hydrogenated canola oil-based margarines and shortenings
to achieve crystal stability in the beta prime (b0 ) form for adequate shelf life of
margarine. Preferably, palm oil needs to be lightly hydrogenated.
6.2.5. Interesterification Interesterification is another process for changing the

melting properties of fats and oils. The process represents an alternative to the use
of partially hydrogenated fats in manufacturing products of a variety of melting
properties.
Thomas (96) and Desrosier (97) investigated chemically catalyzed interesterifi-
cation of nonhydrogenated canola oil with fully hydrogenated vegetable oils. In the
manufacture of zero trans-isomer fat products, interesterification of canola and
other oils with high melting fats is of interest to take advantage of the nutritionally
favorable fatty acid profile of nonhydrogenated canola oil. The process has also
been proposed for use in controlling the problem with beta crystallization of hydro-
genated canola oil. However, this process has not found significant use because of
cost. Kurashige et al. (98) used enzyme-catalyzed interesterification of palm oil
with canola oil to improve the pourability of palm oil compared with merely blend-
ing with canola oil. They used a 1,3-specific lipase. Canola oil was used because of
90 CANOLA OIL
its low content of saturated fatty acids. Industrial scale processes to carry out
enzyme-catalyzed interesterification, however, are still rare.
6.2.6 Dewaxing Canola oil is a natural salad oil. This means that it remains clear
and liquid at refrigerator temperatures. It is used without winterization to produce
bottled oil and salad dressings. However, the oil may contain a small and variable
concentration of compounds (about 20 400 ppm), which may over time appear as
sediment in the deodorized oil. This appears to be dependent on seed growing con-
ditions. For the sake of simplicity, the term "waxes" is applied to these compounds,
but it is known (99) that about 2040% of these compounds are not wax esters. The
crystallization behavior of the mix of compounds that can crystallize from canola
oil is unpredictable in that low concentrations (<50 mg/kg) can sometimes show up
in the oil as a sediment and that the sedimentation can occasionally occur in a few
days, or take several months.
For some markets, it is desirable to dewax the oil to avoid any chance of a hazy
appearance. The process is carried out on the bleached oil by chilling in a contin-
uous heat exchanger to about 5 C and metering about 0.1% of a filter aid into the
chilled oil stream on the way to a filter. Little or no retention time is given before
filtration. This reduces the wax content to <50 mg/kg, which then usually no longer
produces a visible haze. Usually, in-house dewaxing process designs are used, or
processes patterned after sunflower oil dewaxing installations. Throughput rates
with canola oil are much higher than with sunflower oil, primarily because of the
much lower wax content.
6.2.7. Deodorization In edible oil processing, this is the final refining step. Its
primary function is to remove compounds from the oil, which impart odor and taste
typical of the seed from which it is derived and any odoriferous compounds formed
in such processes as bleaching and hydrogenation. Further, at the temperatures
required, heat bleaching of the yellow-red carotenoid compounds is an important
aspect of the process. As mentioned earlier, deodorization can also serve to physi-
cally refine the oil as an alternative to the removal of free fatty acids by alkali
refining. Deodorization is essentially a steam distillation of the odor and flavor
compounds from the oil, as well as other relatively volatile compounds such as
free fatty acids. Tocopherols and sterols are also to some extent removed. In the
deodorization of canola oil, it is important that the oil be essentially free of chlor-
ophyll and phosphatidic material and of heavy metals such as iron and nickel. The
oil is heated to 225260 C under very low pressure (0.10.5 kPa) to exclude air.
This increases the volatility of the compounds to be removed. Steam is blown
through the oil (13%), which increases the volatility further and allows efficient
removal of the volatiles to very low levels. Heat bleaching requires a minimum
of about 30 minutes at deodorizing temperatures. It is usual practice to cool the
deodorized oil to about 6080 C after the process and to sparge it with nitrogen
gas before storing in tanks, or shipping in bulk. Specification for the deodorized
canola oil are presented in Table 18.
TABLE 18. Nonhydrogenated, Deodorized Canola Oil.
Characteristic Value
Color (Lovibond red, 5.25 inch) 0.3 1.5

Flavor bland
Peroxide value (meq/kg) Nil
Anisidine value 0.5 2
AOM stability, minimum 15 h/100PV
6.2.8. Crystallization, Packaging, and Tempering With the exception of bulk

frying fats and salad oils, margarine and shortening oils are converted into semi-
solid forms for final use. Margarine oils are emulsified with an appropriate water
phase first. Shortenings, which do not usually involve a water phase, are crystallized
directly. Products high in hydrogenated canola oil require special care to ensure that
crystallization to the very small beta prime crystals is achieved. Essentially, this
means very rapid and complete crystallization before the product enters the pack-
age, that is, very little or no post-packaging crystallization. Tempering of crystal-
lized fat products refers to the process of allowing certain equilibrium processes in
the crystalline/liquid fat mixture to reach completion. The properly crystallized,
packaged product is held for two to four days at 2527 C. This improves plasticity,
creaming, and baking performance. It is a process that is not completely under-
stood, but it is very important for optimum shortening properties. Margarines are
usually tempered for a shorter time and then refrigerated. deMan et al. (93) studied
the effect of commercial crystallization conditions and product tempering in con-
nection with studies on the crystallization behavior of margarines and shortenings
made from hydrogenated canola oil.
6.2.9. Quality Assurance In the processing of canola oil to edible oil products,
much the same quality control procedures are applied as with other oils. A few
aspects, such as for example the presence of chlorophyll derivatives in crude oil
and their removal in processing, are somewhat unique. AOCS (76) or other standard
methods, such as IUPAC, ISO, or DGF are commonly used.
6.2.9.1. Degumming Quality control in the process requires the analysis of the
crude oil and the degummed oil for phosphorus.
6.2.9.2. Alkali Refining This involves the determination of free fatty acids, phos-
phorus, and, if desired, chlorophyll, before and after the process. In addition, soap
concentration is determined after water washing. Soap determination can present
some difficulty in establishing the endpoint of the titration with oils high in
chlorophylls, because of the dark color of such oils.
6.2.9.3. Bleaching The main aspect in the control of the bleaching process
is determination of chlorophyll. It is accepted in the industry that chlorophylls
92 CANOLA OIL
derivatives must be reduced in the bleaching step to a concentration of <0.025 mg/kg,

measured as chlorophyll a. Lower concentrations are innocuous. Higher concentra-
tions produce a noticeably green tinge in the oil and impair the stability of the oil
when exposed to light. In addition, during deodorization, a grayish tinge may devel-
op in the oil. The AOCS method for chlorophyll determination, Cc 13d-92 (96), is
used. Lovibond red and yellow color is, of course, also measured as part of the
process control in bleaching, but the colors to be achieved may vary considerably
depending on the heat-bleachability of the red-yellow color compounds. When
super-degummed oil is bleached to prepare it for physical refining, phosphorus
determination may also routinely be done on the bleached oil.
6.2.9.4. Hydrogenation Process control in hydrogenation is primarily concerned

with the degree of hydrogenation, determination of melting behavior, fatty acid
composition, and nickel catalyst removal. Rapid analysis for control of the degree
of partial hydrogenation makes use of iodine value and, more frequently, refractive
index. This is followed up with determining the solid fat index (SFI), or solid fat
content (SFC) and, where appropriate, the Mettler dropping point. Determination of
the fatty acid composition of hydrogenated oils is by the AOCS Method Ce le-91
(96). Nickel traces are determined by the nickel sulfide test or by ICP spectroscopy.
6.2.9.5. Dewaxing The dewaxing process at present is not controlled by an ana-

lytical test that determines the content of waxes in the oil. This would be too
complex for routine process control. Rather, the oil is subjected to the AOCS
cold test, and this is supplemented by keeping samples at higher temperatures,
up to room temperature (76). In this way, it is possible to gain some confidence
that the dewaxing procedure is in fact performing as expected. As pointed out ear-
lier, the crystallization behavior of the compounds implicated in sediment forma-
tion is complex. This makes these indirect tests vague as predictors of shelf stability
with respect to the oil remaining clear of visible sediments.
6.2.9.6. Deodorization/Physical Refining Quality control is especially concerned

with flavor, free fatty acid concentration, color, stability, and trace contaminants.
With respect to routine flavor testing, twothree trained panelists test the flavor
of small sample of the oil. The oil must be essentially bland, but very slight beany
or grassy notes are tolerated. Free fatty acids, color, and stability testing is done by
the AOCS official methods (76).
6.2.9.7. Crystallization, Packaging, and Tempering Process control of the crys-

tallization conditions (temperature of precooling, temperature of crystallizing, tem-
perature rise in the package) and of tempering time and temperature are especially
important with margarines and shortenings containing a high proportion of hydro-
genated canola oil. This is because of the greater tendency of such products to
recrystallize over time from the beta prime to the beta prime form. Polarized light
microscopy is sometimes used to provide an immediate assessment of the crystal
state of a product. Beyond that, penetrometer readings as well as taste panels for
NUTRITIONAL PROPERTIES OF CANOLA OIL 93
margarine to assess mouthfeel, and for bakery shortenings, creaming, baking, and
icing performance tests, are the procedures used as a guide to determine tempering
time and shelf stability.
7. NUTRITIONAL PROPERTIES OF CANOLA OIL
7.1. Nutritional Significance of Canola Oil Composition

Dietary fat serves several important nutritional functions. It is the source of essen-
tial fatty acids. Members of the n-6 and n-3 (also known as the o-6 and o-3)
families of fatty acids are important constituents of cell membranes and serve as
precursors of eicosanoids (biologically active compounds such as prostaglandins,
thromboxanes, prostacyclins, and leukotrienes). Fat also serves as a carrier for
the fat-soluble vitamins, and it is important source of energy. In addition, it has
important culinary properties and contributes to the palatability of food.
The current interest in dietary fat, however, stems primarily from its implication
in the origin of several chronic diseases. Interest has centered on both the amount
and type of dietary fat in the development of cardiovascular disease, cancer, hyper-
tension, and obesity. As a result, dietary recommendations in many countries call
for a reduction in total fat intake, to 30% of energy, and in saturated fat intake, to
less than 10% of energy. In addition, some nutrition recommendations specify
recommended levels of n-6 and n-3 fatty acids in the diets. Hence, the source of
fat in the diet has assumed considerable importance over the past few years. Interest
in the nutritional properties of canola oil developed because of its fatty acid com-
position (Table 2). Canola oil is characterized by a low level of saturated fatty acids,
a relatively high level of monounsaturated fatty acids, and an appreciable amount of
the n-3 fatty acid a-linolenic acid (18:3 n-3).
Saturated fatty acids. The adverse effect of saturated fat on blood cholesterol
level and its implication in cardiovascular disease has stimulated concern over
the level of saturated fatty acids in the diet. Canola oil contains a very low level
(<7%) of saturated fatty acids: about half the level present in corn oil, olive oil,
or soybean oil and about one-quarter the level present in cottonseed oil. Further-
more, canola oil contains only 4% of the saturated fatty acids (viz., lauric, myristic,
and palmitic) that have been found to increase blood cholesterol level. Hence,
canola oil fits well with the recommendation to reduce the amount of saturated
fat in the diet.
Monounsaturated fatty acids. The report that monounsaturated fatty acids (viz.,
oleic acid) were just as effective as polyunsaturated fatty acids (viz., linoleic acid)
in lowering plasma total and low-density lipoprotein (LDL) cholesterol (100)
aroused interest in the nutritional properties of canola oil. Canola oil contains
60% oleic acid and is second only to olive oil, among the common vegetable
oils, in oleic acid content. Although avocado oil and high-oleic sunflower oil
also contain high levels of oleic acid (>70%), they are minor constituents in the
average diet.
94 CANOLA OIL
There, also, is interest in dietary monounsaturated fatty acids because of their

possible protective effect against oxidation of LDL cholesterol (101). There is
appreciable evidence that the uptake of LDL cholesterol and the formation of fatty
streaks in the intima of large blood vessels, which is considered an early lesion of
atherosclerosis, is enhanced by the oxidation of the LDL cholesterol (102, 103).
LDL cholesterol was found to be appreciably more stable to oxidation when
subjects were fed diets rich in oleic acid than when fed linoleic acid enriched diets
(104106).
Polyunsaturated fatty acids. Canola oil is intermediate among the vegetable oils
in its polyunsaturated fatty acid content; it contains lower levels than corn oil, cotton-
seed oil, soybean oil, or sunflower oil but appreciably higher levels than olive oil or
palm oil. Interest in polyunsaturated fatty acids stems from their role as essential
fatty acids and their effectiveness in reducing plasma cholesterol level. Linoleic
acid and arachidonic acid, members of the n-6 family of fatty acids, have long
been recognized as essential fatty acids. Arachidonic acid is an important constitu-
ent of cell membranes and the precursor of eicosanoids, hormone-like substances
that are involved in a multiplicity of physiological functions ranging from blood
clotting to immune response.
a-Linolenic acid and other members of the n-3 polyunsaturated fatty acids like-
wise are essential. Docosahexaenoic acid (22:6 n-3) is a major constituent of lipids
of the brain and retina of the eye, whereas eicosapentaenoic acid (20:5 n-3) is the
precursor in the synthesis of an analogous but different series of eicosanoids from
those formed from arachidonic acid. Canola oil contains an appreciable amount
(10%) of a-linolenic acid. In addition, there is a favorable balance between linole-
nic acid and linoleic acid (a ratio of approximately 1:2). Soybean oil is the only
other major edible oil that contains a significant amount of linolenic acid (approxi-
mately 7%), but the ratio to linoleate is approximately 1:7.
Minor constituents: tocopherols and phytosterols. Vegetable oils are the primary
source of tocopherols in the average diet. Canola oil is a relatively rich source of
tocopherols (refer to Table 6); it is similar in total tocopherol content to corn oil,
cottonseed oil, safflower oil, and sunflower oil (6070 mg/100 g). Only soybean
oil contains an appreciably higher level of total tocopherols (100110 mg/100 g).
However, canola oil contains a considerably higher level of a-tocopherol, the main
source of vitamin E biopotency among the tocopherols, than soybean oil (27 vs.
12 mg/100 g). Only cottonseed oil, safflower oil, and sunflower oil are better
sources of a-tocopherol than canola oil (35, 56, and 61 mg/100 g, respectively)
(see Table 6).
Canola oil contains a relatively high level of phytosterols (892 mg/100 g), about
twice the level in soybean oil or sunflower oil (436 and 496 mg/100 g respectively)
(Table 8). b-Sitosterol accounts for about 50%, campesterol 35%, and brassicasterol
14% of the total phytosterols in canola oil. Canola oil is the only common vegetable
oil that contains brassicasterol. Plant sterols have been reported to lower plasma
cholesterol level (107) by inhibiting the absorption of dietary cholesterol and the
reabsorption of biliary cholesterol (108).
7.2. Effect of Canola Oil on Plasma Cholesterol and Lipoproteins

Appreciable research on the effect of canola oil on plasma cholesterol and lipo-
proteins has been reported. The primary impetus for this research was the finding
that dietary monounsaturated fatty acids were as effective as polyunsaturated fatty
acids in lowering plasma total and LDL cholesterol (100, 109). These findings also
provided a possible explanation for the observation that canola oil was as effective
as soybean oil in lowering plasma cholesterol in normolipidemic men (110).
Prevailing theory had held that saturated fatty acids raised plasma cholesterol, poly-
unsaturated fatty acids lowered plasma cholesterol, and monounsaturated fatty
acids were neutral, they neither raised nor lowered plasma cholesterol (111, 112).
Normolipidemic subjects. Studies in Canada, Finland, Germany, and the United
States have shown canola oil to be just as effective in lowering plasma total and
LDL cholesterol levels as fat sources containing high levels of polyunsaturated fatty
acids when each replaced saturated fat sources in the diet (Table 19). All diets
resulted in statistically significant decreases in plasma total (mean of 0.47 to
0.88 mmol/L) and LDL cholesterol (mean of 0.43 to 0.74 mmol/L) levels.
Thus, most of the decrease in total cholesterol was caused by a decrease in LDL
cholesterol, which is consistent with the main objective of dietary intervention
programs aimed at reducing the risk of cardiovascular disease. Furthermore, except
for the study by Valstra et al. (116), decreases in total and LDL cholesterol were the
same on the canola oil and the polyunsaturated fatty acid diets. Decreases in apo-
lipoprotein B, which is the lipoprotein characteristic of the LDL fraction, also were
similar on the canola and polyunsaturated diets.
None of the diets in these studies, except for the study by Kratz et al. (106), had
any effect on plasma high-density lipoprotein (HDL) cholesterol levels. Kratz et al.
(106) reported a decrease in plasma HDL cholesterol level on both the canola oil
TABLE 19. Comparison of the Effect of Canola Oil and Polyunsaturated Fatty Acid
Sources on Plasma Total and LDL Cholesterol of Normolipidemic Subjects.
% Change from Baseline

Dietary PUFA Plasma Lipid Baseline
Source Parameter (mmol/L) Canola Diet PUFA Diet Reference
Sunflower oil Total cholesterol 4.42 20 15

LDL cholesterol 2.76 25 21 113
Safflower oil Total cholesterol 5.39 9 15
Soybean oil Total cholesterol 4.40 18 16

Plasma levels on diets typical of usual fat intake.

Significant differences (p < 0:01) between canola and PUFA diets.
96 CANOLA OIL
and the sunflower oil diets, whereas no decrease occurred on an olive oil diet. The
levels of polyunsaturated fatty acids in the safflower oil, soybean oil, and sunflower
oil diets in these studies were two to three times the levels in the average diet. Men-
sink and Katan (109) also found no adverse effect of a relatively high intake of
polyunsaturated fatty acids with normolipidemic subjects fed diets based on cus-
tomary foods. By contrast, Mattson and Grundy (100) found that very high intakes
of polyunsaturated fatty acids (29% of total energy) resulted in a decrease in HDL
cholesterol level of mildly hyperlipidemic subjects fed formula diets, whereas
similar high intakes of monounsaturated fatty acids had no effect on plasma
HDL cholesterol level.
Hypercholesterolemic subjects. Canola oil was also effective in reducing serum
total and LDL cholesterol levels in subjects with increased blood lipid levels.
Replacing 50 g of fat in the regular diet with 50 g of canola oil mayonnaise resulted
in a significant decrease in serum total and LDL cholesterol levels (9 and 10%,
respectively) in a group of men with a mean serum cholesterol level of 7.1 mmol/L
(117). Serum triacylglycerol and very low-density lipoprotein (VLDL) cholesterol
levels also decreased, and serum HDL levels increased on the canola oil regimen.
Generally, type of fat, at the level of substitution used in this study, has not affect
serum HDL cholesterol levels. Bierenbaum et al. (118) also found that substituting
canola oil for 30 g of the usual oils and spreads in the diets of hyperlipidemic sub-
jects resulted in a decrease in plasma LDL cholesterol level. Plasma total choles-
terol, however, did not differ from baseline levels. Substituting canola oil or canola
oil margarine for butter, at much lower levels (approximately one-fifth of the total
fat and 8% of total energy) than those cited above, resulted in a decrease in the
plasma cholesterol level of mildly hypercholesterolemic subjects (119). Substitu-
tion of canola oil for butter resulted in a 9% decrease in total cholesterol and
13% decrease in LDL cholesterol level, and substitution of canola oil margarine
for butter resulted in a 9% decrease in total cholesterol and an 8% decrease in
LDL cholesterol level.
Canola oil also was effective in lowering plasma total and LDL cholesterol con-
centrations of mildly hyperlipidemic subjects fed a low-fat (30% of total energy)
diet (120). The decline in plasma total cholesterol when canola oil or corn oil
provided 20% of the total energy was 12% and 13%, respectively, which was
significantly greater than the decline when olive oil (7%) supplied the fat. How-
ever, canola oil, corn oil, and olive oil were equally effective in lowering plasma
LDL cholesterol levels (16%, 17%, and 13%, respectively); elevated plasma
LDL cholesterol level is a major risk factor in coronary heart disease.
7.3. Effect of Canola Oil on Thrombogenesis

Cardiovascular disease is characterized by three major events: (1) the formation of
atherosclerotic plaques on the intima of blood vessels, which reduce the size of the
lumen of the vessel; (2) thrombosis or clot formation, which is the event leading
directly to a coronary attack or stroke in many individuals; and (3) cardiac arrhythmias,
uncoordinated contractions of the heart muscle resulting in irregular and ineffective
TABLE 20. Effect of Canola Oil on Clotting Time and Factors Involved in Clot Formation.
Effect of Canola Oil

Parameter on Parameter Reference
n-3 Fatty acid (viz., 20:5n-3 and 22:5n-3) Increased 124126

Content of platelet phospholipids No change 127, 128
Arachidonic acid content of platelet phospholipids Decreased 124, 125, 127, 128
In vitro platelet aggregation Reduced 124, 127
Enhanced 128
Eicosanoid production
Thomboxane Decreased 113
Prostacyclin Increased 113
Clotting time Increased 113
heart beats that frequently results in death. Clot formation has only recently
received attention from researchers even though dietary fat has been implicated
for some time (121). Marked differences among Greenland Eskimos and Danes
in the incidence of coronary heart disease (121) led to interest in the effect of
long-chain n-3 polyunsaturated fatty acids on thrombogenesis. The long-chain n-
3 fatty acids of fish oil, namely, eicosapentaenoic acid (EPA) and docosahexaenoic
acid (DHA), have been found to inhibit platelet aggregation (123). A possible
mechanism for the effect of long-chain fatty acids on thrombogenesis is thought
to be a change in the fatty acid composition of platelet phospholipids and an accom-
panying change in eicosanoid formation.
Canola oil has been shown to alter several parameters linked to clot formation
(Table 20). In general, phospholipid fatty acid composition reflects the fatty acid
composition of the diet: Canola oil resulted in higher levels of oleic acid, whereas
safflower oil, soybean oil, and sunflower oil resulted in higher levels of linoleic acid
(125129). Canola and soybean oil also resulted in higher levels of a-linolenic acid,
although very little linolenic acid was incorporated into phospholipids even when
flaxseed oil was incorporated into the diet (126, 130). Dietary linolenic acid, how-
ever, has been found to alter the long-chain polyunsaturated fatty acid composition
of platelet phospholipids. Several groups (124, 125, 127, 128) reported lower levels
of arachidonic acid with canola oil diets. Canola oil diets also resulted in higher
levels of eicosapentaenoic acid and docosapentaenoic acid than soybean oil, sun-
flower oil, or customary mixed fat diets (124126, 129).
Not all studies, however, found higher levels of long-chain n-3 polyunsaturated
fatty acids with canola oil diets (127, 128). Part of the explanation for the apparent
discrepancy among studies may relate to the balance of fatty acids in the diet.
Chan et al. (126) found that changes in plasma and platelet fatty acid composition
varied not only with the level of linolenic acid in the diet, but also with the ratio of
linolenic to linoleic acid. Canola oil was found to reduce in vitro platelet aggrega-
tion although sunflower oil and safflower oil (i.e., low level of linolenic acid and
high level of linoleic acid) also reduced in vitro platelet aggregation (123, 127).
98 CANOLA OIL
On the other hand, Mutanen et al. (128) found that canola oil and sunflower oil
enhanced in vitro platelet aggregation. Similarly, McDonald et al. (113) found
that both canola oil and sunflower oil increased prostacyclin production (an antiag-
gregating eicosanoid) and decreased thromboxane production (a proaggregating
eicosanoid). Although the effect of canola oil on platelet activity and clot formation
is not as well established as its favorable effect on plasma cholesterol and lipopro-
tein levels, there is evidence that it may impede clot formation.
7.4. The Lyon Diet Heart Study

Coronary heart disease is a complex, multifactorial disease, and the mechanism by
which diet affects its development has not been resolved fully. A clinical study, the
Lyon Diet Heart Study (131), which compared the effect of a Mediterranean-type
diet with that of a prudent postinfarction diet, found a marked reduction in
coronary events (8 vs. 33) among heart patients assigned to the Mediterranean
diet. Both cardiac deaths and nonfatal myocardial infarctions were reduced on the
Mediterranean diet, in which olive oil and canola oil were the prescribed dietary oil
sources and butter was replaced by canola oil margarine. Of particular interest was
the absence of any differences between the patients on the Mediterranean diet and
the prudent diet in terms of serum total, LDL, or HDL cholesterol levels. Likewise,
there was no difference between the groups in platelet aggregation even though
plasma levels of oleic, linolenic, and eicosapentaenoic acids were significantly high-
er and stearic, linoleic, and arachidonic acids were significantly lower for the patients
on the Mediterranean diet. The study was terminated after a mean of 27 months
follow-up of patients assigned to the diets. However, an extended follow-up of
the patients in the original clinical study (mean of 46 months per patient) by de
Lorgeril et al. (132) found the benefits of the Mediterranean diet were still very evi-
dent (44 coronary events on the prudent diet vs. 14 on the Mediterranean diet).
Although the results of this study are very encouraging, there is need to confirm
them in further clinical trials.
7.5. SummaryNutritional Properties of Canola Oil

Canola oil is characterized by a low level of saturated fatty acids (less than 4%
palmitic acid) and relatively high levels of oleic acid (60%) and a-linolenic acid
(10%). It is second only to olive oil, among the common fats and oils, in oleic
acid level and, except for soybean oil, the only common dietary fat that contains
a significant amount of a-linolenic acid. Furthermore, there is a favorable balance
in the levels of linolenic and linoleic acids (viz., 18:3/18:2 ratio of 1:2) in canola
oil. Canola oil has been found equally as effective as soybean oil, safflower oil, and
sunflower oil in reducing plasma total and LDL cholesterol levels in normolipi-
demic subjects. It also was effective in reducing plasma total and LDL cholesterol
levels in hyperlipidemic subjects when it replaced saturated fat in their diets. Cano-
la oil diets also have been shown to affect the fatty acid composition of blood
MAJOR FOOD USES 99
platelet phospholipds and to alter platelet activity and thrombogenesis, although the
evidence supporting these observations is not as convincing as its effect on plasma
cholesterol levels.
8. MAJOR FOOD USES
8.1. Standard Canola/Rapeseed Oil

In describing canola/rapeseed oil food uses, the Canadian experience is of signifi-
cant interest for a number of reasons. First, canola/rapeseed was originally devel-
oped and introduced in Canada commercially so that considerable experience in
using canola oil in edible oil products has been accumulated over a longer period
of time. Second, canola, after its introduction, rapidly became the most important
oilseed crop and the most heavily used edible oil in Canada, as documented below.
Third, the Canadian edible oil products market demanded a variety of high-quality
products, which led edible oil producers to develop many uses for canola oil as well
as find applications especially suited for it.
With respect to the importance of canola oil in Canada, usage has grown from
the early years after its introduction to about 68% of the edible vegetable oil con-
sumed in 2000. It has been at this level for the last decade. Thus, in 1992, 1993,
1994, 1995, and 1999, the corresponding percentages were 63%, 68%, 73%, 72%,
and 68%, respectively (Adapted from COPA Monthly Statistics for Feb. 1993,
1994, 1995, 1996, 2000, and 2001). Most of the oil is used as a liquid, that is,
the nonhydrogenated form (probably >70%). The products using liquid oil are
primarily salad oils and salad dressings, then, as the liquid oil component in
margarine formulations and in household and baking shortenings.
Very lightly hydrogenated canola oil (IV about 90) and more highly hydroge-
nated canola oil (IV lower than 90) are used for frying and in margarine and
shortenings.
8.1.1. Salad Oil, Salad Dressings, Mayonnaises, and Cooking Oil Uses
Canola oil is an natural salad oil. This means that it remains clear (no sedimen-
tation) at refrigeration temperatures (35 C). No winterization or fractionation is
required, except in some instances when, because of seed growing conditions,
the oil may be contain waxes and traces of other high-melting material. These
compounds may crystallize over time and create appearance problems in clear bot-
tles. Experience has shown that only for the most demanding markets it is necessary
to remove these compounds. They present no health hazard and are not sufficiently
concentrated to affect emulsion stability when the oil is used in mayonnaises and
other emulsified salad dressings.
Canola oil is used either pure or, increasingly, in some markets, as a component
in salad oil blends of several oils. Such blended salad oils are usually aimed at
achieving a certain fatty acid composition for nutritional reasons. Canola oil con-
tributes low saturated fat and some linolenic acid.
100 CANOLA OIL
Canola oil is also used as cooking oil, including pan-frying. It is not recom-
mended for deep-fat frying. The reason for this is that its polyunsaturated fatty
acid content, which, even though only moderately high, makes it unsuitable.
In all of these applications, canola oil has gained favour in many areas of the
world in the 1980s and 1990s, including North America and many European coun-
tries. This is because of its low content of saturated fatty acids compared with all
other competing oils in these applications. Further, its high content of oleic acid and
its moderate content of linoleic acid makes canola oil even more competitive in
food applications. Linoleic acid consumption is recognized as being too high in
diets that are high in fat and high in the use of soybean, sunflower, and corn oils,
such as is the case in most industrially developed countries. Its linolenic acid con-
tent is recognised as a nutritional advantage over sunflower and corn oils.
The low total polyunsaturation of canola oil, about 30% versus 58% for soybean
oil, along with the high content of monounsaturates, about 60% versus about 25%
for soybean oil, are responsible for the good flavour stability of this oil, despite the
presence of linolenic acid. Additional minor, but important reasons, for better oxi-
dative stability of canola oil compared with soybean oil are as follows:
1. That a larger percentage of its linolenic acid content is in the sn-2 position in
the triacylglycerols than is the case with soybean oil; this confers somewhat
greater resistance on linolenic acid to oxidation.
2. The presence of some sulfur compounds, which can act as antioxidants (51).
Detailed fatty acid composition of canola, soybean, sunflower, corn (maize), and
flax oils as well as some specialty canola oils and HEAR oil are given in Table 2.
8.1.1.1. Frying Fats Large amounts of canola oil are used as lightly hydroge-
nated (IV 90), stable, but pourable frying fat. Canola oil is uniquely suited to
combine good stability with pourability because of its fatty acid composition.
Low total polyunsaturation requires relatively little hydrogenation to reduce the
levels of linolenic acid and linoleic acid to values that are low enough to confer
good frying stability. During this process, very little stearic acid is formed. Coupled
with the low original content of saturates of only about 6%, an oil of good stability
yet still pourable at room temperature is obtained. In Table 21, the fatty acid com-
position and the solid fat indices (representative values) of lightly hydrogenated
canola oil with 2% residual linolenic acid and IV of about 90 are given along
with the fatty acid composition of the nonhydrogenated oil. The data for soybean
oil are also included to show the difference in fatty acid composition, especially in
the amount of saturates and polyunsaturates, as well as in the solid fat content of
these two lightly hydrogenated oils. The data show the advantage of canola oil with
respect to oxidative stability, which is caused by the much lower content of poly-
unsaturated fatty acids, and the advantage of pourability, which is caused by a much
lower solid fat present at 10 C and 21.1 C (room temperature). This type of oil is
very suitable and heavily used for small-scale frying in restaurants as well as in
MAJOR FOOD USES 101
TABLE 21. Fatty Acid Compositions and Solid Fat Indices of Lightly Hydrogenated
Canola and Soybean Oil (Residual C18:3 Content, 2 %).
Iodine Fatty Acid Composition (w/w %) Solid Fat Index at C
Value C16:0 C18:0 C18:1 C18:2 C18:3 Trans 10.0 21.1 26.7 33.3
Canola
115 4.5 2 60 22 10
90 4 4 79 9 2 25 23 0 0 0
Soy
130 11 4 25 53 7 0 none
95 11 7 54 25 2 15 8 4 0 0
large industrial frying operations. A negative aspect of the hydrogenated canola

frying fat is the somewhat higher concentration of trans-isomers compared with
soybean oil.
The lightly hydrogenated canola oil of IV 90, or slightly lower to produce a fry-
ing fat containing about 1% linolenic acid, has also been shown (133) to be useful
as a winterized, very stable salad oil. A very significant advantage is that this sta-
bilized salad oil is obtained at a yield of about 95% and a 12-hour cold test, com-
pared with lightly hydrogenated, winterized soybean oil at a yield of 7080% and a
6-hour cold test. In some areas of the world, this type of product, based on soybean
oil, is being used, and seems to be expanding. This processing was very popular in
the United States in the 1960s, but it was discontinued because of the low fractio-
nation yields.
It is important to note that lightly hydrogenated canola oil, such as listed in
Table 21, does not contribute significantly to the fat crystal matrix of fat products
in which it is used. In this respect, it is similar to the use of liquid canola oil; that is,
it can be used in products such as margarine and shortenings without contributing to
beta crystallization problems.
Canola oil is also used in more highly hydrogenated forms to produce very stable
frying fats that are essentially free of any significant amounts of polyunsaturates,
but with high amounts of oleic and elaidic acids and moderate amounts of saturates.
Examples of more highly hydrogenated canola oils are given in Table 22, especially
TABLE 22. Fatty Acid Compositions and Solid Fat Indices of Highly Hydrogenated
Canola Oils (%).
Iodine Fatty Acid Composition Solid Fat Indices at C
Value C16:0 C18:0 C18:1 C18:2 C18:3 Trans 10.0 21.1 26.7 33.3 40.0
82 4 5 87 2 <1 32 18 5 0 0 0
77 4 9 84 <1 <1 35 25 11 6 0 0
72 4 13 81 0 0 44 35 17 10 3 0
68 4 18 76 0 0 48 53 34 27 13 1
62 4 25 70 0 0 46 65 53 48 33 12

102 CANOLA OIL
the oils in the IV range from 82 to 72. These oils are very low in polyunsaturated
fatty acids, about 2% linoleic acid, or are entirely free of this acid. They are very
stable, and especially suited when a stable, but relatively firm fat is needed, such as
needed in donut frying. In frying applications, the problems with beta-crystalliza-
tion, formation of a grainy texture, and mouthfeel, are usually not important, unless
the deep-fried foods are stored for some time. In these cases, large fat crystals may
become visible at the food surface, which is undesirable. Blends with hydrogenated
soybean oil, hydrogenated cottonseed oil, or with palm oil are used to control
beta-crystallization problems. Additional methods that are useful for suppressing
beta-crystallization are discussed below in connection with margarine products.
In Canada, only selectively hydrogenated canola oil is used. The practical reason
is that the somewhat higher content of trans-isomers makes the more highly hydro-
genated oil more resistant to beta crystallization compared with nonselectively
hydrogenated canola oil. When trans-isomer content must be minimized in canola
oil products, liquid canola oil or very lightly hydrogenated canola oil is used, such
as shown in Table 21.
8.1.1.2. Soft (Tub) Margarine In Canada and in many other countries where mar-
garine is consumed, soft or tub margarine is now predominant. Consumption of
hard or stick margarine has decreased significantly over the last two decades.
Further, a significant proportion of soft margarine in many countries is of the
zero trans-isomer type as no hydrogenated oils are used. Palm/palm-kernel oil-
based hard fat blends with a suitable solid fat content profile are used to provide
the crystalline fat component. Only about 610% of the total fat are required to
supply the necessary crystalline fat phase. The composition of the palm/palm-
kernel oil-based hard fats varies depending on a variety of factors and is very often
proprietary to the suppliers, which are primarily based in Malaysia. Interesterifica-
tion is usually used to produce these blends. Canola oil is used only as the liquid oil
component of these no trans-isomer products. The prime example of these products
is BecelTM, which has been on the market for many years in parts of Europe and in
Canada. Canola oil is often the preferred liquid oil, because it has the lowest satu-
rated fatty acid content and it supplies some linolenic acid, and yet has good flavor
stability. This oil is also used together with other liquid oils to make up the liquid
oil component.
Soft margarine based on partially hydrogenated hard fat is also produced with
canola oil as the sole liquid, nonhydrogenated component, or along with other
liquid vegetable oils. In these margarines, canola oil can be used as the partially
hydrogenated hard fat as well. One, or a combination of several, of the hydroge-
nated oils listed in Table 21 and 22 can be used. But, because of the tendency of
canola-based hydrogenated hard fat to form beta crystals over time, partially hydro-
genated soybean oil, or other, palmitic acid-containing oils, such as partially
hydrogenated cottonseed oil, are preferred as the hard fat. This avoids the tendency
of the margarine to have a coarse, sandy texture caused by beta crystallization.
Table 23 lists typical compositions of these three types of soft margarine using
canola oil.
MAJOR FOOD USES 103
TABLE 23. Soft (Tub) Margarine Using Canola Oil.
Solid Fat Indices at

Composition
Type Oils Used % 10.0 C 21.1 C 33.3 C
1. No trans a) palm/palm kernel hard fat 8 }

b) canola oil, or other 92 } 57 45 12
liquid vegetable oils,
including blends,
to achieve specific
fatty acid compositions
2. Trans - containing a) hydrogenated canola hard fat 47 }
100 % canola b) canola, or other liquid 53 } 810 68 12
vegetable oils, as above
3. Trans - containing a) hydrogenated soy hard fat 25 }
b) canola, or other liquid 75 } 810 68 12
vegetable oils, as above
In the past, margarine oils made entirely from canola oil were produced espe-
cially in Canada and Sweden. This is still of interest when the hard fat must be
made from canola oil because of cost, or oil availability. Using partially hydroge-
nated canola hard fat requires control of beta crystals, which must be suppressed.
This can be done by (1) adding about 0.30.5% of trisorbitan stearate to the fat
blend and (2) by using several selectively hydrogenated canola hard fats of different
hardness to introduce a greater variety of triacylglycerols into the fat blend. Trisor-
bitan stearate and the greater variety of triacylglycerols interfere with the conver-
sion to the beta form of fat crystals and, therefore, retard the formation of a sandy
margarine texture. Changing the fatty acid position in triglycerides by interesterifi-
cation of some or all of the hydrogenated canola hard stock used in the oil formula-
tion can reduce beta crystallization. It is used in Europe, but not in North America.
All of the above-mentioned measures to control beta-crystallization are costly
and, in the case of trisorbitan stearate, requires a label declaration as an additive,
giving the product an undesirable chemical connotation; hence, its use is limited. In
todays practice, blending with another hard fat that is relatively high in palmitic
acid to raise the concentration of this acid to at least 8% in the final blend is pre-
ferred. To meet this requirement, the oils used are palm at about 15% and partially
hydrogenated cottonseed at about 30%. Further, when it is not required to maximize
canola oil in the fat blend, partially hydrogenated soybean oil is used to supply the
high melting portion of the blend instead of partially hydrogenated canola oil. Soy-
bean oil contains 11% of palmitic acid, enough to confer acceptable crystal stability
in the beta prime form.
8.1.1.3. Hard (Stick) Margarines Liquid and lightly hydrogenated canola oil is
used to produce stick margarine. They do not contribute significantly to the crystal
matrix of the product, as pointed out earlier. The tendency of the more highly
104 CANOLA OIL
hydrogenated canola oils, such as those listed in Table 16, to form beta crystals
means that these are not used to any significant extent as contributors to crystalline
fat. They would have to be used in relatively large amounts, 4050% in the fat
blend, which makes it more difficult and costly to control beta crystallization.
The liquid or lightly hydrogenated oil is generally used at 50% to 60% of the oil
composition, the remainder being relatively highly hydrogenated soybean or
cottonseed oil, or hydrogenated palm oil.
In situations where the use of canola oil must be maximized, that is the hard fat
component as well as the more liquid component must be from canola oil, the same
approach as outlined above for soft margarine must be used. As was already pointed
out, demand for stick margarine is declining, and the need for margarine made
entirely from canola oil is no longer significant even in the countries where canola
oil is used as the main oil.
8.1.1.4. Shortenings, Baking, and Pastry Margarine Similar to the use of canola
oil in making table margarine, liquid canola oil is heavily used to produce shorten-
ing and baking and pastry margarines. Lightly hydrogenated canola oil of about IV
90 is used when better oxidative stability is required than can be achieved with
liquid oil. When it is desired to use canola oil even as the hard fat component in
these formulations, then the considerations related to crystallization as discussed
for margarine apply. Generally, highly hydrogenated canola oil is not much used.
The liquid canola oil or lightly hydrogenated canola oil is blended with hard fats,
such as tallow, palm, partially hydrogenated soybean and cottonseed oils, and fully
hydrogenated versions of these oils (stearine), to meet preferred specifications.
Detail compositions are usually proprietary. For baking applications shortening
and baking margarine, having the fat component with the beta prime crystalline
form is especially important for good performance. For this reason, baking short-
enings based totally on canola oil are not in use.
8.1.1.5. Antioxidant Usage in Canola Oil Products Present practice is to avoid

the use of synthetic antioxidants in edible oil products as much as possible. This
also applies to canola oil products. For many edible oil products, users now specify
that chemical additives cannot be used as ingredients. This avoids having to declare
on the product label that the product contains a synthetic, chemical antioxidant as
an additive. Instead, there is increasing emphasis on preserving during processing
the natural antioxidants present in canola oil, the tocopherols. Tocopherol losses
occur primarily during deodorization of the oil, because of the high temperatures
required. Processors are limiting deodorizing temperatures as much as possible
when high tocopherol concentrations are required, consistent with achieving
good deodorization. It is possible to retain as much as 80% of the original tocophe-
rol content in the deodorized oil.
The tocopherols are especially important as antioxidants in frying, because of
their low rate of evaporation and low rate of destruction at frying temperatures
(134). The tocopherol content of standard canola oil is given in Table 6, together
with the tocopherol content of some specialty canola oils and other common
MAJOR FOOD USES 105
vegetable oils. Alpha and gamma tocopherol are the main tocopherols present in
canola oil.
It is interesting to note that canola oil is more than twice as high in alpha toco-
pherol (about 270 mg/kg) than soybean oil (about 116 mg/kg). Alpha tocopherol is
now recognized as the main tocopherol with vitamin E function in humans (135).
Canola oil is a very good source of vitamin E.
The synthetic antioxidants that are used, when there are no restrictions on using
them, are butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT),
propyl gallate (PG), and tertiary butylhydroxyquinone (TBHQ). The usual added
amounts are 0.02% as single or combined. This amount is the maximum allowable
in Canada and the United States, for example. Regulations governing synthetic
antioxidant usage can differ from one country to another.
Citric acid is used as a metal chelating agent, usually as monoacylglycerol
citrate at 0.01% either by itself or along with BHA and BHT. In some countries,
the United States, for example, citric acid can be used as a chelating agent without
having to declare it on product labels when it is added to the oil in an aqueous solu-
tion in the cooling stage of the deodorization process. Lower amounts of 0.005%
are often used.
8.1.1.6. Canola Oil Use in Selected Areas of the World It is interesting to

review, briefly, the use of canola oil in edible oil products in various countries in
the world.
The United States of America import large amounts of canola oil from Canada in
addition to some domestic production. At the time of writing, about 90% of the
canola oil was consumed in liquid form as salad oil and in salad dressings. This
is a direct result of the emphasis on consuming oils, which are low in saturated fatty
acids, and canola oil is lowest in saturated fat among vegetable oils. Canola oil
is also being used in blended salad oils to achieve certain fatty acid profiles, as
mentioned earlier. The relatively high use of canola oil in the United States is
remarkable, because none was used before 1983.
About 10% of canola oil is hydrogenated and consumed in the form of shorten-
ing. This appears to be mostly as frying fats, also termed frying shortenings in
North America, to take advantage of the low concentration of saturated fat and
low polyunsaturation as discussed before. Interestingly, virtually no canola oil,
liquid or hydrogenated, was used in margarine formulation at the time of this
writing.
Mexico uses significant amounts of canola oil, predominantly as salad oil, salad
dressings, and cooking oil. This is mostly from seed imported from Canada and the
European Union.
Japan uses large amounts of canola oil. It has been an importer of Canadian
canola seeds since the introduction of canola in the early 1970s, and it has for
many years taken about one-half of the Canadian canola crop. The oil is predomi-
nantly used in liquid, nonhydrogenated form as cooking oil, salad oil, and salad
dressings oil, pure, as well as in blends with other oils. Further, in a new develop-
ment, it is used as base oil to produce dietetic cooking and salad oils made up of
106 CANOLA OIL
about 80% diacylglycerols (DAG oil). DAG oil is more easily hydrolyzed in the
digestive system and is then used as an immediate energy source rather than stored
as body fat. It is worth noting that the DAG oil technology is being introduced into
the United States (136).
In China, canola-type rapeseed oil products still contribute a very small propor-
tion of total rapeseed oil products. Oil from both high erucic acid rapeseed and
canola rapeseed represent the largest use of edible oil at present. The oil from these
two sources is almost entirely used as cooking oil. There are very little amounts of
this oil used for margarine or shortening formulations at present. Efforts are being
made to widen the spectrum of edible oil products and convert from HEAR cultiva-
tion to canola cultivation.
India, similar to China, does not use significant amounts of canola-type rapeseed
oil. Instead, mustard seed oil, which is high in erucic acid, similar to HEAR oil, is
the most important oil used. This is used almost entirely as a liquid cooking oil. At
least among the lower income segment of the population, and even in the middle
class, this oil is used undeodorized and is favored for its taste. Undeodorized fresh
canola/rapeseed oil cannot compete with mustard seed oil in flavor. Some blending
of canola oil with mustard oil is done to lower the content of erucic acid in the mus-
tard oil. Unless taste preferences change and there is greater attention to the health
implications of the types of fat in the diet, canola oil will be used only to a limited
extent in the foreseeable future.
The Middle East is beginning to use canola oil in competition with sunflower
and corn oil as salad oil and salad dressings/mayonnaise. Margarine and vanaspati
are in many countries based on hydrogenated soybean oil and on palm oil at pre-
sent. Interest in canola oil is based on its nutritional properties, mainly its low satu-
rated fatty acid content. There is considerable potential for using canola oil not only
as a salad oil, but also, in its lightly hydrogenated form, as a vanaspati-like pourable
frying fat.
Western and Eastern Europe use large amounts of canola oil, with the exception
of France. Salad oils, salad dressings, mayonnaises, and table and baking margarine
use large amounts of liquid canola oil. Lightly hydrogenated canola oil is heavily
used for frying snacks. This is very similar to the Canadian practice. Canola oil use
is driven, in part, by recognition of the positive health effects of its high oleic acid
content, along with its low saturated fat, and the fact that European grown canola
seeds are not genetically modified (non-GM) at present, as opposed to oil from
imported soybeans. In addition, canola salad oil is considered by some to have a
better shelf life than other more highly polyunsaturated vegetable oils.
Southern Europe and France use relatively little canola oil. Instead, olive,
sunflower, and peanut oils predominate. In the case of France, this is somewhat
surprising, because this country is a large producer of canola seeds. But France
uses large amounts of canola oil for biodiesel in the form of fatty acid methyl esters.
Australia/New Zealand produces canola seed and uses the oil in much the same
fashion as North America and parts of Europe.
South America uses sunflower and soybean oil and increasingly palm oil. Canola
oil is not used in food in South America.
MAJOR FOOD USES 107
8.2.1. Canola Rapeseed Oils with Modified Fatty Acid Composition Since
the introduction of standard canola, there has been considerable plant breeding
efforts to produce canola oils with modified fatty acid compositions. These efforts
were primarily to improve oxidative stability, or crystallization properties, or even
produce lauric acid-containing oils and, more recently, canola oil containing
gamma linolenic acid (11). The following is a list of these developments:
Low linolenic acid canola oil (2% vs. 9%)

High oleic acid canola oil (6977% vs. 60%)
High palmitic acid canola oil (10% vs. 4%)
High stearic acid canola oil (30% vs. 2%)
Lauric acid canola oil (about 33% C 12:0)
Gamma linolenic acid canola oil (37% vs. about 1%).
The complete fatty acid compositions of some of these oils, namely, low linolenic
and high oleic acid, lauric acid, and gamma linolenic acid oils are given in Table 2.
Low linolenic acid canola oil was developed in Canada in the 1980s to improve
the oxidative stability of the oil so that light hydrogenation would not be necessary
(7). The reduced linolenic acid content of this oil of about 2% compared with
about 9% in the standard canola oil. This resulted in an increase in linoleic acid
from 20% to 27% and the increase in oleic acid from about 60% to 61%. In Canada
and the United States, this oil is available in limited quantities and is used entirely
for deep-frying in place of the lightly hydrogenated standard canola oil (Table 21).
Its main advantage is the much lower trans-isomer contents of about 13%,
which are formed during deodorization, whereas the lightly hydrogenated oil
contains 2025%. Widespread use of this oil is, however, hampered by its price,
which tends to be too high because of the low seed yields of the available varieties.
Research on its frying stability and the storage stability of french fries by Warner
and Mounts (9) showed that these properties were improved for low linolenic acid
canola oil as compared with the standard canola. Zambiazi (34) and Normand et al.
(35) showed no significant improvement in the frying stability of this oil and the
storage stability of fried foods. It was found that this might be related to the lower
content of tocopherols (Table 6). There are also anecdotal reports from industry
that the frying stability of the oil is not sufficiently improved to warrant its higher
price.
High oleic acid canola oil is another development pursued in Canada, the United
States, Sweden, Australia, and elsewhere (137). As with low linolenic acid canola
oil, the aim was to produce stable frying oil, which will not need hydrogenation and
thus avoid trans-isomers formation. The oleic acid content in oil from seed devel-
oped in Canada is at about 78%, whereas linoleic and linoleic acids are lowered to
approximately 8% and 3% respectively (see Table 2). Saturated fatty acid content is
unchanged from the standard canola oil. There is limited commercial seed produc-
tion for export to Japan. Also, there is increasing acceptance of the oil in Canada
and the United States. The frying performance in tests was found to be similar to
108 CANOLA OIL
lightly hydrogenated standard canola and mid-oleic sunflower oil. Taste tests of
french fries produced with this oil showed similar consumer acceptance as typical
frying fats used for making french fries (Przybylski, unpublished data). In Australia,
canola oil with 69% oleic acid (Monola) is being offered for frying. In potato frying
tests with ten other oils, it was rated higher in sensory and chemical tests than the
other oils (138).
High palmitic acid canola oil was initially developed in Sweden (139). The pur-
pose was to prevent the beta crystallization of hydrogenated canola oil to make it
more freely useable for margarine and shortenings. The oil contains about 1012%
palmitic acid compared with only 4% in the standard oil. The increased use of cano-
la oil in the liquid form in a large variety of edible fat products and concerns about
saturated fatty acids, along with the ready availability of palmitic acid-containing
oils for blending to control crystallization problems, seems to have prevented this
development from gaining commercial significance.
High stearic acid canola oil containing 2530% stearic acid was developed, but
commercial scale production for food uses has been very limited (140).
High lauric acid canola oil was developed in the United States as an alternative
source for coconut and palm-kernel oils for both food and nonfood uses (141). The
oil contains about 35% lauric acid. Until now, this oil has not found any significant
commercial use. The main reason for the lack of acceptance is said to be because of
its significantly different fatty acid composition compared with coconut oil, and the
consequent difference in performance in typical coconut oil applications. Some use
was made of the oil in the United States as a base stock for a trans-isomer free
margarine and in Europe as a machine oil additive (142), but there is no longer
any significant seed production.
Gamma linoleic acid canola oil is of interest (11). It is an example of a
development for the nutrition supplement market.
8.2.2. High erucic acid rapeseed (HEAR) oil In countries that grow canola,
HEAR oil is used only in special food applications. Its primary use is as a fully
hydrogenated oil to be added to peanut butter (143) in amounts of 12% to prevent
oiling mainly in Canada and the United States. The HEAR oil used contains about
4550% erucic acid, the highest erucic acid rapeseed oil available commercially at
present. Because the saturated erucic acid produce behenic acid, which has a very
high melting point, the completely hydrogenated HEAR oil is very effective in
holding high amounts of liquid oil in its crystal matrix. The patent literature also
mentions the use of fully hydrogenated HEAR oil in interesterification with palm
stearin fraction to formulate a no trans-isomer margarine hard fat, but there appears
to be no significant use.
The patent literature of the 1960s and 1970s contains a number of examples of
other uses of fully hydrogenated HEAR oil. These uses were designed to exploit the
beta prime crystallization properties, either as hardstock in small amounts for bak-
ing shortenings or as base stock for conversion to monoglycerides. A listing of these
patents is given by Teasdale and Mag (144). However, it appears that there have
been very few or no sustained commercial uses of these proposed fats.
NONFOOD USES OF STANDARD CANOLA OIL 109
Plant breeding work to raise the erucic acid content in the oil is being done in
Canada and elsewhere. Indications are that the erucic acid content of about 80% is
possible. This is of interest not only for some of the specialty food uses mentioned
above, but also especially for industrial lubricants (See nonfood uses of HEAR oil).
9. NONFOOD USES OF STANDARD CANOLA OIL
9.1. Biodiesel
Triacylglycerol oils, particularly vegetable oils, are suitable as base stocks of simple
ester derivatives for use as fuel for compression ignition engines, commonly known
as diesel engines. The term biodiesel is applied to this type of fuel. Methanol is the
main alcohol used for transesterification of the triacylglycerols to produce bio-
diesel. The pure methyl esters of fatty acids as well as blends of methyl esters
with petroleum-based diesel fuel are used. The environmental factors are the driv-
ing force to use renewable fuel source to produce biodiesel. Fatty acid esters
(including the triacylglycerols) are readily biodegradable in the presence of water
and soil bacteria. They have low toxicity to plants and animals. At this time, pro-
duction costs of biodiesel are higher than for petroleum-based diesel fuel. In Eur-
ope, biodiesel is taxed only at a very low rate to make it more competitive with
petroleum diesel fuel. There is the tendency to use in the pure form. In North Amer-
ica, no tax reduction is given; blends with petroleum diesel fuel are, therefore, more
popular. Blending, usually around 520% of biodiesel, achieves most of the benefits
that biodiesel has on engine performance without raising petroleum diesel fuel costs
significantly. The blends, of course, are not suitable for very sensitive environments.
The canola growing countries in Europe, notably, Germany, Austria, and France,
but also others, have developed significant methyl ester production capacity and
use.
Standard canola oil, mainly because of its fatty acid composition, is relatively
well suited for biodiesel production. Harrington (145), and Knothe et al. (146)
discussed desired properties of fatty acid ester structure for biodiesel. Knothe et al.
(146) also discussed biodiesel standards in different countries, for those interested.
Briefly, the desired properties of vegetable oil fatty acids for methyl ester biodiesel
can be summarized as follows:
1. Long, unbranched hydrocarbon chains

2. Moderate content of saturated fatty acid chains
3. Monounsaturation, but not polyunsaturation
4. Double bond position preferably toward the end of the fatty acid carbon chain
away from the methyl group.
High concentration of fatty acid methyl esters with the above properties has
good combustion characteristics and good flow properties, including good low-
temperature behavior.
110 CANOLA OIL
Standard canola oil is high in C18 fatty acids, about 95%, which is higher than
the other commodity vegetable oils. It is high in C18:1n - 9 oleic acid at about 60%,
much higher than any other vegetable oils, and it is relatively low in polyunsatu-
rated fatty acids, linoleic at about 21%, and linolenic at about 10%. Viscosity,
cold filter plugging point, and cetane number are some of the most important bio-
diesel fuel properties influenced by fatty acid composition.
The n-9 oleic acid in canola oil confers good combustion properties, high cetane
number, and good flow characteristics. The polyunsaturates also confer good flow
properties, but they have poorer combustion characteristics and lower cetane num-
ber, because the presence of double bonds in n - 3 and n - 6 position. Canola oil
saturated fatty acid content is low. Saturated fatty acids in the C16C20 range give
very high cetane numbers, but they impair the flow properties. It can be seen that
the mix of canola oil fatty acids is well suited for biodiesel production, better than
the other commodity vegetable oils, such as soybean, sunflower, or palm.
The desired properties of methyl esters for biodiesel are given in Table 24 (147).
These properties are the German biodiesel standard Deutsche Industrie Norm (DIN)
V 511605. This standard, as well as other European standards, was developed, espe-
cially with canola/rapeseed oil as the starting material in mind. The suggested
TABLE 24. Biodiesel Standard DIN V 51606, Germany.
Limits
Properties Units Test Method min. max.

Density at 15 C g/ml ISO 3675 0.875 0.900
Kinem. viscosity. at 4 C mm2/s ISO 3104 3.5 5.0

Flashpoint (Pensky-Martens) C ISO 2719 100

Cold filter plugging point C DIN EN 116
April 15Sep. 30 0
Oct. 1Nov. 15 10
Nov. 16Feb. 28 20
March 1Apr. 14 10
Sulphur content % by mass ISO 4620 0.01
Carbon residue (10 % distill.) % by mass ISO 10370 0.30
Cetane Number ISO 5165 49
Ash % by mass ISO 6245 0.01
Water mg/kg ASTM D 1744 300
Total dirt mg/kg DIN 51419 20
Copper corrosion (3 hr at 50 C) ISO 2160 1
Neutralization number mg KOH/g DIN 51558 Part 1 0.5
Methanol % by mass To be agreed 0.3
Monoglycerides % by mass To be agreed 0.8
Diglycerides 0.1
Triglycerides 0.1
Free glycerol 0.02
Total glycerol 0.25
Iodine number g Iodine/100g DIN 53241 Part 1 115
Phosphorus mg/kg To be agreed 10
Adapted from Varese et al. (147).

NONFOOD USES OF STANDARD CANOLA OIL 111
USA-ASTM standard for 100% pure biodiesel is similar in many respects (146), but
it is written for the use of soybean oil as the main starting material. Canola oil for
methyl ester production must either be degummed (<20 mg/kg of phosphorus), or
in addition, must be alkali refined and bleached, depending on the methyl ester pro-
duction process requirements (148).
Possibly, a designed canola oil with enhanced properties especially for biodiesel
production may be developed in the future. The high oleic acid canola oil already
available represents a version of such oil, because its high C18:1n - 9 content is a
very desirable methyl ester component of biodiesel. The same can be said of
the high oleic acid varieties of soybean and sunflower oil. It should be noted that
the direct use of canola oil as diesel fuel is also being fostered, especially in some
European countries, notably, Germany. The advantage is that the cost of conversion
to methyl esters and marketing for glycerol, byproduct, (see below), will lower the
total cost of the fuel. Farm tractors and certain stationary diesel engine and station-
ary heating equipment can be adapted for direct use of the oil, without transforma-
tion into esters, and the oil can easily be supplied to these users at suitable
temperatures to maintain good fluidity. The same properties that make canola oil
well suited as a base stock for methyl ester diesel fuel are also advantageous in
using it directly. The oil must be degummed to below 50 mg/kg of phosphorus,
depending somewhat on the engine manufacturers specifications, and it must be
free of any solid particles (148).
Table 25 gives a comparison of some of the basic properties of methyl esters
derived from palm, soybean, and canola rapeseed oil and standard petroleum-based
diesel fuel. In this table, palm oil illustrates the effect of fatty acid chain length and
relatively high saturation on properties, soybean oil illustrates the effect of high
unsaturation, and canola oil represents the effect of low saturation and high mono-
unsaturation. It is worth noting that HEAR oil would also be a good biodiesel base
stock resulting in methyl esters of higher cetane number, but higher viscosity and
somewhat poorer low-temperature behavior can be expected.
It can be seen that compared with petroleum diesel fuel, methyl esters (1) have
higher density, (2) fall into the lower range of viscosity, (3) have the same or higher
cetane numbers, and (4) have lower heating value. Canola methyl esters are in the
middle range of properties among the three oils, with low viscosity, good cetane
TABLE 25. Some Properties of Palm, Rapeseed Canola, and Soybean Oil Methyl Esters
Compared with Diesel Fuel.
Methyl Density Viscosity Cetane Iodine Lower Heating

Ester 15 C, g/L 40 C, mm2/s Number Value Value MJ/kg
Palm oil 877877 4.34.5 64.370.0 52 37.0

Canola oil 882 4.2 51.059.7 114 37.2
Soybean oil 880 4.0 45.756.0 131 37.1
Petroleum
Diesel fuel 830840 12.53.5 51 42.7
Adapted from Varese et al. (147).

112 CANOLA OIL
rating, and heating value. This indicates that the standard canola fatty acid compo-
sition produces a good compromise in the balance of desirable methyl ester
properties.
It is interesting to note that Varese and Varese (147) cite a highly favorable
energy balance factor range of 2.53.5 for the production of canola seed and the
conversion of the oil to canola methyl esters depending on climate, soil, and ester
production method. In this context, it is also worth noting that there are quality
problems with the byproduct glycerol from biodiesel production. Consequently,
the price of biodiesel byproduct glycerol is very low. If these quality problems
are solved, the economics of methyl ester production will become more comp-
etitive to petroleum-based diesel fuels. The vegetable oil methyl esters applied
for biodiesel can also be used as biosolvents for cleaners, ink removers, and similar
applications.
9.2. Solvents
Solvents based on vegetable oil methyl esters are finding increasing applications in
many areas. Biodegradability, very low volatility, low viscosity, tolerance to low
temperatures, and high flash points are the main attractions, along with good sol-
vent properties, such as water wetting, and penetration compared with petroleum-
based solvents. Especially the low volatility of methyl ester solvents is increasingly
important in the manufacture of paints and coatings. Lower oxidative stability is a
negative factor with esters produced from unsaturated oils, but canola oil methyl
esters perform better in this respect than the other more unsaturated vegetable
oils. Its low content of saturated fatty acids is an advantage in providing esters
of low viscosity that are easier to work with.
9.3. Lubricants/Engine Oils/Heat Transfer Oils/Hydraulic Fluids/

Demolding Agents/Inks
Lubricants are synthesized using a wide range of fatty acids and various alcohols,
including polyols. Some of these lubricants have viscosity and lubricity properties
that are not achievable with petroleum-based oils. A negative aspect is that lubri-
cants based on oils containing a high level of unsaturated fatty acids have poor oxi-
dative stability. Canola oil, because of its relatively low polyunsaturation compared
with commodity oils such as soybean has advantages in lubrication application. As
with methyl ester production, the canola oil used in these applications requires
refining to various degrees, depending on the base oil specifications of the manu-
facturer. The same arguments also apply to the use of canola oil as a base for engine
oil, heat transfer and hydraulic fluids, inks, and demolding agents.
There is considerable research and development work underway in many coun-
tries to test low-linolenic, high-oleic and high-palmitic canola oil varieties as base
stocks to determine which of these oils with modified fatty acid compositions is
best suited for the various products and applications.
PRODUCTION OF OILSEEDS AND OILS 113
9.4. Nonfood Uses of High Erucic Acid Rapeseed Oil (HEAR)

The traditional industrial use of HEAR oil was in lubrication of marine steam
engines. This use was caused by the unique film-forming properties of the oil,
which are particularly important in the wet steam engine environment.
Modern uses of HEAR oil are much more varied. One of the largest utilizations
is as base for the production of erucamide, which is a derivative of erucic acid, used
as an antiblock and slip-promoting agent in the production and functioning of plastic
films. The oil is also used as a component in paints and coatings. Erucic acid can be
converted by oxidative cleavage to brassylic (HOOC-(CH2)-COOH) and pelargonic
(nonanoic acid) acids. Former bypolymerization will form nylon 1313 (this number
indicates how many carbon atoms are in the acid component). This type of nylon
has low moisture absorption, greater dimensional stability, and good dielectric
properties, making it a good insulator. Brassylic acid is also an ingredient in the
production of polyester and melamine resin coatings. Paints with these additives
have balanced flexibility, low hydrolytic factor, and good hardness (149). Direct
polymerization of erucic acid showed promising results in plastic formulation but
will challenge plastic chemists in developing technologies and applications for it
(150). Further, HEAR oil serves as base for producing high-pressure greases.
10. PRODUCTION OF OILSEEDS AND OILS
Soybean dominates world oilseed production and represents 55% of the worlds
total oilseed production (Figure 3). The production of canola/rapeseed seeds ranked
second in the world behind soybeans in 20002001.
If one considers that soybean is produced for protein, it can be concluded that
canola production ranks first in terms of true oilseed production. The United States,
Brazil, and Argentina together produced about 75% of all soybean seeds. The world
production of soybean oil is in the top position followed by palm and canola/rape-
seed among major oils and fats (Figure 4) (75, 151).
The major producers of canola/rapeseed in the world are China, Canada, India,
and the countries of the European Union (Figure 5).
Canada, European Union countries, Poland, and Australia have been the major
canola seed exporters. Production of canola seeds in Canada is presented in
Figure 6, showing growing trend for the last decade.
Canola oil underwent a transition period in the 1980s as human nutritionists
recognized the benefits of the low levels of saturated fats in the diet and the benefits
of monounsaturates as compared with saturates and polyunsaturates. In the case of
Canadian canola oil, it moved from oil that competed in the markets around the
world on the basis of price to a premium priced oil. Canola is perceived as premium
oil and is now virtually consumed all over the North American continent. As was
discussed earlier, Canada is the worlds leading consumer of canola oil on a per
capita basis. The countries of the European Union and Canada dominate the export
of canola oil. In Canada, canola oil represents about 70% of all vegetable oils pro-
duced. The nearest rival is soybean oil with a share close to 25%. Functional and
114 CANOLA OIL
Figure 3. Production of oilseeds. Percentage of total world production of 306.9 million metric
tons. (Source: Canadian Grains Council, Statistical Handbook 2001.)
nutritional properties of canola oil, widely recognized by consumers, has allowed

canola oil to dominate the Canadian salad oil market at level of over 80% and the
shortening and margarine markets at levels of 66% and 50%, respectively. Canadian
exports of canola oil have become dominated by sales into the United States as
Figure 4. World production of oils and fats for the year 20002001. Total production of 117.1
million metric tons. (Source: Canadian Grains Council, Statistical Handbook 2001.)
PRODUCTION OF OILSEEDS AND OILS 115
Figure 5. Major world producers of canola/rapeseed seeds (ten-year average from 1990 to
2000). Production averaged 31.6 million metric tons per year. (Source: Canadian Grains Council,
Statistical Handbook 2001.)
American nutrition experts, food companies, and consumers recognized the

nutritional benefits of canola oil.
Oilseed products enter the world trade as either oil or protein meal or are used in
a wide variety of industries. Although the products from various oilseeds may
Figure 6. Canola seed production in Canada. (Source: Canada Grains Council, Statistical
Handbook 2001.)
116 CANOLA OIL
appear totally interchangeable, oils vary considerably in their chemical composi-

tion, nutritional properties, and functionality. The demand for oils is determined
by the basic characteristics that give them their desired functionality. This is
becoming even truer in the 1990s as genetically engineered plants and oils, targeted
at specific niche markets, begin to emerge and became available to food processors.
Modern technology enables manufacturers to substitute one oil for another, but
once the oil has been modified, absolute availability and price become important
determinants of demand.
The production or import of oilseed in a particular country will be determined by
whichever oilseed product is in the greatest demand. For example, in Japan, the
oilseed processing industry will consider the current domestic need for oils and
fats, international markets for oils and fats, and domestic and international demand
for protein meals before a decision is made as to which oilseed will be imported.
Basically, when domestic need for protein is high, soybean will be imported and
processed. If domestic demand for oil is the driving force, Japan imports canola
for processing (151).
REFERENCES
1. Saskatchewan Wheat Pool, The Story of Rapeseed in Western Canada, Regina,

Saskatchewan, Canada, 1974.
2. T. Mag, in F. Shahidi, ed., Canola and Rapeseed. Production, Chemistry, Nutrition and
Processing Technology, Avi Book, Van Nostrand Reinhold, New York, 1990, p. 251.
3. C. F. Ying and J. M. deMan, Can. Inst. Food Sci. Technol., 22, 222 (1989).
4. R. G. Ackman, in F. Shahidi, ed., Canola and Rapeseed. Production, Chemistry,
Nutrition and Processing Technology, Avi Book, Van Nostrand Reinhold, New York,
1990, p. 81.
5. R. Scarth, P. B. E. McVetty, S. R. Rimmer, and B. R. Stefansson, Can. J. Plant Sci., 68,
509 (1988).
6. A. J. Vecchio, Inform, 7, 230, 235, 239, 242 (1996).
7. R. Przybylski, N. A. M. Eskin, L. Malcolmson, S. Durance-Tod, R.Carr, and M. Mickle,
Lebensm. Wiss. u. Technol., 26, 205 (1993).
8. I. Petukhov, L. Malcolmson, R. Przybylski, and L. Armstrong, J. Am. Oil Chem. Soc., 76,
627 (1999).
9. K. Warner and T. L. Mounts, J. Am. Oil Chem. Soc., 70, 983 (1993).
10. R. Wong, J. D. Patel, I. Grant, J. Parker, D. Charne, M. Elhalwagy, and E. Sys, in
I. McGregor ed., Proceedings of the Rapeseed Congress, Program and Abstract,
Saskatoon, Saskatchewan, Canada, 1991, p. 53.
11. P. Tso, S. Ding, K. DeMichele, and Y. S. Huang, in Yung-Sheng Huang and Vincent A.
Ziboh, eds., g-Linolenic Acid: Recent Advances in Biotechnology and Clinical Applica-
tions, AOCS Press, Champaign, Illinois, 2001, p. 321.
12. J. L. Sebedio and R. G. Ackman, J. Amer. Oil Chem. Soc., 58, 972 (1981).
13. R. L. Wolff, J. Amer. Oil Chem. Soc., 70, 425 (1993).
REFERENCES 117
14. R. C. Wijesundera and R. G. Ackman, J. Amer. Oil Chem. Soc., 65, 959 (1988).
15. R. Przybylski, C. G. Biliaderis, and N. A. M. Eskin, J. Amer. Oil Chem. Soc., 70, 1009
(1993).
16. X. Hu, J. K. Daun, and R. Scaith, J. Amer. Oil Chem. Soc., 71, 221 (1994).
17. T. Rezanka and P. Mares, J. Chrom., 542, 145 (1991).
18. H. Kallio and G. Currie, Lipids, 28, 207 (1993).
19. R. G. Ackman, in J. K. G. Kramer, F. D. Sauer, and W. J. Pigden, eds., High and Low
Erucic Acid Rapeseed Oils, Academic Press, Toronto, Ontario, Canada, 1983, p. 85.
20. M. Jaky and E. Kurnik, Fetten-Seifen Anstrichmittel, 82, 267 (1981).
21. H. Liu, C. G. Biliaderis, R. Przybylski, and N. A. M. Eskin, J. Amer. Oil Chem. Soc., 70,
441 (1993).
22. F. Sosulski, P. Zadernowski, and K. Babuchowski, J. Amer. Oil Chem. Soc., 58, 561
(1981).
23. R. Przybylski and N. A. M. Eskin, J. Amer. Oil Chem. Soc., 68, 241 (1991).
24. J. Cmolik, J. Davidek, J. Pokorny, and J. Valisek, in K. Szymanski, ed., Proceedings of
the 7th Rapeseed Congress, Poznan, Poland, 1987, p. 1368.
25. T. K. Mag, H. G. Duff, and J. H. Ritchie, in 1987 AOCS Annual Meeting, New Orleans,
Louisiana, 1987.
26. A. Smiles, Y. Kakuda, and B. E. MacDonald, J. Amer. Oil Chem. Soc., 66, 348 (1989).
27. A. Smiles, Y. Kakuda, and B. E. MacDonald, J. Amer. Oil Chem. Soc., 65, 1151 (1988).
28. T. H. Smouse, in K. Warner and N. A. M. Eskin, eds., Methods to Assess Oil Quality and
Stability, AOCS, Champaign, Illinois, 1994, p. 12.
29. A. Kamal-Eldin and L. A. Appelqvist, Lipids, 31, 671 (1996).
30. G. W. Burton and K. U. Ingold, Ann. N. Y. Acad. Sci., 570, 7 (1989).
31. L. K. Patterson, in M. A. J. Rodgers and E. L. Powers, eds., Oxygen and Oxy-Radicals in
Chemistry and Biology, Academic Press, New York, 1981, p. 89.
32. V. W. Bowry and R. Stocker, J. Amer. Chem. Soc., 115, 6029 (1993).
33. A. L. Tapel, Ann. N. Y. Acad. Sci., 203, 12 (1968).
34. R. C. Zambiazi, Ph.D. disseetation, University of Manitoba, Winnipeg, Manitoba,
Canada, 1997.
35. L. Normand, N. A. M. Eskin, and R. Przybylski, J. Amer. Oil Chem. Soc., 78, 369 (2001).
36. Th. Willner, U. Je, and K. Weber, Fett/Lipid, 99, 138 (1997).
37. R. P. Evershed, V. L. Male, and L. J. Goad, J. Chrom., 400, 187 (1987).
38. M. H. Gordon, and L. A. D. Miller, J. Amer. Oil Chem. Soc.,74, 505 (1997).
39. A. Strocchi, Riv. Ital. Sostanze Grasse, 64, 401 (1987).
40. S. B. Kochar, Progress in Lipid Research, 22, 161 (1983).
41. G. Marchio, R. de Andreis, and E. Fideli, Rivista Italiana delle Sostanze Grasse, 64, 185
(1987).
42. R. Przybylski and N. A. M. Eskin, in I. McGregor, ed., Proceedings of the 8th Rapeseed
Congress, Saskatoon, Saskatchewan, Canada, 1991, p. 888.
43. J. Nourooz-Zadeh and L. A. Appelqvist, J. Amer. Oil Chem. Soc., 69, 288 (1992).
44. Y. Endo, C. T. Thorsteinson, and J. K. Daun, J. Amer. Oil Chem. Soc., 69, 564 (1992).
45. K. Suzuki and A. Nishioka, J. Amer. Oil Chem. Soc., 70, 837 (1993).
118 CANOLA OIL
46. B. Drozdowski, I. E.Goraj-Moszara, Z. Hazuka, M. Zajac, and R. Pawlowicz, in

K. Szymanski, ed., Proceedings of the 7th International Rapeseed Congress, Poznan,
Poland, 1987, p. 1304.
47. K. Ward, R. Scarth, J. K. Daun, and T. C. Thorsteinson, J. Amer. Oil Chem. Soc., 71, 1327
(1994).
48. L. L. Diosady, P. Sleggs, and T. Kaji, in Canola Council of Canada, 7th Progress Report
Research on Canola Seed, Oil, Meal and Meal Fractions, Winnipeg, Manitoba, Canada,
1986, p. 186.
49. C. M. Elson, D. L. Hynes, and D. L. MacNeil, J. Amer. Oil Chem. Soc., 56, 998 (1979).
50. R. C. Wijesundera, R. G. Ackman, V. Abraham, and J. M. deMan, J. Amer. Oil Chem.
Soc., 65, 1526 (1988).
51. D. Barnard, L. Bateman, E. R. Cole, and J. I. Cunneen, Chem. Ind. (London), July, p. 918
(1958).
52. D. M. Chapman, E. A. Pfannkoch, and R. J. Kupper, J. Amer. Oil Chem. Soc., 71, 401
(1993).
53. R. Przybylski, and N. A. M. Eskin, J. Amer. Oil Chem. Soc., 65, 629 (1988).
54. T. P. Labuza, CRC Crit. Rev. Food Technol., 2, 355 (1971).
55. R. Usuki, Y. Endo, and T. Kaneda, Agric. Biol. Chem., 48, 991 (1984).
56. R. G. Ackman, and C. A. Eaton, J. Amer. Oil Chem. Soc., 54, 435 (1977).
57. H. Noureddini, B. C. Teoh, and L. D. Clements, J. Amer. Oil Chem. Soc., 69, 1184
(1992).
58. W. Lang, S. Sokhansanj, F. W. Sosulski, J. Amer. Oil Chem. Soc., 69, 1054 (1992).
(1992).
60. V. M. Arens, G. Ghur, and J. Weibel, Fette Seifen Anstrichm, 79, 256 (1977).
61. N. B. Kyriakidis, and T. Katsiloulis, J. Amer. Oil Chem. Soc., 77, 1235 (2000).
62. L. DSouza, J. deMan, and J. M. deMan, J. Amer. Oil Chem. Soc., 68, 153 (1991).
63. S. McCurdy, Project Report 89-326, POS Pilot Plant Corp., Saskatoon, SK, Canada,
1990.
64. R. Kruger, Project Report 91-19C, Canola Utilization Assistance Program, Canola
Council of Canada, Winnipeg, Manitoba, Canada, 1992.
65. J. A. H. Dahlen and Ch. Kristofferson, in ISF/AOCS World Congress, New York, 1980.
66. E. H. Unger, in F. Shahidi, ed., Canola and Rapeseed: Production, Chemistry, Nutrition
and Processing Technology, van Nostrand Reinhold, New York, 1991, p. 235.
67. G. Penk, in A. R. Baldwin, ed., Proceedings of the World Conference on Emerging
Technologies in the Fats and Oils Industry. American Oil Chemists Society, Cham-
paign, Illinois, 1985, pp. 38 45.
68. Stork Friesland Scandinavia AB, Goeteborg, Sweden.
69. E. W. Lusas, L. R. Watkins, and S. S. Koseogulu, Project ASA RP-88401, Food Protein
Research & Development Center, Texas A&M University, College Station, Texas, 1989.
70. N. Buhr, in D. Erickson, ed., Edible Fats and Oils Processing: Basic Principles and
Modern Practices, American Oil Chemists Society, Champaign, Illinois, 1990, p. 43.
71. D. R. Grant, R. L. Eager, J. M. Pepper, and J. F. Mathews, The 8th Progress Report,
Canola Council of Canada, Winnipeg, Manitoba, Canada, 1987, p. 141.
REFERENCES 119
72. F. H. Schneider and U. Schuette, Fett, Wissenschaft u. Technologie, 86, 331 (1984).
73. L. D. Campbell and B. A. Slominski. J. Amer. Oil Chem. Soc., 67, 73 (1990).
74. Canadian Grain Commission, Inspection Division, Official Grain Grading Guide,
Winnipeg, Manitoba, Canada.
75. Canadian Oilseed Processors Association, Trading Rules, 1993-1994, Winnipeg,
Manitoba, Canada, pp. 6, 7, 23, 33.
76. American Oil Chemists Society, Official Methods and Recommended Practices,
77. H. J. Ringers and J. C. Segers, U.S. Patent 4,049,686, 1980, (to Unilever N.V.).
78. T. K. Mag, in F. Shahidi, ed., Canola and Rapeseed: Production, Chemistry, Nutrition
and Processing Technology, van Nostrand Reinhold, New York, 1991, p. 251.
79. K. F. Carlson, Inform, 4, 272 (1993).
80. Y. Endo, R. Usuki, and T. Kaneda, Agric. Biol. Chem., 48, 985 (1984).
81. T. K. Mag, in D. Erickson, ed., Edible Fats and Oils Processing: Basic Principles
and Modern Practices, American Oil Chemists Society, Champaign, Illinois, 1990,
p. 107.
82. U. I. Brimberg, Fette, Seifen, Anstrichmittel, 83, 184 (1981).
83. J. H. Henderson, J. Amer. Oil Chem. Soc., 70, 831 (1993).
85. P. M. Parker, in 84th AOCS Meeting, Anaheim, California, 1993.
86. L. O. Bergman, Canadian Patent 664,968, June 11, 1963, (to AB Pellerins Margar-
infabrik).
87. H. Szemraj, Tluszcze Jadalne, 18, 13 (1974).
88. P. Rothemund, in Kirk-Othmer, ed., Encyclopedia of Chemical Technology, Vol. 5, 2nd
ed., 1964, p. 348.
89. V. Abraham, and J. deMan, J. Amer. Oil Chem. Soc., 63, 1185 (1986).
90. J. deMan, J. E. Pogorzelska, and L. deMan, J. Amer. Oil Chem. Soc., 60, 558 (1983).
91. G. G. Hatfield, MSc thesis, University of Toronto, Faculty of Applied Chemistry and
Chemical Engineering, Toronto, Ontario, Canada, 1994.
92. N. Nock, in Annual Meeting of the American Oil Chemists Society, Atlanta, Maryland,
1994.
93. J. de Man and L. deMan, Inform, 3, 542 (1992).
94. A. Naguib-Mostafa, J. Amer. Oil Chem. Soc., 62, 756 (1985).
95. V. DSouza, J. deMan, and L. deMan, J. Amer. Oil Chem. Soc., 68, 153 (1991).
96. K. C. Thomas, J. Can. Inst. Food Sci. Technol., 2, 1167 (1988).
97. W. J. Desrosier, MSc. Thesis, University of Saskatchewan, Saskatoon, Saskatchewan,
Canada, 1988.
98. J. Kurashige, N. Matsusaki, and H. Takahashi, J. Amer. Oil Chem. Soc., 70, 849 (1993).
99. R. Przybylski, personal communication, 1994.
100. F. H. Mattson and S. M. Grundy, Lipids, 26, 194 (1985).
101. J. Westhuysen, Ann. Clin. Lab. Sci., 27, 1 (1997).
102. D. Steinberg, S. Parthasarathy, T. E. Carew, J. C. Khoo, and J. L. Witztum, New Engl.
J. Med., 320, 915 (1989).
120 CANOLA OIL
103. S. Parathasarathy and S. M. Rankin, Prog. Lipid Res., 31, 127 (1992).
104. P. Reaven et al., Am. J. Clin. Nutr., 54, 701 (1992).
105. M. Abbey, G. B. Belling, M. Noakes, F. Hirata, and P. J. Nestel, Am. J. Clin. Nutr., 57,
391 (1993).
106. M. Kratz, P. Cullen, F. Kannenberg, A. Kassner, M. Fobker, P. M. Abuja, G. Assmann,
and U. Wahrburg, Eur. J. Clin. Nutr., 56, 72 (2002).
107. A. M. Lees, H. Y. I. Mok, R. S. Lees, M. A. McCluskey, and S. M. Grundy,
Atherosclerosis, 28, 325 (1977).
108. T. Heinemann, G.-A. Kullak-Ublick, B. Pietruck, and K. von Bergmann, Eur. J. Clin.
Pharmacol., 40, S59 (1991).
109. R. P. Mensink and M. B. Katan, New Engl. J. Med., 321, 436 (1989).
110. B. E. McDonald, V. M. Bruce, E. L. LeBlanc, and D. J. King, in 4th Internal Rapeseed
Congress, 1974, pp. 693700.
111. A. Keys, J. T. Anderson, and F. Grande, Metabolism, 14, 776 (1965).
112. D. M. Hegsted, R. B. McGandy, M. L. Myers, and F. J. Stare, Am. J. Clin, Nutr., 17, 281
(1965).
113. B. E. McDonald, J. M. Gerrard, V. M. Bruce, and E. J. Corner, Am. J. Clin. Nutr., 50,
1382 (1989).
114. G. M. Wardlaw, J. T. Shook, M.-C. Lin, M. A. Puangco, and J. S. Kwan, Am. J. Clin.
Nutr., 54, 104 (1991).
115. J. K. Chan, V. M. Bruce, and B. E. McDonald, Am. J. Clin. Nutr., 53, 1230 (1991).
116. L. M. Valsta, M. Jauhianinen, A. Aro, M. B. Katan, and M. Mutanen, Arterioscler.
Thromb., 12, 50 (1992).
117. T. A. Miettinen and H. Vanhanen, Am. J. Clin. Nutr., 59, 356 (1994).
118. M. L. Bierenbaum, R. P. Reichstein, T. R. Watkins, W. P. Maginnis, and M. Geller, J. Am.
Coll. Nutr., 10, 228 (1991).
119. T. Seppanen-Lasskso, H. Vanhanen, I. Laakso, H. Kohtamaki, and J. Viikari, Br. J. Nutr.,
68, 639 (1992).
120. A. H. Lichtenstein et al., Arterioscler. Thromb., 13, 1533 (1993).
121. T. L. V. Ulbricht and D. A. T. Southgate, Lancet, 338, 985 (1991).
122. J. Dyerberg, Nutr. Rev., 44, 125 (1986).
123. P. M. Herold and J. E. Kinsella, Am. J. Clin. Nutr., 43, 556 (1986).
124. S. Renaud, F. Godsey, E. Dumont, C. Thevenon, E. Ortchanian, and J. L. Martin, Am.
J. Clin. Nutr., 43, 136 (1986).
125. B. M. Weaver, E. J. Corner, V. M. Bruce, B. E. McDonald, and B. J. Holub, Am. J. Clin.
Nutr., 51, 594 (1991).
126. J. K. Chan, B. E. McDonald, J. M. Gerrard, V. M. Bruce, B. J. Weaver, and B. J. Holub,
Lipids, 28, 811 (1993).
127. J.-S. Kwon, J. T. Snook, G. M. Wardlaw, and D. H. Hwang, Am. J. Clin. Nutr., 54, 351
(1991).
128. M. Mutanen, R. Freese, M. L. Valsta, I. Ahola, and A. Ahlstrom, Thromb. Haemost., 67,
352 (1992).
129. E. J. Corner, V. M. Bruce, and B. E. McDonald, Lipids, 25, 598, (1990).
130. T. A. B. Sanders and K. M. Younger, Br. J. Nutr., 45, 613 (1981).
REFERENCES 121
131. M. de Lorgeril et al., Lancet, 343, 1454 (1994).

132. M. de Lorgeril et al., Circulation, 99, 779 (1999).
133. B. F. Teasdale, Canadian Patent 726,140, 1966.
134. G. Pongracz, Fat Sci. Technol., 90, 249258 (1988).
135. National Academy of Sciences, Report of the Panel on Dietary Reference Intakes for
Vitamin C, Vitamin E, Selenium and Carotenoids, Food and Nutrition Board, National
Academy Press, Washington, D.C., 2000, p. 187.
136. Anonymous, Inform, 12, 487 (2001).
139. R. Scarth, P. McVetty, and S. Rimmer, in D. McGregor, ed., The Proceedings of the Eigth
Rapeseed Congress, Saskatoon, Saskatchewan, Canada, 1991, p. 143.
140. G. Galili, S. Galili, E. Lewinsohn, and Y. Tadmor, Critical Rev. Plant Sci., 21, 167
(2002).
141. T. Voelker, T. Hayes, A. Crammer, and H. Davies, Plant J., 9, 229 (1996).
143. C. H. Japiske, Canadian Patent 815,481, 1969.
144. B.F. Teasdale and T. K. Mag, in J. K. G. Kramer, F.D. Sauer, and W. J. Pigden, eds., High
and Low Erucic Acid Rapeseed Oils, Academic Press, Toronto, Ontario, Canada, 1983,
p. 197.
145. K. J. Harrington, Biomass, 9, 1 (1986).
146. G. Knothe, R. O. Dunn, and M. O. Bagby, Inform, 7, 827 (1996).
147. R. Varese and M. Varese, Inform, 7, 816 (1996).
148. G. Boerner, private communication, OHMI GmbH, Magdeburg, Germany.
149. S. Latta, Inform, 1, 438 (1990).
150. D. J. Murphy, Trends Biotech., 14, 206 (1996).
151. G. Bastin, Vegetable Oilseeds and Oils Traders Guide, The International Trade Centre,
UNCTAD/GATT, 1988.
3
Coconut Oil
Elias C. Canapi, Yvonne T. V. Agustin, Evangekube A. Moro,
Economico Pedrosa, Jr., Mara Luz J. Bendano
1. THE COCONUT PALM
In the tropics, the coconut palm (Cocos nucifera, L.) is one of the most useful trees.
The palm reaches 30 m or more in height, with a striped, unbranched, and relatively
smooth trunk measuring approximately 25 cm in diameter and crowned with some
30 fronds, each of which is about 56 m long. The roots may extend up to 10 m in
sandy soil, serving as anchor in addition to their primary function as nutrient absor-
bers and breathing organs (1).
As a perennial provider of food, beverage, shelter, animal feed, and feed-stock for
the oleochemical industries, the palm is reverently described as the Tree of Life, Tree
of Heaven, and other such metaphors by people of coconut-producing countries.
In some areas of the tropics, the palm grows with a minimum of attention
through a lifespan of more than 50 years. Its fruit, the coconut, has been referred
anticlimactically as the lazy mans crop; coconuts produced throughout the year
drop to the ground when mature, there to be collected at leisure. Commercial farms,
however, are tended and developed to maintain and improve productivity.
1.1. Geographical Distribution

Within 20 north and south latitudes, the coconut palm is productive, especially
along coastal areas (2). Palms grown beyond the limits of the Torrid Zone are
123
124 COCONUT OIL
generally nonproductive. The major coconut-growing areas are located in Asia,

islands of the Pacific Ocean, Africa, and Central and South America. In 1991 the
world coconut hectarage was 10.9 million (3).
1.2. Agronomy
Coconut farms are established in a wide variety of soil types, from littoral sands to
heavy clay. Palms grown in well-aerated soil are productive and tolerate soil pH
ranging from 5 to 8 (4). In determining the suitability of sites for coconut farms,
a good supply of soil moisture with adequate drainage are considered the foremost
factors. Water supply may come from well-distributed rainfall (13002300 mm/year),
irrigation, or from accessible water table (5). The extremes, excessive rainfall and
drought, affect the nut yield per palm.
The coconut palm thrives well in plenty of sunlight. It has been observed that
the palm produces more female flowers and a higher weight of copra per nut
during the months of May to August when there is abundant sunshine. Furthermore,
young palms grow slowly when planted under old trees in the course of rehabilitating
plantations (6).
Soil, water, and foliar analyses in coconut plantations are conducted to deter-
mine the type of fertilizer and micronutrients required by the palm. Long-term
studies show that the application of potassium or sodium chlorides in optimum
doses as supplements to the standard nutrients for inland farms increases yields
in nuts per tree and copra per nut (7, 8).
1.3. Propagation
The method of propagation for coconut palms is through seednuts. Due to the
highly heterogeneous character of coconut palms, progenies do not necessarily
exhibit all the desirable characteristics of the parent palms. Through selection of
seeds from elite parents, vigorous and productive palms may be evolved. The
long span of six or more years required to evaluate the performance of progenies
of coconut palms has underlined the need to explore other methods of propagation;
the response appears to be the current interest in the development of tissue culture
techniques.
Seedlings. Propagation of the coconut palm starts from the collection of fully
matured nuts (1112 months old) from selected palms. The seednuts are stored
in the shade for at least one month. In the nursery, seednuts are set to germinate
in loose and friable soil provided with adequate moisture and drainage facilities.
Some farmers transfer sprouted seednuts into polybags to allow selection of seed-
lings showing uniformity in growth and vigor (see Figure 1). Further, seedlings in
polybags do not suffer damage and shock when transported or when transplanted
in the fields.
Six-to nine-month-old seedlings with four to five leaves may be transplanted
in fields and spaced to accommodate 115160 palms per hectare. Coconut palms,
THE COCONUT PALM 125
Figure 1. A 9-month-old seedling raised in polybag and ready for transplanting in the
field.
on the average, start bearing fruits in the sixth year and continue to provide fruits
perennially for more than 50 years, barring any serious damage caused by plant
pests, diseases, climatic disasters, or neglect in farm management.
Tissue Culture. The most promising vegetative (asexual) method under develop-
ment for the propagation of elite coconut palms is the tissue culture technique. In
this method, the actively dividing tissue from a selected palm is removed and grown
asceptically in a specially formulated culture medium containing the prescribed
nutrients, hormones, and root and shoot growth-promoting factors. Under the right
induction conditions, the explant develops into a callus, which is induced to differ-
entiate into shoots and roots forming the plantlet. The plantlet is hardened in a pot
kept in a greenhouse before transplanting in the field.
The progress of studies on vegetative propagation have been reported with
encouraging results (911). Refinements of the current state of the art to improve
the number of survivors from the callus stage to seedlings planted in the fields is
needed to make tissue culture of coconut palms a feasible method to merit commer-
cial application.
126 COCONUT OIL
1.4. Varieties
There is a wide range of variations that may occur within a coconut palm colony
with regard to phenotype characteristics such as color, size and shape of nuts, and
shape and symmetry of crown, among others. In some isolated localities, colonies
of coconut palm develop well-defined and uniform phenotypic characters arising
from generations of natural selection.
Makapuno, a peculiar Philippine coconut variety, has a cavity within the kernel
filled with a palatable jellylike substance. These fruits do not germinate although
they contain an embryo. However, the propagation of makapuno has been success-
fully developed through in vitro culture of its rescued embryo and final establish-
ment of the seedling in the field (12, 13).
The two most distinguishable varieties of coconut palm are the tall and the
dwarf. The tall variety grows to a height of more than 20 m while the dwarf variety
attains a height of approximately 3 m upon maturity. Tall dwarf hybrids and the
reciprocal dwarf tall give progenies that have most of the superior properties of
either parent. With properly selected parents, the hybrid palms are generally hardy,
resistant to drought and diseases, precocious, and large fruit bearers.
2. THE FRUIT
The common mature coconut fruit weighs more that 1 kg and is ovoid in shape and
green or yellow in color. The nut has a smooth epidermis over a fibrous mesocarp
(husk) that covers the hard endocarp (shell). Within the shell is the endosperm (kernel,
meat) approximately 12 cm thick. A thin brown layer called testa separates the kernel
from the inner surface of the shell. The cavity within the kernel has an average volume
of 300 mL and contains the endosperm liquid (coconut water) (see Figure 2).
2.1. Kernel
The kernel is the origin of the following products: coconut oil, desiccated coconut,
coconut skim milk, coconut cream, coconut flour, protein powder, and copra meal
(see Table 1).
2.2. Coconut Water

Coconut water is a sterile liquid. From mature nuts, it has the following constituents
in percent: total solids 4.7, fat 0.74, protein 0.55, ash 0.46, and reducing sugar 1.
The liquid has a pH of 5.6 (14). It is a coproduct of desiccated coconut and con-
sidered a health beverage. Concentration of coconut water by reverse osmosis may
further expand its use in the fermentation and food industries.
2.3. Testa
The thin brown layer between the kernel and shell is the testa. This layer is pared
off from the kernels outer surface to eliminate colored bodies in the production of
THE FRUIT 127
Figure 2. Mature coconut fruit and a longitudinally halved specimen.
desiccated coconut. The parings contain oil that has a fair amount of unsaturated
fatty acids C18:1 and C18:2 (see Table 2).
2.4. Shell and Husk

The shell that encloses the kernel is a hard spherical covering 35 mm thick. It is
used mainly for fuel in copra making. Other products derived from coconut shell
are charcoal, activated carbon, filler for synthetic resin, glues, components in mos-
quito-repellent coils, and decorative items.
TABLE 1. Typical Composition of Coconut Kernel

(1, p. 171).
Composition %
Moisture 50.0
Oil 34.0
Ash 2.2
Fiber 3.0
Protein 3.5
Carbohydrates 7.3
128 COCONUT OIL
TABLE 2. Fatty Acid Composition of Paring

(Testa) Oil (1, p. 253).
Fatty Acid Paring Oil (%)
Caproic C6 0.1
Caprylic C8 0.5
Capric C10 1.5
Lauric C12 22.7
Myristic C14 15.8
Palmitic C16 17.2
Stearic C18 1.3
Oleic C18 : 1 25.4
Linoleic C18/2 15.5
The husk is 510-cm-thick fibrous cover adhering to the coconut shell. Like the
shell, it is also used mainly for fuel in farms. Products derived from the husk are
coir, bristle, rubberized fiber, rope, geotextiles, and activated carbon.
3. COPRA
Copra is the dried kernel of coconuts. Fresh kernels contain approximately 50%
moisture. Various drying methods are employed to bring down the moisture content
ideally to 68% (see Table 3). At this level, mold growth in copra is inhibited. The
conversion of kernel to copra is an essential step if the oil is to be drawn by the
conventional mechanical extraction method.
3.1. Sun Drying

Dehusked nuts are split into halves and drained. The halves are exposed to the sun
and in due time the kernels shrink. The partially dried kernels are separated from
the shells for further drying under the sun for 68 days. During occasional rains, the
kernels are protected with adequate cover, such as plastic sheets or any other sui-
table material.
TABLE 3. Calculated Yields Based on an Initial 100 g Fresh Kernel with 50% Moisture
Content (1, p. 162).
Moisture 50 15 12.5 10 7.5 5 0
Oil 34.5 58.65 60.37 62.10 63.82 65.55 69.0

Nonoils 15.5 26.35 27.13 27.90 28.68 29.45 31.0
Copra wt., g 100.0a 58.82 57.14 55.55 54.05 52.63 50.0
a
Fresh kernel.
OIL EXTRACTION 129
3.2. Direct Fire Drying

The direct fire dryer consists of a bamboo grill platform where the split nuts are
placed. Underneath is a fire hearth where coconut shells and husks are burned to
provide heat for the vaporization of water from the kernels. The kernels shrink
and are separated from the shells for further drying. Smoke from the burning
fuel imparts a light brown color on the copra and its oil.
Small farm producers continue to use this traditional method because of the low
cost of construction, simplicity of design, and ready availability of shell and husk
for fuel in coconut farms.
3.3. Hot-Air Drying

In 1978, Lozada (15) introduced a low-cost efficient dryer that is adaptable for
small farms. The dryer uses husks or shells for fuel in a specially designed burner
that can operate with an even supply of heat. The dryer is provided with a steel plate
bottom to isolate the kernels from the smoke generated by the fuel.
There are a number of hot-air dryer models, devices inspired by modern techno-
logy. In principle, these dryers apply hot air to expel moisture from the kernels.
Since smoke does not come in contact with the kernels, the product is a cleaner
and whiter copra.
The relationship of oil, non-oil, and moisture contents from fresh kernel to copra
as moisture diminishes in stages is given in Table 3.
4. OIL EXTRACTION
On an industrial scale, the dry process is the traditional method of extracting oil
from the coconut (16). This is done by crushing copra in an expeller, the trade
name of the machine patented by V. D. Anderson. The meal (or cake) may be
further treated with solvents to extract residual oil.
The wet-process feedstock is fresh kernel instead of copra. The extracted oil
does not have to be refined, unlike the oil from copra. The coproducts of oil
from the wet process are edible.
Sections 4.1 and 4.2 briefly describe the major steps in the extraction of oil from
copra and fresh kernel, respectively.
4.1. Dry Process

The dry process involves mechanical extraction of oil in crushers or expellers with
copra as feedstock (see Figure 3). Mechanical extraction may be supplemented with
a second extraction, using solvents, to recover residual oil from the meal.
Mechanical Extraction. Copra with 1012% moisture content is conveyed to an
automatic scale, passes through a magnetic chamber for the removal of tramp
iron, and is ground to a particle size of approximately 0.3 cm in diameter. The
particles are flaked to facilitate exposure of a large surface area.
130 COCONUT OIL
Figure 3. Mechanical extraction flow chart (16).
The flakes are cooked and conditioned for 20 min at 115 C in horizontal
cookers. Residual moisture is brought down to 3% as fat cells of the conditioned
copra flakes are ruptured and phosphatides are precipitated.
The sized, cooked, and conditioned flakes are fed to the crushers where a conti-
nuous squeezing action extracts the oil. Extraction characteristic follows a smooth
diminishing curve with most of the oil drained at the inlet section of the barrel cage
and tapers off toward the exit. The resulting copra meal has a typical residual oil
content of 7%.
The characteristic of oil and copra meal from a well-maintained and properly
operated expeller are as follows: (1) low-colored oil, 8 Red/50 Yellow in the Lovi-
bond scale; (2) light brown copra meal with 8% maximum residual oil; and (3)
copra meal with a uniform thickness of approximately 0.6 cm and 6% maximum
meal fines.
OIL EXTRACTION 131
A sustained optimum expeller efficiency requires cooking and conditioning of

copra flakes to be controlled between 91 C and 93 C under an appropriate resi-
dence time to reduce moisture content between 3% and 4%. Flakes conditioned
below 91 C with residual moisture above 4% have lower oil yield while flakes con-
ditioned above 93 C with residual moisture below 3% would yield dark-colored
oils, charred meal with the consequent disadvantage of poor oil extractability.
Screened and settled fines should not exceed 10% of fresh feedstock to prevent
the formation of abrasive particles during the cooking/conditioning stage.
Post-expeller Treatment. The oil from the expellers passes to a screening and
settling tank to initiate the separation of fines, which are recycled with fresh feed-
stock to the system. To each ton of supernatant oil, 10 kg of bleaching earth is
mixed. Passage through a polishing filter gives a clear oil ready for storage or
further processing.
Copra meal (cake) from the filter press is pelletized, bagged, and despatched to
animal feed millers.
Solvent Extraction. This operation supplements mechanical extraction and con-
sequently minimizes residual oil in the copra meal.
Copra undergoes an accelerated preliminary extraction with a controlled residual
oil content of 1418% in the expeller meal; expeller throughput rate is almost
doubled. Hexane (bp 68.7 C) is widely used as the solvent for extraction. In an
extraction unit that operates on a countercurrent system, the cake is met by oil-
rich miscella (hexane oil) and leaves the extractor as it is rinsed with pure hex-
ane. The solvent-extracted cake has a residual oil content of approximately 3.5%.
Hexane in miscella and cake is recovered for reuse in succeeding operations.
Escaping hexane vapors are trapped by cold mineral oil spray to preclude hazards
and maximize solvent recovery.
4.2. Wet Process

Feedstock in the wet process (17) is fresh kernel. In addition to oil, other edible
coproducts are recovered from the kernel, namely, coconut flour, protein, carbohy-
drates, and vitamins.
To encourage wider commercial application of this process, the following advan-
tages are emphasized: superior quality of the oil product and the recovery of nutri-
ent coproducts that would otherwise be lost in copra. The process is briefly
described to give a general view on the extraction of oil from fresh kernels.
Dehusked mature nuts are shelled to separate the kernels, followed by paring to
remove the testa. The testa is set aside for the extraction of paring oil, a byproduct.
The pared kernels are finely comminuted through a wedge and die plate mill and
through a roller mill. The comminuted mass is passed through a screw press, which
expels the coconut milk. The milk is filtered through a screen conveyor. The cream
that is separated from the milk by centrifugation is heated to reduce its moisture
content. By further centrifugation, oil is separated. Trace moisture in the oil is
reduced to 0.10.2% level by atmospheric heating. A typical yield of 6.8 tons nat-
ural coconut oil is extracted from 25 tons of fresh kernel.
132 COCONUT OIL
The skim milk is spray-dried to recover proteins and carbohydrates. Residue

from the screw press and parings are milled to recover oil and coconut flour.
5. REFINING
Refining of crude fats and oils involves a series of steps for the removal of impu-
rities from the glycerides to make the product suitable for human consumption and
improve product shelf life. The impurities are fatty acids, phosphatides, metal ions,
color bodies, oxidation products, solid paricles, and volatiles that include objection-
able odors. Crude coconut oil is refined by any of the following methods: (1) che-
mical refining (batch or continuous) and (2) physical refining. The comparative
performance of both methods is summarized in Figure 4.
5.1. Chemical Refining (16)

The free fatty acids (FFA) of crude coconut oil are neutralized with dilute sodium
hydroxide solution resulting in the formation of soap: RCOOH NaOH !
RCOONa H2O. The soap and other impurities in the water phase are collec-
tively called soapstock.
In the batch refining process, the separation of the water phase containing soap-
stock from the oil is by gravity. Some amount of neutral oil is lost by saponification
and by occlusion in soapstock.
Continuous refining process, on the other hand, offers the following major
advantages over the batch-type operation: (1) saponification of neutral oil is
minimized due to a short contact time of 3045 s between oil and sodium hydr-
oxide; (2) the time consumed for the separation of the aqueous soapstock and
wash water from the oil is reduced considerably by passage through centrifugal
separators.
After refining (neutralization), the oil is bleached. Color bodies in the oil are
adsorbed on the surface of the bleaching clay and activated carbon particles.
Experiments cited by Brimberg p (19) showed that bleaching process follows the
rate formula, ln c=c0 k t, where t is the time from the addition of bleaching
clay; c is the concentration of pigment at time t, c0 is the concentration at t0 , and k is
the rate constant.
Efficiency in the bleaching process may be improved for certain oils when a
small amount of water is present during the mixing steps (20).
Deodorization is the last step in chemical refining. Volatile odoriferous sub-
stances, including low-molecular-weight fatty acids, are removed by stripping
with steam under reduced pressure. The final product is generically called RBD
(refined, bleached, and deodorized) coconut oil.
A lower grade product called Cochin oil is coconut oil that is chemically refined
and bleached but not deodorized.
Batch Neutralization. Crude coconut oil from storage is fed into cylindrical
tanks equipped with heating coils and stirrers. The oil is heated to 80 C while being
133
Figure 4. Material balance: chemical and physical refining of crude coconut oil. Units were converted from the original lb. and Be to kg. and sp. gr. (18).
134 COCONUT OIL
stirred, and a measured amount of sodium hydroxide solution is introduced by

spraying on the surface of the oil. An excess of 510% sodium hydroxide over
the stoichiometric requirement is added to ensure appropriate neutralization of
free fatty acids. In this stage, hydrated gums migrate to the water phase. Heating
and stirring are stopped when soap breaks are formed. A break forms as soap
coagulates with some occluded neutral oil, excess sodium hydroxide, and other
impurities. The aqueous soapstock is allowed to settle and subsequently drawn
off for acidulation with sulfuric acid to recover a mixture of fatty acids, occluded
neutral oil, and other impurities. The mixture is called acid oil.
The neutral oil is washed with soft water at 1015% dosage to remove traces of
soap. The washed neutral oil is dried under vacuum and stored in buffer tanks.
Continuous Neutralization (Short Mix). A regulated volume of crude coconut
oil is heated through a plate-type heat exchanger to a temperature of 9095 C, con-
veyed to a centrifugal mixer where a required volume of 2N sodium hydroxide solu-
tion is added through a metering pump. The oil-soap-stock mixture is passed
through a primary centrifuge for the separation of soapstock from the neutralized
oil. Neutralized oil from the primary separator is heated to 90 C and mixed with
1015% hot water to wash off residual soap. From a second-stage separator, the
washed oil cascades over a series of plates in the vacuum dryer maintained under
a pressure of 3000 Pa. The dried neutral oil is stored in buffer tanks.
Bleaching. In a cylindrical reactor provided with heating coils and a variable
agitator, the neutral oil is brought to a temperature of 90 C under a pressure of
6001000 Pa and agitated. A predetermined amount of bleaching earth and acti-
vated carbon is added via a dosing device installed at the top of the mixer. After
30 min, the slurry is dried and later passed through a plate and filter press for the
removal of spent earth and carbon. The neutralized and bleached oil is cooled down
to 50 C before being transferred to storage tanks. Steam is injected through the
spent earth to recover entrained oil.
Deodorization. The neutralized/bleached oil is pumped into a deaerator oper-
ated under a pressure of 500 Pa to evacuate entrained air. From the deaerator, the
oil passes through a shell and tube economizer and is heated to a temperature of
240 C by means of a thermal oil heater. The stripper and deodorizing column oper-
ates under a pressure of 6001000 Pa; volatile components such as low-molecular-
weight fatty acids, ketones, aldehydes, and other odoriferous substances are
stripped off by live steam. The rising vapors laden with volatile components pass
through a cyclone scrubber where fresh fatty acid oil is sprayed on top of the vessel
to recover outgoing fatty acids.
The deodorized coconut oil flows down to the drop tank and holding chamber of
the distilling column. Residence time in the deodorizer ranges from 1.5 h to 2 h.
The oil from the column is withdrawn by a hermetically sealed and heat-resistant
pump and conveyed to the economizer, which preheats the incoming deodorizer
feedstock. The oil is cooled down to 50 C, at which temperature citric acid is
added. Citric acid enhances the stability of oil by immobilizing iron and copper,
which are pro-oxidants. (Note: The deodorization step is omitted in the production
of Cochin oil.)
COCONUT OIL COMPOSITION 135
5.2. Physical Refining

Coconut oil refiners have gained interest in the physical refining system as a sub-
stitute for chemical refining for the following reasons: (1) physical refining has low-
er oil losses vis-a-vis chemical refining; (2) pollution problems associated with
soapstock acidulation is precluded; (3) lower installation cost; (4) lower steam,
water, and power consumption; and (5) distilled fatty acids are of a higher grade
than the acid oil from chemical refining (21).
Design and technological breakthroughs have improved efficiency of the system.
Performance of new packed-column deodorizers for physical refining is predicted
by computer calculation, for which a special program has been developed on the
basis of experience with earlier columns (22).
The main feature in physical refining of crude oils is the application of steam dis-
tillation to remove the free fatty acids and volatile components from the oil. The tech-
nical feasibility of physical refining depends largely on the pretreatment stages for
the removal of phosphatides, color bodies, metal ions, and nonvolatile impurities.
Without an effective pretreatment, steam refining may fail to produce an oil of color
and stability characteristics comparable to the classically refined product (23).
Degumming. Crude coconut oil is continuously heated to 8090 C, and 85%
phosphoric acid at a range of 0.050.10% by volume of feedstock is dispersed in
the oil via an in-line static mixer. The mixture is conveyed into a coagulation tank
and agitated for 2030 min.
Bleaching. The oil is pumped to a slurry tank passing through a preheater at a
temperature of 9095 C. The oil, dried and deaerated under vacuum, receives a
predetermined dose of bleaching earthactivated carbon (10 : 1 ratio), agitated for
2030 min for the removal of color bodies and other adsorbable impurities. The
bleached oil is filtered through a plate and frame filter press and further purified
in drum-type polishing filters.
Steam Stripping and Deodorization. The degummed and bleached oil is deaer-
ated and heated to 240 C as it passes through economizers and thermal oil heating
units. The hot oil cascades down the stripping column operating under a reduced
pressure of 6001000 Pa pulled by four-stage steam ejectors. The oil is met by
sparge steam injected at the bottom of the column. Residence time is 11.5 h for
the removal of the odoriferous substances. Rising vapors are sprayed with fresh
fatty acids through a nozzle-type sprayer to recover stripped fatty acid distillates.
The deodorized coconut oil exits from the column and passes through an economi-
zer for partial cooling by the incoming bleached oil feedstock. Final cooling takes
place in a plate-type cooler where the temperature is maintained at 4050 C. Citric
acid is added before passing the product through a polishing filter.
6. COCONUT OIL COMPOSITION
Coconut oil belongs to unique group of vegetable oils called lauric oils. The most
abundant fatty acid in this group is lauric acid, CH3(CH2)10COOH. Other sources of
lauric oils are palm kernel, babassu, cohune, and cuphea.
136 COCONUT OIL
More than 90% of the fatty acids of coconut oil are saturated. This accounts for
its low iodine value ranging from 7 to 12. The saturated character of the oil imparts a
strong resistance to oxidative rancidity. Assessment of the oil by active oxygen
method (AOM) yielded results between 30 h and 250 h (24). Although oxidative sta-
bility is reduced in RBD oils, due to losses in the natural antioxidants of crude coco-
nut oils, the addition of citric acid at the end of deodorization as the oil is cooled to
100 C was effective in regaining considerable oxidative stability in the oil (25).
Fatty acids with less than 12 carbon atoms are classified as medium-chain fatty
acids (MCFA). Esters of MCFA with glycerol, known as medium-chain triglycerides
(MCT), are components in medical foods and infant food formulations. With more
than 15% C6, C8, and C10 fatty acids, coconut oil is the richest source of MCFA.
Approximately 0.5% of crude coconut oil is not saponified by caustic treatment.
The unsaponifiable matter consists mainly of tocopherols, sterols, squalene, color
pigments, and carbohydrates. The odor and taste of coconut oil is largely due to
d- and g-lactones, which are present in trace quantities (24). Among the unsaponifi-
ables, tocopherol contributes to the oxidative stability of crude coconut oil. A
typical sample of crude coconut oil contained 55 ppm total tocopherols of which
40.7 ppm is a-tocopherol (25). Most of the unsaponifiables are removed in the
process of refining, bleaching, and deodorizing of crude coconut oil.
The various triacylglycerol (TAG) components of coconut oil may be separated
and quantified by gas chromatography with the use of stable silicon gum stationary
phase under temperature-programed conditions and identified by reference to stan-
dard TAG solutions. The carbon number of a TAG component is the sum of carbon
atoms of the fatty acids attached to the glycerol moiety. For example, the carbon
numbers of trilaurin and oleodistearin are 36 and 54, respectively. The relative
amounts of each TAG in a sample of fat serves to establish its identity. For coconut
oil, this test may also serve to distinguish it from other lauric oils (see Table 4 and
Figure 5).
TABLE 4. Triacylglycerol Number Composition (wt %)

of Coconut Oil (26, p. 228).
TAG Carbon Number Range (Mean) %
C28 0.71.0 (0.8)

C30 2.84.1 (3.4)
C32 11.514.4 (12.9)
C34 15.617.6 (16.5)
C36 18.319.8 (18.8)
C38 15.117.7 (16.3)
C40 9.211.1 (10.2)
C42 6.58.0 (7.3)
C44 3.64.6 (4.2)
C46 2.13.0 (2.6)
C48 1.62.6 (2.3)
C50 0.82.0 (1.7)
C52 0.42.0 (1.6)
C54 0.11.5 (1.2)
CHEMICAL AND PHYSICAL TESTS 137
Figure 5. Distribution of triacylglycerols by carbon number, coconut oil (23).
7. CHEMICAL AND PHYSICAL TESTS
For purposes of identifying natural fats and ascertaining their quality, a number of
analytical tests are routinely employed. The test results of a sample of fat under
assessment should fall within the range of established constants to confirm its
identity. For coconut oil the usual tests are fatty acid composition, acid value/
percent free fatty acid, saponification value, iodine value, ReichertMeissl value,
Polenske value, unsaponifiable matter, peroxide value/stability test, s.p./m.p., color,
and solid fat content (see Tables 5 and 6)
The principal criteria presently being used to measure quality of coconut oil are
free fatty acids content and color. In addition, sensory evaluation through taste and
odor of oil serves to confirm the acceptability of the product.
Details on sampling methods and test procedures are described in various
reference sources (2832).
7.1. Fatty Acid Composition

The individual fatty acid composition of fats and oils are routinely assayed by
gas chromatography. Samples are transesterified with methanol to convert the
138 COCONUT OIL
TABLE 5. Range of Fatty Acid Composition of Coconut Oil

Cross Regional (26, p. 227).a
Fatty Acid Range Weight (Mean) %
Caproic C6 0.40.6 (0.5)

Caprylic C8 6.99.4 (7.8)
Capric C10 6.27.8 (6.7)
Lauric C12 45.950.3 (47.5)
Myristic C14 16.819.2 (18.1)
Palmitic C16 7.79.7 (8.8)
Stearic C18 2.33.2 (2.6)
Oleic C18 : 1 5.47.4 (6.2)
Linoleic C18 : 2 1.32.1 (1.6)
C20 t 0.2 (0.1)
C20 : 1 t 0.2 (t)
a
Five countries and a total of 21 samples.
fatty acids into relatively volatile methyl ester derivatives. The esters are volatilized
and swept in a stream of inert gas through a stable inert powder support that has
been treated with a liquid (stationary phase) that is not volatile under the test con-
ditions. The esters are eluted in succession from the column in accordance with
their individual retention times in the stationary phase. Each emerging fatty acid
ester is recorded ideally as an individual peak; the area under the curve is a function
of the quantity of eluted component.
7.2. Acid Value / Percent Free Fatty Acid

Acid value is the number of milligrams of potassium hydroxide required to neutra-
lize the free acids in 1 g of oil.
A measured amount of oil is dissolved in ethyl alcohol and titrated with 0.1 N
sodium hydroxide. For coconut oil, an equivalent term, percent free fatty acid
TABLE 6. Coconut Oil Product Specifications (27).
Crude Cochin RBD
Moisture and impurities % max. 1.0 0.1 0.03

Free fatty acid (as lauric) % max. 3.0 0.07 0.04
Color (5 1/4 in. cell) Lovibond R/Y max. 12/75 1/10 1/10
Saponification value 250264 250264
Unsaponifiable matter % max. 0.4 0.1 0.1
Peroxide value, max. 2.0 0.5 0.5
Slip/melting point, C 2426 2426
Refractive index 40 C 1.4481.450 1.4481.450
Flavor/odor Coconut flavor Bland/odorless
and odor
CHEMICAL AND PHYSICAL TESTS 139
as lauric (%FFA), is more commonly used. % FFA as lauric mL 0.1 N

NaOH 20/g oil; acid value (mg KOH/g oil) % FFA as lauric 2.8.
7.3. Saponification Value

Saponification value is the number of milligrams of potassium hydroxide required
to neutralize the free acid and saponify the esters in one gram of fat.
A slight excess of alcoholic potassium hydroxide is reacted with a measured
amount of oil sample and boiled gently under reflux with an air condenser. A blank
is similarly treated. Upon complete saponification, the treated sample and blank are
titrated with 0.5 N hydrochloric acid. Saponification value B S 0:5 N
HCI 56.1/g oil; where B titer of blank in mL and S titer of sample in mL.
7.4. Iodine Value

Iodine value is the number of grams of iodine absorbed in one gram of oil. The
unsaturated carbon atoms of fatty acids absorb iodine according to the following
reaction
CH CH I2 ! CHI CHI
A slight excess of either Hanus (IBr) or Wijs (ICI) solution is added to a mea-
sured amount of oil sample and shaken for 30 min. A blank is similarly treated. A
specified volume of 15% potassium iodide is added to the treated sample and blank,
followed by titration with 0.100 N sodium thiosulfate.
Iodine Value B S 0:100 N thiosulfate 12.7/g oil; where B titer of
blank in mL and S titer of sample in mL.
7.5. Reichert-Meissl Value

The Reichert-Meissl value is the number of milliliters of 0.10 N sodium hydroxide
required to neutralize volatile fatty acids (mainly C4 and C6) of 5.0 g of oil.
Polenske value is the number of milliliters of 0.10 N sodium hydroxide required
to neutralize insoluble fatty acids (mainly C8, C10, C12) of 5.0-g oil. Typical
ReichertMeissl and Polenske values of coconut oil are 8.4 and 11.5, respectively
(33). Both values are significantly high, as expected in coconut oil, which is a rich
source of MCFA.
A sample of fat, weighing 5
0.1 g is saponified in glycerol-sodium hydroxide
solution. The saponified sample is acidified with dilute sulfuric acid, followed by
distillation of the volatile fatty acids. A volume of 110-mL distillate is collected,
chilled to 15 C, and filtered. The filtrate is titrated with 0.1 N sodium hydroxide.
A blank determination is carried.
ReichertMeissl value B S 1:1; where B titer of blank in mL and
S titer of sample in mL.
The water-insoluble fatty acids on the filter paper are washed, dissolved in ethyl
alcohol, and titrated with 0.10 N sodium hydroxide. Polenske value B S; where
B titer of blank in mL and S titer of insoluble fatty acids in mL.
140 COCONUT OIL
7.6. Unsaponifiable Matter

Unsaponifiable matter in fats is the ether or hexane-soluble components extracted
after a fat sample is refluxed with alcoholic potassium hydroxide.
7.7. Peroxide Value/Stability Test

Peroxide value is expressed in milliequivalents peroxide oxygen per kilogram oil.
Peroxides formed in oil during storage initiate the development of oxidative rancidity.
A measured amount of fat is dissolved in acetic acidchloroform solvent in the
presence of potassium iodide. Peroxides in the fat liberate iodine, which is titrated
with 0.10 N sodium thiosulfate. A blank run is similarly treated. Peroxide value
(meq peroxide oxygen/kg oil B S 0:10 N thiosulfate 1000/g oil; where
B titer of blank in mL and S titer of sample in mL.
Stability test (Swift test/active oxygen method) gives an indication of the oils
resistance to oxidation during storage. The test gives the time in hours required for a
sample of oil to reach a peroxide value of 100 when subjected to aeration under
specified conditions of temperature and airflow rate. The period of time in hours
is determined by interpolation between two peroxide value determinations, which
must fall between 75 meq and 175 meq.
7.8. Slip/Melting Points or Melting Range

The slip/melting points or melting range of fat products is the temperature at which the
transition from solid to liquid state is observed. Unlike pure substances, which have
sharp melting points, fats are mixtures of TAGs exhibiting a range of melting tempera-
tures. Some methods require elaborate equipment while others, such as the capillary
tube method are simple and can provide precise results for routine determinations.
By capillary action, molten fat sample is drawn into the tube, which is open at
both ends. The fat is solidified by chilling and then melted in water at a regulated
rate of increase in temperature. The temperatures at which the fat slips and appears
clear in the liquid state are noted as its slip/melting points or melting range.
7.9. Refractive Index

The refractive index of a medium is the ratio of the speed of light in vacuo at a
given wavelength to the speed of light in the medium. Measurement is done by
means of a suitable refractometer at a specified temperature for a particular sample.
Refractive index is useful for identification of fats and the observation of progress in
reactions during hydrogenation.
USES 141
TABLE 7. Solid Fat Content of Coconut Oil

(% by Pulse-NMR) (28, p. 386).
Fat Content (%)
SP/MP C 24
Iodine value 8.5
% Solids
20 C 36
30 C 0
35 C 0
7.10. Color
The color of coconut is measured in a Lovibond Tintometer, using a 1-inch or
5 14 -inch cell for dark and light colored oils, respectively. Results are given in red
and yellow units describing the combination that matches the sample color. Alter-
natively, the optical density of the oil can be measured with the use of a spectro-
photometer in a suitable cell at a wavelength of maximum absorbance.
7.11. Solid Fat Content/Solid Fat Index

Nuclear magnetic resonance (NMR) is widely used for the determination of solid
fat content in fats and oils. The old dilatometry method, which gives solid fat
index values, is time consuming and inadequate as a quick means for control pur-
poses in the processing of fats. The NMR method is quicker, more precise, and clo-
ser to the absolute solid fat content than dilatometry (see Table 7).
7.12. Titer
Titer is the solidifying point of the mixed fatty acids derived from a sample of fat.
Fat is saponified and subsequently acidulated. The layer of mixed fatty acids is
separated, washed, dried, and allowed to cool. As the solids begin to separate,
the temperature rises due to the liberation of latent heat. The highest temperature
reached is taken as the titer.
8. USES
Coconut oil is used in a wide range of food and nonfood products. It is a raw mate-
rial for the production of medical foods and infant food formulations. In industry,
the fatty acids of coconut oil provide a versatile feedstock for an array of products
from diesel fuel substitute to hygienic products.
8.1. Edible Products

In coconut-producing countries, RBD coconut oil is used extensively as frying oil.
Physical blends and interesterified mixtures of coconut oil and hydrogenated
142 COCONUT OIL
palm oil are processed into margarines and shortenings. Coconut jam, a syrupy
emulsion derived from coconut cream in cane sugar, is consumed as dessert, bread
spread, and rice cake topping. Filled milk, in liquid or powder form, contains
coconut oil (in lieu of butter fat) and a polyunsaturated oil emulsified in skim
milk. As spray oil for crackers and cookies, coconut oil improves shelf life of these
products because of its resistance to oxidative rancidity. Coconut oil is widely used
as cream fat and as a component in biscuit cream and confectionery oil.
8.2. Medical and Infant Food Formulations

Medium-chain fatty acids (MCFA), mainly C8 and C10, obtained by hydrolysis of
coconut oil, is reesterified with glycerol to form a mixture of randomized medium-
chain triglyceride (MCT) through a method developed by Babayan (34). MCTs are
absorbed and oxidized rapidly with an energy rated at 34.7 kJ/g. A number of med-
ical and infant food formulations have MCTs as the principal source of fat supple-
mented with polyunsaturates (35). In a field study by Intengan et al. (36) a
structured (interesterified) 75 coconut oil25 corn oil preparation gave better
weight gain and nutritional recovery, vis-a-vis a polyunsaturated vegetable oil,
when given to malnourished children as supplemental fat source in their diet.
8.3. Nonfood Products

A commonly used nonfood product derived from coconut oil is soap. Laundry bar
soaps made by boiled or cold procedures have excellent lathering property even in
moderately hard water. A blend of tallowcoconut oil in ratios from 67 : 33 to
85 : 15 form an ideal fat charge for toilet soaps (37). Soap from such blends exhi-
bit desirable characteristics related to lather quickness, low mechanical erosion, and
absence of swelling or cracking of soap bars.
One of the major uses of acid oil from soapstock and the distillates from physical
refining is in the manufacture of animal feeds. The fatty components increase the
caloric density of the feed.
Derivatives of fatty acid from coconut oil are feedstock for a number of diverse
nonfood products. Coconut oil fatty acids and glycerol are released by hydrolysis or
alcoholysis of the fat. The fatty acids or their methyl esters, which are subsequently
fractionated, constitute the starting materials for the oleochemical industry. The
byproduct, glycerol, is purified by vacuum distillation. The purified product is,
among others, a component of pharmaceutical preparations, an important ingredient
in toothpastes, a raw material in the manufacture of nitroglycerol, and the fluid in
hydraulic jacks and shock absorbers.
9. STORAGE
Although coconut oil has a natural resistance to oxidative rancidity, (see discus-
sion), the bulk storage systems should take into account all factors that contribute
ECONOMICS 143
to the deterioration of the product during storage. These factors are light, prooxi-
datants, air, heat, and moisture. Light may have the least effect because the oil is
handled in a closed system.
From the milling of copra for the extraction of crude coconut oil through the
refining steps and final storage in tanks, the oil is in continuous contact with iron
and possibly copper-containing alloys, both of which are prooxidants. The addition
of citric acid (25) or any other appropriate antioxidant in the last stages of deodor-
ization of RBD oils (see discussion) affords protection to the oil from oxidative ran-
cidity. Crude coconut oil has natural and protective antioxidants.
The introduction of air during handling procedures may be minimized by filling
tanks through a subsurface entry point instead of allowing the oil to fall through air
into the storage tanks.
Storage tanks are provided with mechanical agitators and heating devices,
although the latter may hardly be used in most areas of the tropics. During brief
spells when temperature falls below slip/melting points of coconut oil, the product
is warmed with agitation to prevent localized overheating.
The effect of moisture on fat in storage is well known. In the presence of
enzymes, mainly from microorganisms, hydrolysis of fat is accelerated giving
rise to an unpleasant soapy taste peculiar to coconut oil exposed to conditions
favoring hydrolytic rancidity.
10. ECONOMICS
World coconut oil output fell 10% in 2001/02 to 3.3 million metric tons. U.S.
imports of coconut oil were 1,150 million pounds in 2001/02, up 35 million
from the previous year. In anticipation of a downturn in global production lasting
into 2003, importers were probably taking advantage of favorably low prices early
this year to secure larger stocks. The U.S. import unit value for coconut oil declined
from $361 per metric ton in 2000/01 to $327 in 2001/02. For palm-kernel oil, the
other main lauric oil, U.S. imports in 2001/02 fell to 330 million pounds from 364
million in 2000/01. But total supplies of palm-kernel oil were higher because of an
ample level of beginning stocks. A recovery in domestic disappearance from a low
2000/01 pace was possible because of the larger supplies (38).
Faced with a general deterioration of market prospects, the coconut industry
continued to receive special attention in major producing countries. In Indonesia,
support measures tended to emphasize intercropping, rehabilitation measures, and
product diversification. In the Philippines, in 2001, coconut producers have been
included in the public food distribution scheme with a view to protect farmers
from the impact of declining prices for coconut products. A number of accompany-
ing rural development programes aim at providing alternative livelihood opportu-
nities for small coconut farmers (39).
Table 8 gives world oilseed production data and includes copra. Table 9 gives
world vegetable oils production and includes coconut oil. Copra meal production
data are included in Table 10.
144 COCONUT OIL
TABLE 8. World Oilseed Production, 106 t 1995/96 to Date (38).
Years

Item 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03a
Production
Soybeans 124.90 132.22 158.07 159.82 159.90 175.10 183.78 184.49
Cottonseed 35.15 33.61 34.35 32.62 32.93 33.53 36.61 33.37
Peanuts 27.47 28.96 27.29 29.77 28.99 31.12 33.11 31.84
Sunflowerseed 25.72 23.80 23.21 26.63 27.22 23.29 21.25 23.33
Rapeseed 34.44 31.53 33.23 35.89 42.47 37.52 35.87 32.17
Copra 5.13 6.05 5.33 4.38 5.46 5.90 5.26 5.30
Palm kernel 4.87 5.21 5.05 5.62 6.41 6.91 7.24 7.40
Total 257.67 261.38 286.53 294.72 303.37 313.36 323.10 317.89
a
Forecast.
Source: Foreign Agricultural Service, USDA.
TABLE 9. World Vegetable Oils Production, 106 t 1995/96 to Date (38).
Years

1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03a
Production
Soybeans 20.17 20.53 22.57 24.65 24.74 26.80 28.72 29.85
Palm 16.26 17.64 16.97 19.25 21.80 23.93 24.88 25.37
Sunflowerseed 9.01 8.61 8.29 9.18 9.63 8.41 7.57 8.32
Rapeseed 11.24 10.52 11.43 11.81 13.64 12.96 12.20 11.41
Cottonseed 4.15 3.70 3.70 3.57 3.57 3.52 3.82 3.56
Peanut 4.15 4.38 4.18 4.44 4.15 4.30 4.75 4.51
Coconut 3.16 3.69 3.29 2.71 3.34 3.63 3.26 3.23
Olive 1.45 2.46 2.53 2.50 2.37 2.48 2.53 2.35
Palm Kernel 2.10 2.22 2.20 2.43 2.75 2.95 3.11 3.17
Total 73.08 73.76 75.16 80.54 85.97 88.98 90.85 91.79
a
Forecast.
TABLE 10. World Protein Meal Production, 106 t 1995/96 to Date (38).
Years
Item 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03a
Soybeans 89.08 90.82 98.84 107.54 107.74 116.47 124.71 129.58

Cottonseed 13.11 11.89 11.79 11.36 11.45 11.30 12.10 11.31
Rapeseed 18.58 17.53 18.85 19.12 22.27 21.18 19.99 18.64
Sunflowerseed 10.21 10.06 9.51 10.51 10.72 9.43 8.45 9.25
Fish 6.52 6.64 5.08 5.80 6.29 5.75 5.43 5.61
Peanut 5.73 6.01 5.41 5.76 5.27 5.52 6.13 5.79
Copra 1.74 1.97 1.74 1.44 1.77 1.90 1.68 1.70
Palm Kernel 2.54 2.70 2.67 2.93 3.32 3.56 3.75 3.82
Total 147.49 147.62 153.88 164.47 168.82 175.12 182.23 185.69
a
Forecast.
TABLE 11. Edible Coconut Oil: U.S. Supply and Disappearance, 106 lb 1991 to Date (38).
Item 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a 2003b
Stocks October 277 188 251 164 163 84 150 393 152 136 260 227 148
Imports 841 1,163 999 1,100 874 1,188 1,438 791 926 1,115 1,093 860 970
Exports 22 0 19 18 12 12 6 11 14 8 7 8 10
Domestic disappearance 910 1,084 1,067 1,083 941 1,111 1,189 1,021 927 983 1,119 930 958
a
Preliminary and estimated.
b
ERS and WAOB forecast.
Source: Bureau of the Census.
146 COCONUT OIL
Table 11 gives U.S. edible coconut supply and disappearance for 19912003. Dis-
appearance, as defined by the USDA-ER, means beginning food stocks, production
and imports minus exports, shipments to U.S. territories, and ending stocks.
Coconut oil is generally considered an expensive oil, and it normally commands
a premium over other vegetable oils.
REFERENCES
1. J. A. Banzon and J. R. Velasco, Coconut Production and Utilization, Philippine Coconut

Research & Development Foundation, Pasig, Metro Manila, Philippines, 1982, p. 19.
2. A. R. Diokno, The Brown Man and the Coconut, Manor Press, Manila, Philippines, 1962,
pp. 1121.
3. J. L. Arranza, in Proceedings of the World Conference and Exhibition on Lauric Oils,
Manila, Philippines, February 2025, 1994, p. 4.
4. S. S. Magat (Chairman), The Philippines Recommends for Coconut, PCARRD Phil.
Recommends Series No. 2-13, Los Banos, Laguna, Philippines, 1993, p. 3.
5. Ref. 1, p. 9.
6. Ref. 1, p. 6.
7. S. S. Magat et al., Phil. J. Coconut Studies, XIII(2) (1988).
8. G. D. Padrones et al., Response of MAWA Hybrid to Increasing Levels of NaCl, Annual
Report Agricultural Research, Philippine Coconut Authority, 1990, pp. 3840.
9. E. P. Rillo, Phil. J. Coconut Studies, XIV(2), 1620 (1989).
10. A. G. del Rosario et al., in Proceedings of the 5th Scientific Meeting of the Federation of
Crop Science Studies of the Philippines, Iloilo City, April 2629, 1989.
11. N. T. Thanh-Tuyen and D. Apurillo, Phil. J. Coconut Studies, XVII(1) (1992).
12. E. V. de Guzman and D. A. del Rosario, The Philippine Agriculturist, University of the
Philippines Publication Series A, Vol. XLVIII, No. 23, (1964), pp. 8294.
13. E. V. de Guzman and G. C. Manuel, Phil. J. Coconut Studies, II(1), 3539 (1977).
14. J. A. Banzon et al., Coconut as Food, PCRDF, Pasig, Metro Manila, 1990, p. 51.
15. E. P. Lozada, Coconuts Today, IX(1), 923 (1992).
16. Oil Milling Manual, Coconut Industry Investment Fund Oil Milling Group, Metro Manila,
Philippines, personal communication, 1993.
17. R. Hagenmaier, Coconuts Today, I(1), 1723 (1983).
18. D. C. Tandy and W. J. McPherson, JAOCS, 61(7), 1256(1984).
19. U. I. Brimberg, JAOCS, 59(2), 74.
20. Alfa-Laval, in the Fats and Oils Industry, Fats & Oils Division, S-147 80 Tumba, Sweden,
27829-E2, Sept. 1989.
21. A. Forster and A. J. Harper, JAOCS, 60(2), 265 (1983).
22. Alfa Laval Fats and Oils Packed Column for Edible Oil Refinery, Fats & Oils Division,
S-147 80 Tumba, Sweden, BP 5004E1, pp. 27.
23. Ref. 21, p. 266.
24. F. V. K. Young, JAOCS, 60(2), 376 (1983).
REFERENCES 147
25. M. H. Gordon and I. A. Rahman, JAOCS, 68(8), 574576 (1991).

26. J. B. Rossell et al., JAOCS, 62(2), 228 (1985).
27. Product Specification, Coconut Industry Fund Oil Milling Group, Metro Manila,
Philippines, personal communication, 1993.
28. K. Helrich, ed., Official Methods of Analysis of the Association of Official Analytical
Chemists, 15th ed., AOAC Inc., Arlington, Virginia, 1990.
29. Food Chemicals Codex, 3rd ed., Committee on Food Chemicals Codex, National
Academy Press, Washington, D.C., 1981.
30. Ref. 28, Second Supplement to 3rd ed., 1986.
31. C. Paguot, Standard Methods for the Analysis of Oils, Fats and Derivatives, 6th ed.,
Pergamon Press, Maxwell House, Fairview Park, Elmsford, New York.
32. W. W. Christie, Lipid Analysis, 2nd ed., Pergamon Press, Maxwell House, Fairview Park,
Elmsford, New York, 1982.
33. N. Gopalakrishnan et al., JAOCS, 64(4), 539 (1987).
34. V. K. Babayan, JAOCS, 45, 2325 (1968).
35. C. Dayrit et al., Phil. J. Coconut Studies, XVII(2), 1517 (1992).
36. C. Intengan et al., Phil. J. Coconut Studies, XVII(2), 1114 (1992).
37. N. O. V. Sonntag, JAOCS, 58(2), 156A (1981).
38. Oil Crop Situation and Outlook Year Book. Economic Research Service, USDA.
(Oct. 2002). Available: http://www.usda.gov.accessed july 2004.
39. Review of Basic Food Policies. Food and agriculture organization. (July 2004). Available:
http://www.fao.org.
4
Corn Oil*
Robert A. Moreau
United States Department of Agriculture,
Agricultural Research Service
1. OVERVIEW
In 2001, 2.04 MT (million tons) of corn oil was produced worldwide (representing
about 2% of the total worldwide vegetable oil production), with the top three pro-
ducers being the United States (57%), the EU-15 (10%), and Japan (5%) (1). Com-
pared with the 2001 world production of other vegetable oils, corn oil ranks tenth,
behind soybean (26.66 MT) > palm > canola/rapeseed > sunflower > peanut >
cottonseed > coconut > palmkernel > and olive (1).
Corn oil has traditionally been considered a premium vegetable oil, and at the
time of the writing of this chapter, the average U.S. wholesale price for corn oil
is $0.29/0.43 per pound (crude/refined) compared with soybean ($0.31/0.37) and
peanut ($0.57/0.71) oils (2, 3).
Unlike most other vegetable oils that are obtained directly from seeds that
contain high levels of oil, corn oil (maize oil) is obtained from seeds (kernels)
that contain only 35% oil. Obtaining oil directly from the kernels is technically

Mention of tradenames or commercial products in this article is solely for the purpose of providing
specific information and does not imply recommendation or endorsement by the U.S. Department of
Agriculture.
149
150 CORN OIL
TABLE 1. Fractions Obtained by Bench-Scale Dry Milling and Wet Milling

of Corn. Note: Commercial Wet Milled Fiber is a Mixture of Course and Fine
Fiber.
Wt % Oil in
Grams of Bench Scale Wt % Oil in Bench Commercial
Fraction Fraction/100 g of Kernela Scale Fractiona Fractionb,c
Dry Milling
Grits 82.63 0.63 0.03
Germ 11.24 15.00 0.37 2025b
Bran 6.13 2.11 0.03 12 c
Wet Milling
Starch 68.48 0.02 0.01
Gluten 8.34 0.89 0.00
Fine Fiber 8.58 1.36 0.05 12c
Course Fiber 5.48 2.19 0.05 12c
Germ 4.90 35.56 0.41 4450b
Steepwater 3.33 0.06 0.008
a
From (5).
b
From (6).
c
From (7).
possible, but corn kernel oil would be costly to produce (because of the low
levels of oil in the kernels). Because corn kernels contain high levels of starch
(6075%), a process, wet milling, was developed to efficiently isolate pure starch
from corn kernels. The first corn wet mill in the United States started producing
cornstarch in 1842, and by 1860, several corn wet mills were in operation (4).
Corn germ is generated during both wet milling and dry milling (Table 1), but
the amount of germ and the composition of germ differ when produced via these
two routes (Table 1). It was estimated that in 1996, about 90% of commercial corn
oil in the United States was from wet milled germ and the remaining from dry
milled germ (4).
During industrial wet milling, the kernel is separated into five fractions: (1)
starch, (2) steepwater solubles (7%), (3) fiber (course & fine, 10%), (4)
corn gluten meal (6%), and (5) germ (7%) (5). The steepwater solubles and
fiber fractions are blended together to produce an animal feed called corn
gluten feed, which contains about 21% protein and 6070% fiber. The high
fiber content restricts its use to mainly feeds for ruminants. Corn gluten meal
contains about 60% protein and low fiber (<1%), and it is a premium feed for
nonruminants (poultry and swine). Corn germ is rich in oil (>30%), and it is
the source of all commercial corn oil, and like wheat germ oil, corn oil could
more accurately be called corn germ oil. Like corn oil and wheat germ oil,
rice bran oil is another edible oil that is similarly extracted from a grain processing
fraction.
EXTRACTION AND REFINING 151
2. EXTRACTION AND REFINING
2.1. Conventional Corn Germ Extraction

Unlike oilseeds, where solvent extraction alone can be used to obtain oils, extrac-
tion after flaking of wet milled corn germ produces a substantial amount of
fines that interfere with the efficiency of the extraction process. Traditionally,
oil is removed from the wet milled germ using a conditioning (heating) process,
followed by mechanical expelling (prepress) and hexane extraction (Table 2).
Extrusion has also been employed as a means of germ preparation for solvent
extraction, producing a crude corn oil of high quality and high yield (8). Hexane
is removed from the oil-rich miscella by evaporation, heat, and vacuum, and the
hexane is recycled. Hexane is also removed and recycled from the germ cake.
Corn germ meal contains 2325% protein and is usually sold as an animal
feed ingredient and is often added to increase the protein content of corn gluten
feed.
Among the seven major U.S. wet milling companies that are members of the
Corn Refiners Association in 2003, only three companies report that they currently
market crude and refined corn oil (9). The other major wet mill companies sell their
corn germ, either for extraction and corn oil production, or for animal feed use,
depending on seasonal markets and prices.
Oil is usually obtained from dry milled corn germ by full press (via an expeller).
List et al. (10) compared some of the chemical components in corn oil from wet
milled germ versus dry milled germ and reported lower levels of free fatty acids,
lower levels of phosphorous, and higher levels of tocopherols in the latter. Although
several high oil corn hybrids are available and are becoming increasingly pop-
ular, most of the crop has been used for animal feed (with the increased fat provid-
ing more calories) and little or none has been wet milled to obtain starch, corn oil,
and other products.
TABLE 2. Processes for Conventional Extraction of Corn Germ.
Process Equipment Used Germ Pretreatment Oil Remaining (%)a
Untreated Germ 4450 (germ from wet mill)

2025 (germ from a dry mil)
Full (Hard) Press Expeller Heat 611
Partial Press & Expeller & Extractor Heat 13
Hexane
Extraction
Extrusion & Extruder (Expander) Heat and 12
Hexane to produce collets additional
Extraction and extractor moisture
a
From (6).
152 CORN OIL
2.2. Alternative Extraction Processes

A process for the hexane extraction of oil from the flaked kernels of high-oil (>8%)
corn hybrids was patented (11). An interesting aspect of the proprietary process is
that it uses the same common flaking and extraction machinery that is usually
employed for soybean oil extraction.
Ethanol has also been proposed as an organic solvent for the extraction of corn
oil from corn germ or from whole kernels. Some advantages of ethanol over hexane
are its higher flash point, it being a food-grade solvent, and the fact that it can be
readily produced from corn via fermentation. Some disadvantages compared with
hexane include its higher boiling point (requiring more energy to remove solvent
from meal) and its increased polarity, which means that it extracts more polar
extractants (e.g., phospholipids) that may need to be removed during refining.
Hojilla-Evangelista et al. (12) developed a process, the sequential extraction pro-
cess (SEP), which involves the ethanol extraction of whole flaked corn kernels to
first extract the corn oil, and then additional steps to extract and fractionate proteins
and starch. The economics of the SEP process have been rigorously evaluated, and
some recent process modifications have been proposed to improve the efficiency
and lower the cost (13, 14). Recently, a method was developed to remove corn
oil and zein from ground corn and a proprietary process was reported to separate
the corn oil and zein using membrane filters (15, 16).
Supercritical CO2 extraction methods were evaluated for a number of oilseeds in
the 1980s. Corn germ extraction by supercritical CO2 was evaluated by ARS
researchers in Peoria, IL. Methods for corn germ extraction using 100% supercri-
tical CO2 were developed (10, 17) and patented (18). Others have demonstrated that
addition of ethanol modifier (0 10%) to supercritical CO2 can decrease the extrac-
tion time and improve the functionality of germ proteins (19). Although there are
no technical barriers, extraction of corn germ with supercritical CO2 is more costly
than conventional extraction methods.
Several aqueous and enzyme-assisted methods have been reported for corn oil
extraction. In 1982, Stolp and Stute reported a proprietary aqueous process to
obtain corn oil from corn germ (20). An aqueous enzymatic (using commercial cel-
lulases, hemicellulases, polygalcaturonases, galactomanases, and various pectino-
lyitc enzymes) process for extracting corn oil from corn germ was also reported
(21, 22). Although these aqueous and enzymatic processes have been available
for a number of years, they have not been used in commercial production, mainly
because of higher costs when compared with conventional extraction techniques.
Recently, Verser and Eggeman received a U.S. patent for a process that uses enzy-
matic milling and produces ethanol via an acetic acid process that also yields corn
oil (23).
2.3. Refining Steps

The major component of crude corn germ oil is triacylglycerols, but the crude oil
also contains other minor nonpolar and polar lipid components (Table 3). Free
TABLE 3. Polar and Nonpolar Lipid Classes in Corn (germ) Oil, Corn Kernel Oil, and Corn Fiber Oil.
Wt% Total Lipid

Oil TAG FFA St:E St FPE tocols GL PL Reference
Germ Oil (crude) 95.6 1.7 nr 1.2* nr 0.06 nr 1.2 24

Germ Oil (crude) 96.8 0.31 0.47 0.48 0.01 0.17 nr nr 5
Germ Oil (RBD) 98.9 0.03 nr 1.1* nr 0.05 0 0 24
Kernel Oil (crude) nr Nr 0.763.09 0.541.28 0.0470.839 0.0230.127 nr nr 25
Fiber Oil (crude) 84.5 2.11 5.61 1.17 4.11 0.76 nr nr 5
Fiber Oil (crude) nr Nr 2.99.2 1.94.3 6.59.5 nr nr nr 26

Value is after saponification, meaning that it is the sum of St and St:E.
Abbreviations: nrnot reported; TAGtriacylglycerols; FFAfree fatty acids; St:phytosterol fatty acyl esters; Stfree phytosterols; FPEphytosterol ferulate esters; tocols
tocopherols and tocotrienols; GLglycolipids; PLphospholipids; RBDrefined, bleached, and deodorized oil.
154 CORN OIL
TABLE 4. Steps in the Commercial Processing of Corn Oil.
Steps Purpose Byproduct Produced
Alkali Refining
Degumming Remove gums Lecithin
Alkali Treatment Remove free fatty acids Soapstock a source of fatty
acids
Bleaching Remove pigments and oxidation Spent Clay
products
Dewaxing (winterizing) Improves stability, especially at Waxes and saturated
lower temperatures triacylglycerols
Deodorization Remove volatiles and off-flavors Deodorizer distillatea
source of phytosterols and
tocopherols
Physical Refining
Degumming Prevent precipitates at low Lecithin
temperature
Bleaching Remove pigments and oxidation Spent clay
products
Dewaxing (winterizing) Improves stability, especially at Waxes and saturated
lower temperatures triacylglycerols
Steam refining/ Removes free fatty acids and other Distillatea source of free
deodorization volatiles fatty acids, and a potential
source of phytosterols and
tocopherols
fatty acids, pigments, volatiles, phospholipids, and waxes are the major undesirable
components in crude corn oil, and these are removed by several refining steps (Table 4).
Whereas soybean oil processing usually is preceded by water degumming, during
corn oil processing, degumming is often not included if corn oil is going to be
processed via alkali refining (degumming is necessary if physical refining is
used). In corn oil processing, most companies remove free fatty acids by alkali
refining, which involves adding base and neutralizing (and sequestering) the free
fatty acid soaps (and phospholipids) into a byproduct called soapstocks (26).
A chemical comparison of the soapstocks obtained from alkali refining of oils
from corn germ, peanut, and cottonseed revealed that corn oil soapstocks had a
high level of phytosterols (almost 7%) and phosphorous, and an intermediate level
of free fatty acids (27). Alternatively, free fatty acids can be removed by the process
of physical refining or steam refining, which involves treating the oil at high
temperature and vacuum to volatilize the free fatty acids. Physical refining is only
advisable if the oil is of high qualityotherwise, the oil becomes dark and has poor
stability (personal communication, R. Ormsbee).
Physical refining begins by removing phospholipids by water degumming (28).
Failure to adequately remove the phospholipids (by either alkali refining or degumm-
ing) results in a corn oil that will form dark colors and off flavors when heated (4).
After a subsequent bleaching step, the next step in physical refining is a steam dis-
tillation at high temperature and very low pressure (vacuum), which volatizes the
COMPOSITION 155
free fatty acids. Leibovitz and Ruckenstein (29) reported higher yields of oil with
physical refining than with alkali refining. Others have noted that oils that contain
phytosteryl esters (especially ferulate-phytosteryl esters, such as those found in
corn fiber oil and rice bran oil) are extensively hydrolyzed during conventional
alkali refining, but they remain relatively intact during physical refining (Personal
communication, R. Nicolosi). Other strategies for removing free fatty acids from
crude oil include liquidliquid extraction and a new method involving solvent
extraction in a perforated rotating disk (30). Deodorization of corn oil involves
treatment at high temperature (>200 C) and vacuum (210 mm Hg), and it
removes undesirable odors and flavor components (24). Unfortunately, the deodor-
ization process also removes some phytosterols and tocopherols.
The byproduct of deodorization, the deodorizer distillate, has been used as a
major industrial source of phytosterols (which are used as ingredients in phytosterol-
containing functional foods and supplements or used as precursors in the synthesis
of some steroid pharmaceuticals) and tocopherols (a major source of natural
vitamin E). A recent study compared the levels of tocopherols and phytosterols
in industrial deodorizer distillates obtained from chemical and physical refining
of corn, canola, sunflower, and soybean oils (31). In another study, the effect of
the type of vegetable oil refining process (physical versus chemical) on the levels
of phytosterols was compared (32), and interesting differences were observed. Pig-
ments are usually removed by treating the oil with acid-activated bleaching clay (6).
Another refining step that ensures stability of oils at low temperature is dewaxing or
winterization, which involves cooling the oil to 510 C, and removing precipi-
tates via filtration (29).
3. COMPOSITION
3.1. Comparison of Corn Germ Oil, Corn Kernel Oil, and Corn Fiber Oil
Although all of the current commercial corn oil is produced from corn germ oil,
considerable research has been devoted to the study of extracting the entire corn
kernel to produce corn kernel oil and extracting corn fiber (a byproduct of wet
milling) to obtain corn fiber oil (Table 3). The levels of total phytosterols (the
sum of free phytosterols and phytosterol fatty acyl esters) in corn germ oil (refined)
averages about 1%, which is higher than the levels found in most other common
vegetable oils (24). Some of these phytosterols are removed during refining, but
even after refining, the levels of total phytosterols in commercial corn oil are about
1% (Table 3). Hexane-extracted corn kernel oil contains higher levels of the three
phytosterol lipid classes (free phytosterols, phytosteryl fatty acyl esters, and phyto-
steryl ferulate esters) than those found in corn (germ) oil (30; Table 3). Moreau et al.
(7) reported that a unique oil, very rich in the two phytosteryl esters (their chemical
properties will be described in a later section) could be extracted from corn fiber. Corn
fiber oil contains the highest levels of natural phytosterols and phytostanols of any
known plant extract (33). The relevant patent for corn fiber oil covers the process to
156 CORN OIL
extract the oil, the composition of matter, and the cholesterol-lowering applications
(34). Recently, researchers at Eastman patented an alternative process for obtaining
corn fiber oil and other products from corn fiber (35, 36).
3.2. Fatty Acid Composition of Corn Triacylglycerols

Edible oils are often compared by performing alkaline hydrolysis (saponification)
of the triacylglycerols and by comparing the fatty acid profiles. In the 1950s and
1960s, a marketing slogan for corn oil was that it was high in polyunsaturates,
mostly attributed to its high levels of linoleic acid, an essential fatty acid (one that is
not synthesized by humans and must be obtained in the diet), abbreviated as 18:2
(Table 5). Another desirable characteristic of corn oil is that it contains relatively
low levels (<15%) of saturated fatty acids and very low levels of linolenic acid,
abbreviated as 18:3 (which is especially susceptible to oxidation, leading to rancid-
ity). Although the levels of linoleic acid in U.S. corn oil average about 60%, it has
been noted that its levels in corn oil produced outside of the United States are closer
to 50%, with most of the difference being accounted for by higher amounts of oleic
acid (26).
Several studies have reported that when the same corn hybrids are grown in mul-
tiple locations, the corn oil produced from plants grown in cooler regions contains
higher levels of linoleic acid (41). It was also noted that the average levels of
linoleic acid in commercial corn oil in the United States increased from 57.8%
to 62.0% between 1974 and 1986 (41).
In response to the current demand for high monounsaturate-vegetable oils,
efforts have been devoted to developing corn hybrids that produce corn germ oils
with high levels of oleic acid. A high oilhigh oleic acid corn hybrid has been
patented (42). In addition to the extensive literature on the fatty acid composition of
crude and refined corn germ oil, the fatty acid composition of crude corn kernel oil
(39) and crude corn fiber oil (43) have also recently been reported. The fatty acid
composition of kernel oil and fiber oil are very similar to that of corn germ oil
(Table 5).
3.3. Triacylglycerol Molecular Species

Reversed-phase high performance liquid chromatography (HPLC) techniques have
been developed to quantitatively analyze the triacylglycerols molecular species of
fats and oils from a variety of plants and animals. Investigations of the
triacylglycerols molecular species of refined corn oil indicated the successful
identification of 19 to 27 individual molecular species, with oleatelinoleatelinoleate
and linoleatelinoleatelinoleate being the two most abundant molecular species
(Table 6). Silver ion HPLC was also used to quantitatively analyze corn oil triacyl-
glycerols (46). The method separated the triacylglycerols into 11 fractions, with the
largest 2 fractions having five and six double bonds, which translates to the struc-
tures oleatelinoleatelinoleate and linoleatelinoleatelinoleate (confirming the
two most abundant molecular species identified in reversed-phase HPLC).
TABLE 5. The Fatty Acid Composition of Corn (germ) Oil and Corn Fiber Oil.
mol% of Total Fatty Acids

Oil 16:0 18:0 20:0 18:1 18:2 18:3 Ref
Germ oil (RBD) US 11.0 0.5 1.8 0.3 0.2 0.2 25.3 0.6 60.1 1.0 1.1 0.3 24
Germ oil (RBD) US 9.216.5 03.3 0.30.7 2042.2 39.465.6 0.51.5 37
Germ oil (RBD) US 10.90 1.80 nr 24.2 58.0 0.70 38
Germ oil (RBD) US 11.0 0.6 1.7 0.3 nr 25.8 0.9 59.8 1.2 1.1 0.4 6
Germ oil (RBD) Int 12.9 1.4 2.6 0.6 nr 33.1 2.5 48.8 2.4 1.4 0.4 6
Kernel oil (crude) Int 9.211.8 1.11.7 0.30.5 19.530.4 53.065.3 1.22.1 39
Corn fiber oil (crude) 13.8 0 1.7 0 0.3 0 23.8 0.1 56.4 0.1 2.6 0 40
Abbreviations: nrnot reported; USUS hybrids; IntInternational hybrids; 16:0palmitic acid; 18:0stearic acid; 20:0arachidic acid; 18:1oleic acid; 18:2linoleic acid;
18:3linolenic acid; RBDrefined, bleached, and deodorized oil.
158 CORN OIL
TABLE 6. Quantitative Analysis of Triacylglycerol Molecular Species in Refined Corn

Germ Oil.
TAG Molecular Species Area %a Area%b Area %c
LLO 19.98 21.5 23.0

LLL 17.79 25.4 22.6
LLP 13.71 14.7 15.2
OOL 11.82 10.7 10.6
PLO 10.85 10.0 10.4
PPL 2.48 2.5 1.7
OOP 3.48 2.9 2.4
LLS 2.64 2.2 1.8
LOS 1.77 1.8 1.3
OOO 4.35 2.8 3.2
PPO 1.55 0.9 0.4
PLS 0.78 0.8 0.4
LLLn 0.91 1.2 0.8
LnLO 2.20 0.9 2.3
OOS 0.56 0.6 0.5
POS 0.20 0.3 0.3
PLnL 0.43 0.5 0.5
PPP 0.0 0.0 0.1
OOLn 1.09 0.1 1.0
PLnO 0.0 0.1 0.5
PPS 0.36 0.0 0.1
SSL 0.0 0.1 0.3
LnLS 0.0 0.1 0.0
SSO 0.0 0.0 0.0
PPLn 0.0 0.0 0.2
SSP 0.0 0.0 0.1
SSS 0.0 0.0 0.1
a
From (44).
b
HPLC-Mass Spec values from (45).
c
HPLD-Flame Ionization Detector Values from (45).
Abbreviations; Lnlinolenic acid; Llinoleic acid; Ooleic acid; Sstearic acid; Ppalmitic acid
3.4. Unsaponifiables and Phytosterols

Commercial corn oil has been recognized as having the highest levels of unsaponi-
fiables (1.32.3%) of all commercial vegetable oils (6). The three most abundant
chemical components in the unsaponifiable fraction of corn oil are phytosterols,
tocopherols, and squalene.
Corn germ oil contains two phytosterol lipid classes, free phytosterols and phy-
tosteryl fatty acyl esters (Table 3). Phytosterols have been recognized as 1 of the 12
important classes of phytonutrients (47). Most chemical identification of phytoster-
ols in vegetable oils has been conducted by saponifying (alkaline hydrolysis) the oil
and measuring the resulting free phytosterols, usually by GLC (Table 7). The major
phytosterols in corn germ oil are sitosterol (formeraly called b-sitoster-
ol) > campesterol > stigmasterol (Table 7; Figure 1). Snyder et al (50) developed
TABLE 7. Phytosterols in Corn (germ) Oil and Corn Fiber Oil (Values Represent Free and Esterified Phytosterols, Measured After Saponification).
mol % of Total Phytosterols

Oil campesterol stigmasterol b-sitosterol sitostanol 5-avenasterol 7-stigmasterol 7-avenasterol ref
Germ oil (RBD) 18.624.1 4.37.7 54.866.6 nr 4.28.2 1.04.2 0.72.7 37

Germ oil (RBD) 24.3 7.7 61.6 nr 3.8 0.7 0.8 48
Germ oil (RBD) 22.1 5.7 69.8 nr 2.4 nr nr 49
(free phytosterols)
Germ oil (RBD) 16.7 6.8 66.9 nr 8.0 nr nr 49
(esterified phytosterols)
Corn fiber oil (crude) 4.9 0.4 1.4 0.1 34.3 0.1 43.1 0.7 1.8 0 nr nr 40
Corn aleurone oil (crude) 4.7 0.3 tr 20.8 0.1 50.6 0.4 3.5 0.1 nr nr 40
160 CORN OIL
HO HO
Cholesterol Campesterol
HO HO
Stigmasterol Sitosterol (-Sitosterol)
HO HO
Sitostanol (stigmastanol) 5-avenasterol
HO HO
7 Stigmasterol 7-avenasterol
Figure 1. Structures of the common sterols (phytosterols) in corn and their comparison with
cholesterol, the main sterol in animals. Note that cholesterol has 27 carbons, campesterol has 28
carbons, and all of the other phytosterols shown have 29 carbons.
a supercritical CO2 method to concentrate and fractionate the phytosterols in corn

germ oil. Analyses of the total phytosterols in corn fiber oil revealed that the major
phytosterol was sitostanol (Table 7). Sitostanol is a phytostanol (meaning that it
is completely saturatedit contains no carboncarbon double bonds, whereas
COMPOSITION 161
phytosterols typically contain at least one). Natural phytostanols are rare in plants,
and the only reports of its presence in greater than trace amounts have been in
grains and Tall Oil (a byproduct of paper pulping) (51). The phytostanols (sitostanol
and campestanol) used in commercial sitostanol-ester margarines are produced by
catalytic hydrogenation. Sitostanol is the phytostanol produced by the catalytic
hydrogenation of the two most common plant phytosterols, sitosterol and stigmas-
terol. Campestanol is the phytostanol produced by the hydrogenation of campester-
ol (33). Our Laboratory recently reported (40) that most of the sitostanol in corn
fiber oil is found as the phytostanyl ester of ferulic acid, and most of the sitostanol
in corn fiber (and in corn kernels) is found in the aleurone layer (Table 7). In some
grains, the aleurone layer comprises several cell layers, but in corn, it comprises a
single layer of (phytosterol-rich) living cells (52). A method to enrich corn-fiber-
derived aleurone cells by milling them to a small particle size, followed by floata-
tion to separate the aleurone and pericarp was recently reported (53). The composi-
tion of corn fiber oil was also found to be influenced by the type of hybrid and the
location of growth (26) and by various alternative milling technologies (54). As
noted in Section 3.2, corn oil soapstocks contain almost 7% phytosterols (27)
and may represent an economical feedstock for phytosterols.
Squalene is an unsaponifiable compound in corn oil that has not received much
attention. Squalene was previously reported to be the major hydrocarbon in corn oil
(43) and we recently reported that both corn germ oil and corn fiber oil contain
about 0.2% squalene (40).

Corn oil has long been known to be a rich source of tocopherols, with g-tocopherol
being the most abundant tocopherol, followed by a-tocopherol and then d-tocopher-
ol (Table 8). Among the tocopherols, a-tocopherol has received the most attention
because of its high vitamin E activity, but the other isomers also are known to have
valuable antioxidant properties. Recent evidence suggests that g-tocopherol may be
superior to a-tocopherol in preventing the oxidation of low-density lipoproteins and
in delaying thrombus formation (57). Wang et al. (55) recently reported significant
levels of tocotrienols (the most abundant was g-tocotrienol followed by a-tocotri-
enol) in corn kernel oil, and saponfication of the kernels caused about a two-fold
increase in the levels of extractable tocotrienols and g-tocopherol. It is currently
believed that, in addition to their valuable antioxidant properties, tocotrienols
also possess cholesterol-lowering propertiesprobably associated with their ability
to inhibit cholesterol biosynthesis (58). We recently reported high levels of g-toco-
pherol in corn fiber oil (about 0.36 wt %), noting that heat pretreatment (150175 C
for 1 hour) of the corn fiber (prior to extraction) caused a nearly ten-fold increase
in the levels of extractable g-tocopherol (increasing its concentration in the oil
to about 3%) (56). The HPLC method that we used to quantitatively analyze g-toco-
pherol in this previous report (56) was based on UV-280 nm detection. Recently, we
repeated our previously published measurement of g-tocopherol levels in heat
TABLE 8. Tocopherols and Tocotrienols in Corn (germ) Oil, Corn Kernel Oil, and Corn Fiber Oil.
mg/kg oil
Oil a-tocopherol b-tocopherol g-tocopherol d-tocopherol a-tocotreinol g-tocotrienol d-tocotrienol Ref.
Germ, crude 191 0 493 118 nr 23 nr 6

Germ, RBD 134 18 412 39 nr nr nr 6
Germ, RBD 23573 0356 2682468 2375 0239 0450 020 37
Kernel, crude 67276 020 5831048 1271 4690 60133 nr 39
Kernel, crude (w/o saponified) 23 nr 240 nr 53 97 nr 55
Kernel, crude (with saponified) 48 nr 532 nr 124 197 nr 55
Fiber oil, crude (untreated) nr nr 3,600a nr nr nr nr 56
Fiber oil (heat-pretreated) nr nr 34,933a nr nr nr nr 56
a
Subsequent studies in our laboratory (not yet published) have revealed that these exceptionally high g-tocopherol values (obtained via UV-280 nm detection) were not accurate.
When we repeated the HPLC analyses using fluorescence detection, the levels of g-tocopherol in corn fiber oil (extracted before and after heat-preatment) were similar to above
reported levels in corn germ oil and corn kernel oil (< 500 mg/kg oil).
TABLE 9. Cis and Trans Fatty Acids in Refined Corn Oil and Corn Oil Margarines.
grams/100 g product
C T CC Ct TT CCC
Oil % Fat 16:0 16:1 18:0 18:1 18:1 18:2 18:2 18:2 18:3 20:0 20:1 total trans ref
Corn oil 100 10.9 0 1.8 24.2 0 58.0 0 0 0.7 0 0 0 38

Corn oil margarine 81.98 8.99 0 5.16 16.97 19.63 25.69 0.5 0 0.47 0.19 0.16 20.13 63
(stick)
Corn oil spread, 38.83 3.91 0.02 2.44 9.44 5.20 14.9 0.46 0 0.26 0.19 0.07 5.66 63
light (tub)
Corn oil spread, 19.47 2.03 0.01 0.83 4.75 2.69 7.82 0.1 0 0.17 0.04 0.04 2.79 63
extra light (tub)
Abbreviations: Ccis; Ttrans. For other abbreviations, see Table 5.

164 CORN OIL
pretreated corn fiber using fluorescence detection (excitation at 294 nm and emis-
sion at 326 nm) and we discovered that the large UV-280 peak that we had pre-
viously identified as g-tocopherol was an artifact. Our most recent results (not
yet published), obtained with fluorescence detection, indicate that the levels of
g-tocopherol in corn fiber oil (before and after fiber heat pretreatment) are similar
to the levels in corn germ oil and corn kernel oil (Table 8).
3.6. Carotenoids
High levels of carotenoids have been reported in corn kernels. Most of the carote-
noids (74 86%) are localized in the endosperm, 2 4% in the germ, and 1% in the
bran (59). The most plentiful carotenoids in corn kernels are lutein and zeaxanthin.
Consuming foods that are rich in these carotenoids may decrease the risk for age-
related macular degeneration (60). The levels of carotenoids in commercial corn oil
are relatively low, due in part to their low concentrations in the germ and in part to
their removal during the bleaching step of processing. The nutritional value of corn
oil carotenoids has received little attention. Although it is generally believed that
carotenoids function as antioxidants, there is evidence that, under certain condi-
tions, carotenoids in vegetable oils and certain other food matrices may serve as
pro-oxidants, especially at higher concentrations (61).
3.7. Trans-Fatty Acids

The unsaturated fatty acids in all common vegetable oils exist only in the cis-con-
figuration. During the production of margarines, spreads and shortenings via cata-
lytic hydrogenation, carboncarbon double bonds are converted to carboncarbon
single bonds; however, the process also catalyzes the conversion of some cis-fatty
acids to trans-fatty acids (62). There is a large variation in the levels of cis- and
trans-fatty acids in commercial corn oil margarines and spreads (Table 9), with
the levels of trans-fatty acids ranging from a low of 2.79 to a high of 20.13 g
of total trans-fatty acids per 100 g of product. Concerns about possible linkages
between trans-fatty acids and cardiovascular disease and certain types of cancer
(64) have caused some groups to seek to reduce or eliminate the levels of trans-fatty
acids in foods, either by using butter, or by using processes other than hydrogena-
tion to raise the melting point of corn oil and thereby produce new types of margar-
ines and spreads. With the controversy associated with trans-fatty acids (produced
during chemical hydrogenation), some individuals may not realize that a natural
group of trans-fatty acids, conjugated linoleic acid or CLA (the trans-double
bond is produced by anaerobic rumen bacteria via biohydrogenation), have
been discovered in dairy and beef products. Our current understanding of interna-
tional research indicates that CLA may have several health-promoting properties
(65). The discovery of these health-promoting properties of certain trans-fatty acids
(CLA) may provide incentive to undertake an objective reevaluation of the risks of
the trans-fatty acids in margarines from corn and other vegetable oils.
PROPERTIES OF CORN OIL 165
4. PROPERTIES OF CORN OIL
4.1. Chemical and Physical Properties

The basic properties of corn oil include its pleasing flavor, its high level of poly-
unsaturated (essential) fatty acids, a low level of saturated fatty acids, and a low
level of linolenic acid (4). The other main physical and chemical properties of
corn oil are summarized in Table 10.
4.2. Stability
As frying is a major use of corn oil, numerous studies have compared the stabi-
lity of corn oil and other vegetable oils during frying (6, 66). One study demon-
strated that when compared with canola and soybean oils, corn oil produced the lowest
TABLE 10. Some Physical and Chemical Characteristics of Corn Oil.
Property Value
a
Iodine value 127133
Saponification number 187193a
Free fatty acids after RBD 0.05% maxb
Color lovibond 3.0 red maxb
Gardner 6 max
Refractive index
20 C 1.4753a
26 C 1.4726a
Specific Gravity
25/25 C 0.91875a
Viscosity
40 C 30.80 cPa
60 C 18.15 cPa
Dielectric constant 26 C 3.954a
Surface tension, 25 C 34.80 dyn/cma
Interfacial tension,k H2O at 24 C 18.60 dyn/cma
Thermal conductivity at 130 C 4.2017 10 5 J/s/cm2/ Ca
Unsaponifiables 13%
Weight per gallon at 60 C 7.7 poundsd
Melting Point 11 to 8 Cd
Smoke Point 230 to 238 Cd
Flash Point 332 to 338 Cd
Fire Point 366 to 371 Cd
Cloud Point 14 to 11 Cd
CAS number 8001-30-7
EINECSe number 232-278-2
Japan registry Corn oil: 002275
a
From (6).
b
From (9).
c
From (37).
d
From (6).
e
Abbreviation for European Inventory of Existing Chemical Substances (EU).
166 CORN OIL
level of oxidation products and retained the highest level of tocopherols, during
5 days of continuous frying (44). Another oxidative stability study concluded that
corn oil hybrids with higher levels of saturated fatty acids were more stable than
traditional corn oils (67). A new optical method was recently developed, providing
a new parameter for assessing the oxidative stability of corn oil during frying
(68). A method was also developed to measure oxidative stability (PV) in corn
and soybean oils using near infrared spectroscopy (NIR) (69). Sundram et al. (70)
developed an oxidation-resistant proprietary blend of palm fat and corn oil.
4.3. Nutritional Properties

Numerous clinical studies (>30) during the last half-century have supported the
hypothesis that corn oil has cholesterol-lowering properties (26). This acknowledg-
ment of corn oils apparent superiority over other vegetable oils in its cholesterol-
lowering properties has been termed the maize oil aberration (71). In the 1950s
and 1960s, experts believed that the high levels of polyunsaturated fatty acids in
corn oil were the reason for its cholesterol-lowering properties (4). Others sug-
gested that because corn oil contains the highest levels of unsaponifiables and phy-
tosterols of any common vegetable oil, these components cause the cholesterol-
lowering effect (72). A recent study comparing corn oil with cottonseed oil found
that, although corn oil contained more unsaponifiables, cottonseed oil was more
effective at lowering total serum cholesterol, which the authors attributed to the
specific types of unsaponifiables in cottonseed oil (73). Lichtenstein et al. (74)
reported that hydrogenation of corn oil reduces its cholesterol-lowering properties
in humans. Recent clinical studies (with phytosteryl and phytostanyl ester products)
have indicated that a person must ingest 1.6 to 3.3 g of phytosterols per day to
achieve a 5% to 10% reduction in total serum cholesterol (71). To ingest 1 g of
phytosterols from corn oil, a person would need to consume about 100 g (about
1/2 cup and about 900 calories) of corn oil per day, an amount that is feasible,
but probably not practical. A recent hamster feeding study (75), comparing corn
oil and olive oil, presented evidence that, although both oils reduced LDL-
cholesterol (bad cholesterol), olive oil was better at increasing HDL-cholesterol
(good cholesterol). Other recent studies indicated that corn fiber oil is more
effective than corn (germ) oil at lowering serum cholesterol in hamsters (76).
Recently, Ostlund et al. (77) reported that the levels of phytosterols in commercial
corn oil (about 1%) are enough to sufficiently reduce cholesterol absorption in
humans. Wagner et al. (78) compared a high monounsaturated fatty acid-rich
(MUFA) diet (olive/sunflower oil mixture) with a polyunsaturated fatty acid-rich
(PUFA) corn oil diet and found that the latter had more influence on lipo-
protein metabolism. They concluded that the hypocholesterolemic effect of the
PUFA-rich diet must also be connected with the high amount of unsaponifiable mat-
ter, mainly phytosterols in corn oil (78). Obviously, more work is required to eval-
uate the health-promoting properties of corn (germ) oil and corn fiber oil.
Some recent clinical studies with corn oil have focused on its high levels of
g-tocopherol and other unique properties. A Swedish study revealed that consumption
MAJOR FOOD USES OF CORN OIL 167
of corn and sesame oils significantly increases the levels of serum g-tocopherol
(79). Recently, a U.S. patent was issued for the use of g-tocopherol and g-tocopherol
derivatives as antioxidants and nitrogen oxide scavengers to treat and prevent high
blood pressure, thromboemobic disease, cancer, natriuretic disease, the formation
of neuropathological lesions, and immunomodulation (80). Schurgers et al. (81)
reported that the consumption of corn oil increases the absorption and metabolism
of vitamin K.
The antioxidant properties of tocopherols (such as those found in corn oil) have
been suggested to be involved in treating atherosclerosis by preventing the oxida-
tion of low-density lipoproteins (57). Another study indicated that the particular
ratio of tocopherols in corn oil (a high ratio of g-tocopherol/a-tocopherol) may
achieve better protection against DNA damage than a-tocopherol alone (82). The
beneficial effects of corn oil on blood pressure, platelet aggregation, and diabetes
have been reported by others (22).
5. MAJOR FOOD USES OF CORN OIL
5.1. Cooking/Salad Oil

Of the 1.3 billion pounds of refined corn oil produced in the United States in 2000,
approximately half was used for cooking and salad oils, and about a fourth was used
for margarines and spreads (4). Corn oil has long been a popular cooking oil,
because of its mild flavor, its stability (because of its low levels of linolenate),
and its reputation as a healthy edible oil (because of its high levels of polyunsatu-
rated fatty acids). Because of its higher levels of polyunsaturates than most other
commodity vegetable oils (especially soy), corn oil is considered a premium oil and
is sold at a premium. In recent years, with the increased popularity of monounsa-
turate-rich oils (olive, canola, and now NuSun sunflower oil), corn oil is still con-
sidered a premium vegetable oil, but there has been a drop in the price differential
between corn oil and other commodity vegetable oils.
5.2. Margarines and Spreads

Corn oil margarine (common stick margarines contain 80% fat) and corn oil
spreads (common tub margarine spreads contain 20% to 65% fat) are popular
products. In the 1870s, Unilever began manufacturing margarine in Europe, and
the U.S. Dairy Company began production of artificial butter in the United States
(83). Sales of corn oil margarine in the United States climbed gradually and reached
a level of about 1 million pounds in 1930 (4). During the period of 1950 to 1980
(when there was growing consumer interest in the health-promoting properties of
the polyunsaturates in corn oil), the production of corn oil margarine in the United
States climbed from 15 to 250 million pounds per year (4). During the last 20 years,
there has been growing concern about possible harmful effects of the trans-fatty
acids in margarines and spreads (which can range from 1020% of the total fatty
acids) made from corn oil and other vegetable oils, and consumer concerns about
168 CORN OIL
trans-fatty acids have become a major issue for margarine manufacturers. Methods
have been developed to produce margarines, shortenings, and spreads by interester-
ifying (84) or blending oils (including a recent patent that details a process to pro-
duce trans-free shortening by blending corn oil and palm fat) (70), thus
eliminating the need for chemical hydrogenation and eliminating the formation
of trans-fatty acids. Zero trans-fatty acid margarines and spreads currently account
for a major portion of the sales in several European countries, and a growing num-
ber of U.S. manufacturers are now marketing zero trans-fatty acid margarines and
spreads in the United States. In July 2003, the U.S. Food and Drug Administration
amended its regulations on nutrition labeling to require that trans-fatty acids be
declared in the nutrition label of conventional foods and dietary supplements on
a separate line immediately under the line for the declaration of saturated fatty acids
(85). This ruling will take effect on January 1, 2006.
6. NONFOOD USES OF CORN OIL
6.1. Cosmetics/Skin Care Products

Corn oil is allotted important regenerative properties because of its high level of
unsaponifiables. Like all of the oils rich in essential fatty acids, it has a restructuring
activity and reinforces the cutaneous barrier. It thus helps to maintain a hydrated
epidermis. Corn oil is also a very good skin conditioner and an ingredient in sheath-
ing products for dry hair, body massage oils, emollient hand creams, face care pro-
ducts, skin care products, after-sun oils and creams, and nourishing lip balms (86).
6.2. Biodiesel
In most parts of the world, the term biodiesel now denotes a diesel fuel that is
produced by converting a vegetable oil to methyl (or ethyl) esters. In the United
States soybean oil has been the primary feedstock for biodiesel, mainly becase it
is commonly the least expensive and most abundant vegetable oil. Although there
are economic reasons why corn oil (and other U.S. vegetable oils) has not been used
as feedstocks for biodiesel, there are no technical reasons why a corn oil biodiesel
could not be successfully developed (personal communication. M. Haas).
6.3. Other Industrial Uses

Some of the other industrial uses for corn oil include insecticide formulations,
paints, varnishes, rubber substitutes, rust preventatives, soaps, leather tanning,
and textiles (87).
7. CONCLUSIONS
Corn oils desirable properties include its mild nutty flavor, high levels of unsatu-
rated fatty acids, low levels of saturated fatty acids, very low levels of linolenic
REFERENCES 169
acid, high levels of unsaponifiables (including phytosterols and tocopherols), and

stability during frying. Although the focus of consumers has been shifting toward
high- monounsaturate-oils (olive, canola, and NuSun sunflower oils), additional
research is still needed to objectively compare the health-promoting properties of
polyunsaturates versus monounsaturates. The food industry currently relies heavily
on corn oil margarine and other food products that contain hydrogenated corn oils.
Additional research also is needed to objectively evaluate the risks associated with
trans-fatty acid-containing products made from partially hydrogenated corn oil,
especially in light of the recent evidence that at least some types of natural
trans-fatty acids (conjugated linoleic acids) may have multiple health-promoting
properties.
REFERENCES
1. F. D. Gunstone, Vegetable Oils in Food Technology, Blackwell Publishing, Oxford, 2002,

pp. 6, 12.
2. Chemical Market Reporter, June 7, 2004.
3. M. Ash and E. Dohlman. Oil crop outlook. (2003, April 11). Economic Research Service,
USDA, Available: http://www.ers.usda.gov.
4. Anonymous, Corn oil, manufacturing, properties, uses and analysis of corn oil, Corn
Refiners Association, St. Louis, MO, 1996, p. 24.
5. R. A. Moreau, V. Singh, S. Eckhoff, M. J. Powell, K. B. Hicks, and R. A. Norton, Cereal
Chem., 76, 449 451 (1999a).
6. L. R. Strecker, M. A. Bieber, A. Maza, T. Grossberger, and W. J. Doskoczynski, in Y. H.
Hui, ed., Baileys Industrial Oil and Fat Products, Fifth Edition, Volume 2, Edible Oil and
Fat Products: Oils and Oilseeds, Wiley, New York, 1996, pp. 125158.
7. R. A. Moreau, M. J. Powell, and K. B. Hicks, J. Agric. Food Chem., 44, 21492154
(1996).
8. A. Maza, U. S. Patent 6,201,142, March 13, 2001.
9. Anonymous. (2003a). 2003 Corn Annual, Corn Refiners Association. Available : http://
www.corn.org.
10. G. R. List, J. P. Friedrich, and D. D. Christianson, J. Am. Oil Chem. Soc., 61, 18491851
(1984).
11. J. F. Ulrich and S. C. Anderson, U. S. Patent 6,313,328, November 6, 2001.
12. M. P. Hojilla-Evangelista, L. A. Johnson, and D. J. Myers, Cereal Chem., 69, 643647 (1992).
13. M. P. Hojilla-Evangelista and L. A. Johnson, J. Am. Oil Chem. Soc., 79, 815823 (2002).
14. F. Feng, D. J. Myers, M. P. Hojilla-Evangelista, K. A. Miller, L. A. Johnson, and S. K.
Singh, J. Am. Oil. Chem. Soc., 79, 707709 (2002).
15. J. R. Kwiatkowski and M. Cheryan, J. Am. Oil Chem. Soc., 79, 825830 (2002).
16. M. Cheryan, U.S. Patent 6,433,146, August 13, 2002.
17. D. D. Christianson, J. P. Friedrich, G. R. List, K. Warner, E. B. Bagley, A. C. Stringfellow,
and G. E. Inglett, J. Food Sci., 49, 229233 (1984).
18. D. D. Christianson and J. P. Friedrich, U.S. Patent 4,495,207, January 22, 1985.
170 CORN OIL
19. E. Ronyai, B. Simandi, S. Tomoskozi, A. Deak, L. Vigh, and Z. Weinbrenner, J. Supercrit.

Fluids, 14, 7581 (1998).
20. K-D. Stolp and R. W. Stute, U.S. Patent 4,341,713, July 27, 1982.
21. M. Bocevska, D. Karlovic, T. Jovan, and D. Pericin, J. Am. Oil Chem. Soc., 70, 12731277
(1993).
22. D. J. Karlovic, M. Bocevska, J. Jakovlevic, and J. Turkulo, Acta. Alimentaria, 23, 389
400 (1994).
23. D. Verser and T. Eggeman, U.S. Patent 6,509,180, January 21, 2003.
24. F. T. Orthoefer and R. D. Sinram, in S. A. Watson and P. E. Ramstad, eds., Corn: Chemistry
and Technology, American Association of Cereal Chemists, St. Paul, MN, 1987, pp. 535
552.
25. R. A. Moreau, V. Singh, and K. B. Hicks, J. Agric. Food Chem., 49, 37933795 (2001).
26. V. Singh, R. A. Moreau, A. E. Haken, S. R. Eckhoff, and K. B. Hicks, Cereal Chem., 77,
692695 (2000).
27. M. Dowd, J. Chromatogr. A., 816, 185193 (1998).
28. R. Antoniassi, W. Esteves, and A. J. D. Meirelles, J. Am. Oil Chem. Soc., 75, 14111415
(1998).
29. Z. Leibovitz and C. Ruckenstein, J. Am. Oil Chem. Soc., 60, 395399 (1983).
30. C. G. Pina, and A. J. A. Meirelles, J. Am. Oil Chem. Soc., 77, 553559 (2000).
31. T. Verleyen, R. Verhe, L. Garcia, K. Dewettinck, A. Huyghebaert, and W. De Greyt, J.
Chromatogr. A., 921, 277285 (2001).
32. T. Verleyen, U. Sosinka, S. Ioannidou, R. Verhe, K. Dewettinck, A. Huyghebaert, and
W. De Grety, J. Am. Oil Chem. Soc., 79, 947953 (2002).
33. K. B. Hicks and R. A. Moreau, Food Technol., 55, 6367 (2001).
34. R. A. Moreau, K. B. Hicks, R. J. Nicolosi, and R. A. Norton, U.S. Patent 5,843,499,
December 1, 1998.
35. C. M. Buchanan, N. L. Buchanan, J. S. Debenham, M. C. Shelton, M. D. Wood, M. J.
Visneski, B. K. Aruumugam, J. K. Sanders, L. R. Lingerfelt Jr., and L. Blair, U.S. Patent
6,352,845, March 5, 2002.
36. C. M. Buchanan, N. L. Buchanan, J. S. Debenham, M. C. Shelton, and M. D. Wood, U.S.
Patent 6,388,069, May 14, 2002.
37. D. Firestone, in Physical and chemical characteristics of oils, fats, and waxes, American
Oil Chemists Society, Champaign, Illinois, 1999, pp. 3132.
38. Anonymous. (2003). USDA National Nutrient Database for Standard Reference, Release
15. Nutrient Data Laboratory Home Page. Available: http://www.nal.usda.gov/fnic/food-
comp/Data/SR15/sr15.html.
39. F. D. Goffman and T. Bohme, J. Agric. Food. Chem., 49, 49904994 (2001).
40. R. A. Moreau, V. Singh, A. Nunez, and K. B. Hicks, Biochem. Soc. Trans, 28, 803806 (2000).
41. P. J. White and E. J. Weber, in L. A. Johnson and P. J. White, eds., Corn Chemistry and
Technology, Second Edition, American Association of Cereal Chemists, St. Paul, MN.
In press.
42. K. J. Leto and J. F. Ulrich, U.S. Patent 6,248,939, June 19, 2001.
43. R. A. Moreau, in: F. D. Gunstone, ed., Vegetable oils in food technology, Sheffield
Academic Press: Sheffield, 2002, pp. 278296.
REFERENCES 171
44. L. R. Strecker, A. Maza, and F. G. Winnie, in D. R. Erickson, ed., Proceedings of the World
Conference on Edible Fats and Oils Processing: Basic Principles and Modern Practices,
45. W. C. Byrdwell, W. E. Neff, and G. R. List, J. Agric. Food Chem., 49, 446457 (2001).
46. W. E. Neff, R. O. Adlof, G. R. List, and M. El-Agaimy, J. Liq. Chromatogr, 17, 39513968
(1994).
47. J. W. Fahey, B. A. Clevidence, and R. M. Russell, Nutr. Rev., 57, S34S40 (1999).
48. B. Ham, B. Butler, and P. Thionville, LC-GC, 18, 11741181 (2000).
49. R. E. Worthington and H. L. Hitchcock, J. Am. Oil Chem. Soc., 61, 10851100 (1984).
50. J. M. Snyder, J. W. King, S. L. Taylor, and A. L. Neese, J. Am. Oil Chem. Soc., 76, 717721
(1999).
51. R. A. Moreau, B. D. Whitaker, and K. B. Hicks, Prog. Lipid Res., 41, 457500 (2002).
52. V. Singh, R. A. Moreau, and P. H. Cooke, Cereal Chem., 78, 436441 (2001).
53. V. Singh and R. A. Moreau, Cereal Chem., 80, 123125 (2003).
54. V. Singh, R. A. Moreau, K. B. Hicks, and S. R. Eckhoff, Cereal Chem., 78, 4649 (2001b).
55. C. Wang, J. Ning, P. G. Krishnan, and D. P. Matthees, J. Am. Oil Chem. Soc., 75, 609613
(1998).
56. R. A. Moreau, K. B. Hicks, and M. J. Powell, J. Agric. Food Chem., 47, 28692871 (1999).
57. T. Saldeen, D. Y. Li, and J. L. Mehta, J. Am. Coll. Cardiol., 34, 12081215 (1999).
58. R. A. Parker, B. C. Pearce, R. W. Clark, D. A. Gordon, and J. J. Wright, J. Biol. Chem.,
268, 1123011238 (1993).
59. E. J. Weber, in S. A. Watson and P. E. Ramstad, eds., Corn: Chemistry and Technology,
Am. Assoc. Cereal Chem., St. Paul, MN, 1987, pp. 311350.
60. O. Sommerburg, J. E. E. Keunen, A. C. Bird, and F.J. G. M. Kuijk, Brit. J. Ophtalmol., 82,
907910 (1998).
61. A. Subagio and N. Morita, Food Res. Int., 34, 183188 (2001).
62. M. G. Enig, L. A. Pallansch, J. Sampugna, and M. Keeney, J. Am. Oil Chem. Soc., 60,
17881795 (1983).
63. J. Exler, L. Lemar, and J. Smith. (2001). USDA Food Nutrition and Information Center.
Available: http://www.nal.usda.gov/fnic/etext/000020.html.
64. C. Ip, Am. J. Clin. Nutr., 66, 1523S1529S (1997).
65. R. E. Lawson, A. R. Moss, and D. I. Givens, Nutr. Res. Rev., 14, 153172 (2001).
66. C. Gertz, S. Klostermann, and S. P. Kochhar, Eur. J. Lipid Sci. Technol., 102, 543551
(2000).
67. N. Shen, S. Duvick, P. White, and L. Pollak, J. Am. Oil Chem. Soc., 76, 14251429
(1999).
68. E. Sebben, D. C. Slaughter, and R. P. Singh, J. Food Proc. Pres., 22, 265282 (1998).
69. G. Yildiz, R. L. Wehling, and S. L. Cuppett, J. Am. Oil Chem. Soc., 79, 10851091 (2002).
70. K. Sundram, D. Perlman, and K. C. Hayes, U.S. Patent 5,874,117, February 23, 1999.
71. G. W. Meijer, Lipid Technol., November, 129132 (1999).
72. T. J. Howell, D. E. MacDougall, and P. J. H. Jones, J. Lipid Res., 39, 892900 (1998).
73. J. D. Radcliffe, C. C. King, D. M. Czajka-Narins, and V. Imrhan, Plant Foods Human
Nutr., 56, 5160 (2001).
172 CORN OIL
74. A. H. Lichtenstein, L. M. Ausman, W. Carrasco, J. L. Jenner, J. M. Ordovas, and E. J.

Schaefer, Arterioscler. Thromb., 13, 154161 (1993).
75. A. van Tol, A. H. M. Terpstra, P. van den Berg, and A. C. Beynen, Atherosclerosis, 147,
8794 (1999).
76. T. A. Wilson, A. P. DeSimone, C. A. Romano, and R. J. Nicolosi, J. Nutr. Biochem., 11,
443449 (2000).
77. R. E. Ostlund Jr., S. B. Racette, A. Okele, and W. F. Stenson, Am. J. Clin. Nutr., 75, 1000
1004 (2002).
78. K. H. Wagner, R. Tomasch, and I. Elmadfa, Eur. J. Nutr., 40, 161167 (2001).
79. M. Lemcke-Norojarvi, A. Kamal-Eldin, L-A. Appelqvist, L. H. Dimberg, M. Ohrvall, and
B. Vessby, J. Nutr., 131, 11951201 (2001).
80. W. J. Wechter, U.S. Patent 6,242,479, June 5,2001.
81. L. J. Schurgers, M. J. Shearer, B. A. M. Soute, I. Elmfadfa, J. Harvey, K-H. Wagner,
R. Tomasch, and C. Vermeer, J. Lipid Res., 43, 878884 (2002).
82. I. Elmadfa and E. Park, Eur. J. Nutr., 38, 286 (1999).
83. Anonymous. (2001). National Association of Margarine Manufacturers. Available: http://
www.margarine.org.
84. G. R. List, T. L. Mounts, F. Orthoefer, and W. E. Neff, J. Am. Oil Chem. Soc., 72, 379382
(1995).
85. Anonymous. (2003). US Food and Drug Administration, Federal Register 68 (133) pp.
4143341506. Available: http.//www.access.gpo.gov/su_docs/fedreg/a030711c.html.
86. Green cottage resource site. (2003). Available: http://www.greencottage.com/oils/
corn.html.
87. Kentucky Corn Growers Association. (2003). Available: http://www.kycorn.org/corn-
education/cornclass/8uses.pdf.
5
Cottonseed Oil
Richard D. OBrien,
Lynn A. Jones, C. Clay King,
Phillip J. Wakelyn, and Peter J. Wan
1. INTRODUCTION
Cotton is both a food (cottonseed oil) and a fiber (cotton lint) crop. For each 100 kg
(220.46 lbs) of cotton fiber produced, the plant also produces about 150 kg (330.69 lbs)
of cottonseed. The cotton plant primarily is and has always been grown for the
textile fiber (cotton) component of the plant. Consequently, the production of
seed, which varies directly with cotton fiber production, is dominated by factors
determining the production of cotton fiber. Cottonseed is about 1520% of the value
of the cotton crop.
A typical cottonseed crushing operation will separate the seed into oil [160 kg/t
(320 lbs/t)], hulls [260 kg/t (540 lbs/t)], meal [455 kg/t (910 lbs/t)], and linters
[83.5 kg/t (167 lbs/t)] (1). The hulls and meal are sources of vegetable protein
feed for animals; and the linters are used as a chemical cellulose source in personal
care products, in batting for upholstered furniture and mattresses, and in high-
quality paper (2).
Cottonseed oil, Americas original vegetable oil, dominated the United States
vegetable oil market for almost 100 years. The English and European vegetable
oil industry was based on a variety of oilseeds and tree fruits available in the
home countries and their colonies, bu t cottonseed was the principal raw material
for vegetable oil processors in the United States until the mid-twentieth century. In
173
174 COTTONSEED OIL
a little more than 50 years, through research and experimentation, chemists have
developed a clear, odorless, bland flavored cottonseed oil and a creamy, white
shortening that set the standards for edible fats and oils worldwide. The scientific
and technical advances developed to process cottonseed oil became the corner-
stones of the edible fats and oils industry as it is known today. Numerous
processes were developed or perfected especially for cottonseed oil, which later
found application for other oils. Today, vegetable oil processors worldwide,
including the United States, have a wide range of raw materials to choose from,
but cottonseed pioneered the American vegetable oil industry. Table1 identifies
most of the important dates and events in the history of cottonseed oil processing
and use (3).
2. COTTONSEED OIL INDUSTRY DEVELOPMENT
Crushing of cottonseed for oil was documented in the early Hindu medical books; a
topical medicine was prepared by pounding the cottonseeds followed by boiling to
extract the oil. An early Chinese process of oil extraction was the wedge press; but
it is not clear if this was used to extract oil from cottonseed. For many centuries the
use of cottonseed oil did not develop much beyond this crude stage and was con-
fined to local areas. Then, during the first part of the nineteenth century, plants in
Europe began to crush small quantities of Egyptian cottonseed. Fats and oils
shortages, caused by rapid population increases during the Industrial Revolution
and the English blockade during the Napoleonic Wars, invoked the natural law
of supply and demand, which increased the cost of the available fats and oils pro-
ducts to levels that made it impossible for many Europeans to afford them in their
diets or for illumination. The resultant demand for less expensive fats and oils
products led businessmen to extract oil from a variety of oilseeds and nuts; included
in these endeavors was cottonseed from Egypt and India (4, 5).
The sparsely populated United States had adequate supplies of animal fats while
western Europe was experiencing shortages. Nevertheless, the extraction of oil from
cottonseed became an attractive solution for another problem. Cotton cultivation
had increased rapidly in the United States after Eli Whitney invented the cotton
gin in 1793 creating a surplus supply of cottonseed, after that required for planting,
fertilizer, and animal feed. The excess cottonseed stocks became huge, worthless,
rotting piles that dotted the countryside in the southern states. A few entrepreneurs
began crushing cottonseed for oil between 1820 and 1830, but none of these ven-
tures survived more than a few years. Limited demand, as well as difficult and
expensive transportation to the markets, were formidable obstacles, but probably
foremost was the hull problem with the cottonseed varieties grown in America.
Oil extraction from the American cottonseed was hampered by the fact that the
seed of the short-staple variety grown has a tough hull covered with short cotton
fibers called linters. These residual fibers and hulls absorbed valuable oil, resulting
in low oil production and poor-quality, oily livestock feed. European cottonseed
crushings had been with Egyptian long-staple cotton varieties, which did not
have a tough hull covered with linters. In 1857, William Fee invented a huller,
COTTONSEED OIL INDUSTRY DEVELOPMENT 175
TABLE 1. Some Important Dates in the Evolution of Cottonseed Oil Processing

and Utilization.
Year Event Effect
1793 Eli Whitney invented the cotton gin Cotton cultivation substantially increased
Early Winterization process developed in Separation of an oil into liquid and solid
1800s France for the illumination industry fractions
1857 William Fee invented the Separation equipment which solved a
cottonseed huller major oil extraction problem
1859 Petroleum industry launched Eliminated illumination and lubrication as
potential cottonseed oil applications
1880 Alexander Winters bleaching patent Color adsorbed with Fullers earth
1880s Refining with caustic soda Alkali refining introduced and accepted
1882 Olive oil adulteration European import tariffs imposed with
complete expulsion from Italian markets
1882 Cottonseed oil bottled in the U.S. Limited acceptance due to flavor & odor
1883 O. G. Burnham patented a Allowed the addition of higher levels of
refrigerated chill roll cottonseed oil to lard
1884 Lard adulteration U.S. label restrictionsmandatory
identification of lard compound
1886 W. B. Albright bleaching patent Steam used to recover oil from earth
Late Acidity analysis Chemists learned that FFA level in crude
1880s cottonseed oil was a good quality indicator
1891 Henry Eckstein introduced steam Removed the objectionable flavors and
distillation for cottonseed oil odors from the oil
1897 Sabatier discovered vapor phase Process to convert a liquid oil to a
hydrogenation using nickel catalyst semi-solid
1899 Wesson Oil R
introduced Cottonseed oil deodorized with a vacuum
and superheated steam
1903 William Norman granted British Assigned to Joseph Crossfied & Son, a
patent for hydrogenation British concern that used the process
to harden whale oil for soap manufacture
1909 Procter and Gamble acquired the Developed the hydrogenation process for
U.S. rights to Normans patent food oil use
1911 CriscoR
shortening introduced First hydrogenated all-vegetable shortening
Early Closed internal scraped wall heat Shortening and margarine plasticization
1930s exchanger developed by Votator R
equipment offering improved sanitation
and product uniformity
1930s Pre-Press Solvent Extraction Oil recovery improved from 80% to 97%
Process Developed
1933 Introduction of superglycerinated Initiation of the era of tailor-made fats and
shortening oils products
1941 World War II Cottonseed oil shortages forced
soybean oil utilization
1944 Soybean oil production outranks Cottonseed oil shortages and lower cost
cottonseed oil for the first time soybean oil utilization
1950 Soybean oil replaces cottonseed oil Hydrogenation and emulsifiers allow
as the dominate oil for shortenings utilization of the more economical oil
1951 Cottonseed oil looses dominance Transportation and grocery store handling
in margarine to soybean oil improvements allowed product changes
1958 Cottonseed oil use in salad oil Flavor problems with more economical
manufacture drops below 50% soybean salad oil decreased
1960s Expander Solvent Extraction Improved oil recovery and quality
Technology Introduced
176 COTTONSEED OIL
which effectively separated the tough, linty hulls from the meats to solve the major
extraction problem with American cottonseed. (6, 7). Fee also invented an
improved hydraulic press, and although many advances were made in the mechan-
ical process of cottonseed crushing, the basic technology available remained
unchanged well into the twentieth century.
Cottonseed oil usage for illumination purposes in lamps and candles to supple-
ment whale oil and lard was curtailed by the petroleum industry in 1859. Apart
from its use as a smokeless oil for miners lamps during the late nineteenth century,
cottonseed oil would have no future as an illuminant (8). However, the processing
techniques developed in connection with lighting ultimately became important to
the food industry. Candles require a wax or fat with a high melting point to remain
solid until melted from the flame heat; conversely, a lamp oil must be a clear liquid
to maintain a flame. During the early nineteenth century, a fractionation process
was developed in France to make vegetable oils more like whale oil and tallow.
This technique, called winterization, was achieved by holding the oil in outside
tanks during the winter. The low winter temperatures caused the higher melting
fractions to crystallize and settle to the bottom of the tank, leaving a top layer of
clear oil. The clear oil could be obtained by either decanting from the top or filter-
ing the mixture to separate the liquid and solid fractions (9). The hard fractions
found application as a tallow replacement in candle making, and the liquid fraction
could be used as a lamp oil.
Even with all of the early setbacks, cottonseed crushing and refining became a
profitable venture in the United States after 1870. Large quantities of cottonseed oil
were exported for use in soap manufacture and some found its way into cooking,
salad oils, and the oleomargarine product developed by the French chemist,
Hippolyte Mege-Mouries. Adulteration of olive oil with cottonseed oil resulted in
import tariffs in the olive growing countries and complete expulsion from Italy in
1883. These restrictions curtailed cottonseed oil exports to create an oversupply of
cottonseed oil, which again decreased the value (6).
High lard prices offered an opportunity to increase the domestic consumption of
cottonseed oil. Lard was the choice fat for cooking and baking in many parts of the
western world. One of the chief reasons was the particular consistency, which was
optimum for incorporation into breads, cakes, pastries, and other baked products. It
had become common practice to mix beef fat or tallow with lard, so adulteration
with cottonseed oil was a natural step for the use of the inexpensive oil. Initially,
meat packers secretly added cottonseed oil to lard. This practice was uncovered in
1884, when Armour and Company, seeking to corner the lard market, found that it
had purchased more lard than the existing hog population could have produced.
Public disclosure of this practice led to a Congressional investigation, which
made identification of these mixtures with the descriptive name lard compound
or compound mandatory (6, 10).
Major improvements in oil processing had to take place before substantial quan-
tities of cottonseed oil could replace lard as the preferred fat for baking and frying.
Crude cottonseed oil was unsuitable for use in most food applications because of
its dark color, high free fatty acid content, and objectionable flavor and odor.
Cottonseed oils crushed in Europe and Asia prior to the nineteenth century did not
have the objectionable qualities experienced after technology improvements to
improve yield were implemented. Oilseed milling became an established trade in
Europe, without the benefit of any substantial technology advancements since the
days of the Roman Empire. These inefficient practices produced an oil with a mild
flavor and odor due to a very low level of nontriglycerides. Oil quality changed with
the improvements in milling machinery and technology during the seventeenth and
eighteenth centuries. Extraction technology progressed from edgestone to wedge
press to hydraulic press during this period. Hydraulic pressing was the predominant
means used to separate oil from cottonseed for most of the nineteenth century. Eco-
nomics promoted research activity to recover the oil left in the press cake; mechan-
ical pressing left nearly 20% of the available oil in the press cake. This effort led to
the prepress solvent extraction process in the 1930s that was able to recover better
than 97% of the available oil in cottonseed. Further productivity and quality
demands brought expander-solvent extraction technology into cottonseed crushing
in the late 1960s (11). The more efficient high-pressure expression and solvent
extraction systems provided higher oil yields, which contained substantial quanti-
ties of nontriglyceride materials like gums, color bodies, and other impurities with
strong flavors and odors. As a result of the presence of these impurities, these oils
require more processing to provide a palatable product (12, 13).
Acid-refining techniques were introduced in the United States during the last
half of the eighteenth century but never assumed importance. Alkali-refining pro-
cedures developed in Europe around 1840 using calcium and potassium hydroxides,
also never found acceptance in the United States. Later, during the 1880s, alkali
refining using caustic soda was introduced and found acceptance. Cottonseed oils
were alkali refined in open kettles, and the foots were separated by gravitation until
continuous systems utilizing centrifuges were introduced (1416). Refining with
caustic soda reduced both free fatty acid content and a large portion of the red-
yellow color pigment, gossypol, in cottonseed oils. Caustic-refined cottonseed
oils were found acceptable for packing sardines and marginally acceptable for mix-
ing with lard when it was scarce and expensive. Neutralized cottonseed oil after
caustic refining still had an acrid flavor and somewhat unpleasant odor, which
limited consumer acceptance (17).
Additional color removal from cottonseed oil was necessary to make it more
acceptable for edible oil applications. People expected lard to be white; therefore,
cottonseed oil used to dilute lard or as an ingredient to replace lard had to be light in
color. At first, refiners used a bleaching technique previously used to remove the
orange color from palm oil. This bleaching process was accomplished by exposing
the oil to sunlight in large shallow tanks for up to 18 months on factory roofs. This
method had serious drawbacks: the lengthy process tied up capital for months, and
the sunlight deteriorated the keeping quality of the oil. Cottonseed oil processors
then began to experiment with bleaching agents, such as carbon and different kinds
of clays or bleaching earths, to adsorb the color pigments. The use of bleaching
earth for vegetable oil decolorization was an American achievement. In 1880,
Alexander Winters obtained a U.S. Patent for purifying animal and vegetable fats
178 COTTONSEED OIL
by treatment with pulverized fullers earth. Six years later, William B. Allbright
secured a patent for fullers earth applied with steam to facilitate removal of oil
from the used earth.
Although caustic refining removed free fatty acids and other impurities and
bleaching with fullers earth solved the color problem, the unpleasant flavor and
odor limited the acceptance of cottonseed oil as a salad or cooking oil, as an ingre-
dient for margarines, and the levels that could be used in lard compounds or
substitutes. Attempts to chemically remove the offensive flavor or mask them
with spices or flavors were unsuccessful. The first successful attempt at actually
removing odor and flavor consisted of blowing a current of live steam through
oil at elevated temperatures and atmospheric pressure. Reportedly, Henry Eckstein,
a Fairbank Company chemist, observed this technique at an English firm and under-
stood its importance for the future of cottonseed oil. He introduced this technique
around 1891, and it was quickly adopted by most American processors with other
refinements identified by Boyce, Cuff, and others. Later, David Wesson perfected
and introduced a deodorization process that exposed oil to superheated steam in a
vacuum. This process provided a bland-flavored oil with improved stability or keep-
ing quality. Deodorization enabled processors to increase the amount of cottonseed
oil in lard compounds to 80% or more (7, 8, 17, 18).
Cottonseed oil was bottled in the United States as early as 1882, some of it with
labels printed in French, Spanish, and Italian to make immigrants think that they
were buying olive oil from the country of their origin. Acceptance of liquid cotton-
seed oil was limited by the acrid flavor and unpleasant odor (9). In 1899, David
Wesson convinced the Southern Cotton Oil Company to process, package, and dis-
tribute Wesson Oil. This cottonseed salad oil was caustic refined, adsorption
bleached, and deodorized with the Wesson technique after winterization. The win-
terization process, a nineteenth-century French development to make whale oil sub-
stitutes, removed the high melting fractions from cottonseed oil to maintain clarity
at refrigerator temperatures. This requirement became a necessity with the intro-
duction of mechanical refrigeration and the increased popularity of mayonnaise,
salad dressings, and other emulsified sauces. Oils that solidify at refrigerator tem-
peratures cause emulsions to break, resulting in an unsightly product separation.
Introduction of the catalytic hydrogenation process gave the cottonseed oil pro-
cessors independence from the meat-packing industry and initiated a new era. Com-
pounds were dependent on oleo stearine as the stiffening agent, which were
supplied by the meat packers who recognized their monopolistic situation and
maintained a high cost for this material. Hydrogenation provided a means for con-
verting liquid cottonseed oil to semisolid with a consistency similar to the meat-fat
stearine products. Sabatier discovered vapor-phase hydrogenation, using nickel as a
catalyst, in Paris in 1897. Six years later, William Norman, a German working in
England, obtained a British patent for hydrogenation, and in 1906 commercial
hydrogenation of whale oil began. Procter & Gamble acquired the U.S. rights to
the Norman patent in 1909 and produced the first all-vegetable shortening in
1911. The product was named kispo, which was later changed to Crisco, short
for crystallized cottonseed oil. Crisco is still the leading household shortening in
the year 2004 and even though it has undergone several composition changes, it still
utilizes some cottonseed oil in the formulation. The success of this hydrogenated
cottonseed oil inclined other manufactures to litigation actions, which led to a
U.S. Supreme Court decision invalidating Procter & Gambles exclusive use of
the Norman patent and cleared the way for all processors to employ the hydrogena-
tion process (8, 19).
Meat packers continued to offer lard compounds, later shortened to com-
pounds, employing the hydrogenation process only for the production of hardened
oils to serve as occasional substitutes for oleostearine. Cottonseed oil producers had
the foresight to abandon the lard substitute product concept and description to offer
their hydrogenated products as a new food ingredient that has become known as
shortening. Thereafter, shortening development followed two divergent courses;
all-hydrogenated or a blended type formulation somewhat similar to the compound
concept.
The hydrogenation process gave cottonseed oil shortenings a definite advantage
over the compound shortenings offered by the meat packers. This key process per-
mitted them to change the composition of the inherently liquid oil products to pro-
vide a more consistent product. The properties of the hydrogenated cottonseed oil
shortenings could be controlled, whereas, the consistency of lard varied according
to the time of the year, the variety and age of the hogs, and the type of feed. The
superior neutrality, oxidative stability, and uniformity of the hydrogenated cotton-
seed oil shortenings found favor with commercial bakers and homemakers alike
at the expense of the meat-fat products. This was confirmed by the consumers
willingness to consistently pay higher prices for the hydrogenated cottonseed oil
shortenings.
Solidification of shortening and margarine into smooth plastic products also
posed formidable challenges for the edible oil industry. Pure lard contains relatively
little high-melting fractions; therefore, slow cooling was a satisfactory procedure
for production of smooth, plastic lard products. Agitation of shortening products
in jacketed, water chilled tanks, which had been the usual practice for lard, resulted
in grainy products with the lard compounds containing cottonseed oil as well as the
hydrogenated shortenings. An internally refrigerated chill roll developed for crys-
tallizing lard compounds was patented by O. G. Burnham in 1883. The chill roll
consisted of a revolving hollow metal cylinder chilled by circulating cold or brine
water. A thin layer of lard compound or shortening applied to the revolving drum
solidified and was scraped off into a mixing unit, called a picker pan, where crystal-
lization continued and air was incorporated. After picking, the product was forced
through a small orifice at high pressure to complete the homogenization of the air
and oil. This crystallization process allowed incorporation of higher levels of
cottonseed oil in lard compound products (20).
Initially, margarine was chilled by pouring or spraying molten emulsions into a
vat of running cold water. The emulsion entered at one end of the vat and solidified
before it reached the other end. The solidified material floated to the surface in the
form of a flaky mass, which was skimmed off. After the excess water was drained
from the solidified emulsion, it was worked to achieve the desired consistency. The
180 COTTONSEED OIL
chill roll used to plasticize lard compounds and shortenings was adopted and used
in nearly the same manner (21).
In the early 1930s, development of improved heat-transfer equipment for freez-
ing ice cream led to the perfection of a closed, continuous, internal scraped wall
chilling unit, which replaced chill rolls for plasticizing shortening and margarine
products. Votator1 equipment chilled and plasticized shortenings and margarines
in seconds as it flowed continuously through a closed, mechanically controlled sys-
tem. Adoption of this system offered a number of advantages: (1) improved sanita-
tion afforded by a closed system; (2) improved product uniformity provided by
more accurate control of nitrogen injection, temperature, pressure, agitation, and
throughput; (3) cost reductions resulting from labor savings, reduced space require-
ments, and lower refrigeration demand; and (4) a reduction in product losses (22).
Versions of these closed internal chilling systems are still employed to plasticize
shortenings and margarines.
In 1933, the introduction of superglycerinated shortening brought about signifi-
cant changes for the baker and the shortening industry. These shortenings contained
monoglycerides and diglycerides, which promoted emulsification of fat in water.
Emulsified shortenings allowed bakers to produce more moist, higher volume, longer
shelf life cakes with a fine grain and texture, lighter icings and fillings with higher
moisture levels, and yeast-raised bread and rolls with an extended shelf life. This
development added a new dimension to the fats and oils industry, which ushered in
the era of tailor-made shortenings. Shortening applications were expanded beyond
baking usage to snacks, dairy analogs, confections, foodservice, and other areas. At
the beginning of World War II, the United States domestic use of cottonseed oil
was 65% in shortenings, 8% in margarines, and 18% for miscellaneous food
products, which included salad dressing, mayonnaise, salad oil, and packing of
fish and cured meats (23).
2.1. Post World War II Cottonseed Oil Markets

Cottonseed oil was the edible oil of choice in the United States for many years, both
as a liquid for salad and cooking oils and hydrogenated to produce shortenings and
margarines. During the beginning years of the U.S. soybean oil industry, its flavor
and reversion problems were cottonseed oils allies to support cottonseed oils
dominant position and premium pricing. This dominant position was maintained
well into the 1950s until the boll weevil and New Deal crop restrictions opened
the door to U.S. markets for competing vegetable oil by limiting the availability
of cottonseed oil (6). Cottonseed oil shortages and an increased demand for edible
oil encouraged research by government and industrial laboratories to develop pro-
cesses to make soybean oil more palatable. Cottonseed oil lost its dominant position
but retained its premium pricing over soybean oil and the food fat it originally
challenged for market sharelard.
Cottonseed oil shortages during World War II allowed soybean oil production to
grow despite serious flavor, stability, and reversion problems. Reminiscent of cot-
tonseed oils early days when it was used to adulterate lard, fats and oils processors
incorporated as much soybean oil as possible into cottonseed-oil-based formula-
tions to take advantage of the 49 cent per pound lower cost. Initially, the amount
that could be blended into margarine and shortening products was restricted to less
than 30% to ensure that an objectionable flavor and odor would not develop.
Soon after World War II, soybean surpassed cottonseed as the principal source of
vegetable oil in the United States. In the early 1950s, researchers identified the
cause of soybean oils offensive flavor as the 78% linolenic fatty acid content
with a classic experiment. Nine percent linolenic fatty acid was intersterified into
cottonseed oil, which typically contains less than 1% of this unsaturated fatty acid.
Blind flavor panels profiled this modified cottonseed oil as soybean oil. This finding
gave processors a means to improve the acceptance of soybean oil: hydrogenation
to reduce the linolenic fatty acid content (24). Margarine and especially shortening
products, were the most likely candidates for hydrogenated soybean oil utilization
because these formulations already employed hydrogenated cottonseed oil base-
stocks. However, the beta crystal habit of soybean and most other vegetable oils
could not duplicate the smooth, creamy consistency contributed by the beta-prime
crystal habit of cottonseed oil. Therefore, it was necessary to formulate soybean oil-
based shortenings requiring a smooth, plastic consistency with some hydrogenated
cottonseed oil or another beta-prime crystal promoter. However, this was not the
case with margarine; the use of multiple high-trans basestocks, uniform low-
temperature shipping, storage, and display practices in the United States allowed
table-grade margarines and spreads to be formulated with 100% soybean oil.
Soybean salad oil with an iodine value similar to winterized cottonseed salad oil
and the linolenic fatty acid reduced to 34% was introduced to the U.S. market
around 1958. This product was lightly hydrogenated and subsequently winterized
to remove the hard fractions developed during hydrogenation. Even with additional
hydrogenation costs, this salad oil cost less than winterized cottonseed salad oils.
Specially processed soybean salad oil was eventually accepted as a replacement
for cottonseed salad oil, but most processors proceeded cautiously, offering blends
with cottonseed during the introductory periods, beginning with 2030% soybean
salad oil and gradually increasing the proportion, eventually reaching 100% several
years later.
With the success of soybean oil and a growing consumer appreciation for nutri-
tional foods, other vegetable oils began to more actively compete for a share of the
vegetable oil market. The importance of the essential fatty acids, linoleic and lino-
lenic, were recognized and diet modification to restrict intake of cholesterol, satu-
rated fatty acids, trans fatty acids, and total fat calories have been recommended to
lower serum cholesterol levels in order to reduce the risk of heart attacks. The fats
and oils usage data on Table 2 (2528) reflects these trends: (1) a move away from
animal fats to vegetable oils; (2) replacement of previously established fats and oils
with different source oils; (3) introduction of new vegetable oils; (4) a rise and fall
of some individual source oils; (5) source oil changes reflecting the results of
medical studies; (6) introduction of new oil seed hybrids; and more. Ironically,
the implications in this table could be used as the basis for a social commentary
on life in the United States in the late twentieth, early twenty-first century. For
cottonseed oil, these data show a rapid decline in U.S. domestic usage volume after
1950 to a low point in the 1980s when it started to regain domestic market share.
182 COTTONSEED OIL
TABLE 2. U.S. Edible Fats and Oils Disappearance and Per Capita Consumption.
1000 Metric Tons

Year 1950 1960 1970 1980 1990 2000
Canola 261.7 791.1

Coconut 58.5 78.0 357.4 468.1 406.9 439.1
Corn 101.2 140.6 201.9 305.3 521.2 776.1
Cottonseed 655.4 555.7 404.2 237.2 386.0 305.7
Olive 35.8 23.1 30.4 26.3 95.7 206.4
Palm 0.5 82.6 135.6 116.1 170.1
Palm Kernel 11.8 24.0 42.6 NR 164.2 110.2
Peanut 46.7 28.1 87.5 50.8 89.4 110.7
Safflower 45.4 26.3 46.3
Soybean 655.9 1365.8 2836.3 4134.1 5517.6 7352.8
Sunflower 29.0 90.7 161.9
Lard 929.9 856.8 746.2 464.0 374.2 436.4
Tallow 70.8 148.8 235.0 451.3 433.2 679.5
Butter 601.9 504.9 487.6 461.3 496.7 463.6
Total 3167.9 3726.3 5557 6763.1 8979.9 12049.8
Per capita consumption, kg
Vegetable oils 10.9 12.1 17.7 20.4 23.9 28.6
Animal fats 9.9 8.4 6.4 5.5 4.3 5.2
Total 20.8 20.5 24.0 25.9 28.2 33.8
NR not reported
This renewed popularity was most likely at the expense of tallow and palm oil,
which lost consumer appeal due to unfavorable publicity highlighting nutritional
concerns with saturated fatty acids or cholesterol; both of these edible fats and
oils crystallize in the beta-prime habit like cottonseed oil.
Two primary contributing factors to the supply of cottonseed for processing, and
thus cottonseed oil, in the United States are (1) the supply of seed is determined by
the demand for cotton lint, which has been heavily influenced by the cotton provi-
sions of the U.S. Farm Program, and (2) the raw material usage of cottonseed has
changed dramatically at the end of the twentieth century. Cotton acreage planted in
the United States is largely determined by the support programs of the USDA and is
based on world and domestic stocks and the expected lint requirements. Thus, the
supply of cottonseed cannot respond to market signals the way soybean production
has since the 1940s. The second influence on cottonseed stocks for crushing has
been the competition for whole seed by the dairy-feeding industry. Shifts in dairy
production in the past two decades have put a premium on cottonseeds unique
combination of protein, energy and fiber, and this has affected the availability of
whole cottonseed for dairy feeding. The tonnage of cottonseed crushed in the
United States dropped by a third between 1980 and 2002 while the total supply
of seed rose during the same period. These divergent trends resulted in approxi-
mately 74% of the cottonseed available in 1980 in the U.S. being crushed for
products as opposed to crush levels near 36% twenty-two years later. These shifts
are indicated in Table 3 (29, 30).
TABLE 3. Cottonseed Supply and Disappearance Data for Selected Years (1,000 Metric Tons).
Supply Disappearance
Year Beginning Beginning Amount of Feed, Seed Ending
Aug. 1 Stocks Production Total Crush Supply Crushed (%) Exports and Residual Total Stocks
18991900 3,767 2,249 53%

19041905 5,413 3,035 56%
19141915 6,491 6,501 5,244 81% 3 1,255 6501a
19291930 38 5,812 5,870 4,551 78% 1,279 5830a
19371938 38 7,116 7,154 5,739 81% 1,112 6848a
19401941 36 4,795 4,832 3,990 83% 723 4713a 119
19501951 261 3,724 3,986 3,380 85% 5 543 3926a 60
19601961 95 5,340 5,435 4,855 89% 5 405 5264a 171
19701971 73 3,690 3,763 3,382 90% 35 148 3,565 198
19801981 960 4,056 5,016 3,698 74% 121 836 4,655 361
19901991 332 5,415 5,749 3,056 53% 48 2,054 5,158 591
20002001 249 5,839 6,427 2,498 39% 213 3,329 6,039 396
a
Calculated.
184 COTTONSEED OIL
The export market for U.S. cottonseed oil also affects supplies (3133). Avail-
able cottonseed oil is consumed domestically or exported, most likely to the most
profitable market. Supply and demand practices have channeled the available sup-
plies to the highest bidders which long maintained a premium price structure, while
disposing of the available cottonseed. Cottonseed oil exports were a strong market
during the initial years of the U.S. cottonseed crushing industry and again in the
1980s. Foreign tariffs in 1882 and U.S. imposed tariffs in 1921 caused export
declines (6). In the 1990s, price competition from South American countries and
the above-mentioned domestic seed supply problems again ate away at export
volumes.
The technologies, processes, and methods developed to produce shortening,
salad oil, margarine, and other edible oil products with cottonseed oil have been
adapted for use with soybean, corn, canola, and other oilseeds as well as the animal
fats; lard and tallow. However, fats and oils interchangeability is limited by the
physical and chemical characteristics of the individual source oil. Each source oil
has distinctive flavor characteristics, fatty acid compositions, and triglyceride
structures. Cottonseed oils properties have helped to maintain it as an important
source oil for food products worldwide. U.S. cottonseed oil has enjoyed a strong
export market, along with the high demand for its performance characteristic in
specific food products. Cottonseed oils functional characteristics; such as a pleas-
ing flavor described as nutty, good flavor stability resulting from an absence of tri-
unsaturates, which oxidize rapidly, and a beta-prime crystal habit probably due to a
high-palmitic fatty acid content, have helped to maintain it as a desirable vegetable
oil. These characteristics and the fact that it was Americas original vegetable oil
has made it the standard to which other oils are compared.
2.2. World Production of Vegetable Oils

Cottonseed oil dominated the U.S. and world edible oil markets until just before
World War II, when it was displaced by soybean oil. Cottonseed oil had occupied
this dominant position because of ample supplies, generally a lower price than com-
peting oils, and manufacturers preferred cottonseed oil for most edible products.
The major reason for the loss of this dominant position was an increased demand
for edible oils, which cottonseed oil could not meet. An adequate supply of raw
material was not available due to a declining demand for cotton. The cotton fiber
surplus resulted in cotton farm acreage being converted to growing soybeans. Initi-
ally, oil demand influenced soybean cultivation to fill the edible oil shortage. Later,
recognition of a world shortage of food protein changed this to make soybean oil
production secondary to the demand for meal.
About 80 countries in the world grow cotton (34). Planting time for cotton varies
by locality, varying from February to June in the Northern Hemisphere; harvest
time is in the late summer or early/late fall. In the Western Hemisphere, cotton
is cultivated between about 37 N and 32 S latitude and in the Eastern Hemisphere,
between about 47 N and 30 S. Cultivation of cotton differs markedly from one
country to another, depending on the degree of mechanization (35). The Peoples
COTTONSEED OIL PROPERTIES 185
TABLE 4. World Vegetable Oil Production.
2001/2002 World Vegetabe Oil Production (Metric Tons)
Rank With Without % U.S.

Order U.S. Usage U.S. Usage Usage
1 Soybean 28660.1 Palm 24614.4 0.86
2 Palm 24828.1 Soybean 20968.0 26.84
3 Canola 12174.5 Canola 11497.3 5.56
4 Sunflower 7554.2 Sunflower 7384.1 2.35
5 Peanut 4291.0 Peanut 4177.6 2.64
6 Cottonseed 3812.0 Cottonseed 3464.1 9.13
7 Coconut 3253.2 Coconut 2754.2 15.34
8 Palm Kernel 3103.5 Palm Kernel 2942.5 5.19
9 Olive 2524.7 Olive 2318.3 8.17
10 Corna 1115.4 Corn 512.6 45.96
Total 91316.8 Total 80633.2 11.70
a
U.S. production only.
Republic of China, the United States, India, Pakistan, and Uzbekistan are the largest
producers of cotton and cottonseed (34) and presumably the largest user of cotton-
seed oil. Turkey, Brazil, and Egypt also grow cotton and produce cottonseed oil,
mostly for domestic use (36). Brazil, Argentina, and the United States are the major
exporters of cottonseed oil.
The world vegetable oil markets and cottonseed oil have experienced many
changes over time. One of the more interesting changes for cottonseed oil is a
decrease in position of vegetable oil predominance from second place to sixth while
the quantities produced have steadily increased. Soybean oil has maintained a
dominant position over the past 40 years but has competition from oils grown pri-
marily for their oil content; sunflower oil from Europe, ground nut (peanut) oil from
Africa, rapeseed oil from Canada and Europe, and palm oil from tropical countries.
Another interesting point, illustrated in Table 4, is that palm oil easily becomes the
world leader when the U.S. consumption is taken away from the world market
statistics.
3. COTTONSEED OIL PROPERTIES
Cotton is a warm-weather shrub or tree of the Malvaceae family, the tribe Gossy-
pieae, and the genus Gossypium that grows naturally as a perennial but for commer-
cial purposes is grown as an annual (34). Botanically, cotton bolls are fruits (37).
The principal domesticated species of cotton of commercial importance are hirsu-
tum, barbadense, arboreum, and herbaceum (38). Many different varieties of these
species have been developed through conventional breeding to produce cotton
plants with improved agronomic properties and with improved cotton fiber and
cottonseed properties (39).
186 COTTONSEED OIL
The overwhelming proportion of cottonseed oil in the United States is obtained

from the seeds of Gossypium hirsutum, which is the shorter staple upland cotton
(about 92% of world cotton production and about 97% of U.S. production). Gossy-
pium barbadense, the extra long staple, or Egyptian, cotton, comprises only about
3% of the U.S. crop and about 8% worldwide. Gossypium aboreum and herbaceum
are of very minor importance but are grown in China, India, and Pakistan. Crude
cottonseed oil has a strong, characteristic flavor and a dark, reddish brown color
from the presence of highly colored material extracted from the seed. It is a member
of a particularly useful group of vegetable oils whose fatty acids consist substan-
tially of 16 and 18 carbon fatty acids containing no more than two double bonds.
Cottonseed oil is stable in the beta-prime crystal form, which is desirable in most
solidified products because it promotes a smooth, workable consistency usually
referred to as plasticity. Deodorized cottonseed oils reverted flavor is usually
described as nutty or nut-like, which is acceptable at higher degrees of oxidation
than other vegetable oils. Its characteristics make it a highly desirable food oil
for use in salad and cooking oils, shortenings, margarines, and specialty fats and
oils products.
A number of factors are responsible for minor variations in the composition of
cottonseed oil before it is extracted from the seed. These include climate, growing
region, variety of cotton grown, the agronomic practices employed, and the
handling/storage of the cottonseed before crushing. Due to the interaction of all
these factors in any one sample of oil, it is difficult to make clear generalizations
about quality variations. Factors influencing cottonseed oil quality before it is
extracted from the seed are discussed below.
Climate. Cotton is very productive at high temperatures when adequate water is
available. Free fatty acid levels may rise during hot, humid conditions, but heavy
irrigation above normal levels appears to have no effect on oil hydrolysis to produce
free fatty acids. Abundant rainfall, particularly in May and June, favor high seed-oil
content. Low rainfall tends to increase the proportion of protein, but prolonged
drought results in smaller seeds. Increased maturity, rather than warm weather,
seems to be responsible for higher oil content in late-harvested seeds. Protein is
accumulated in the seeds gradually, but oil is produced just before the boll opens.
Stansbury and co-workers (40) found that high temperatures during seed develop-
ment influenced saturated fatty acid development. They determined that the average
iodine value decrease (reduction in unsaturated fatty acids) per F temperature was
0.76 during boll development and 1.172 during seed development.
Geographical Regions. Differences in climate and soil conditions cause geogra-
phical regions to have a major influence on cottonseed composition, which can
be greater than that caused by the variety of seed grown (41). Tharp (42) reviewed
the variations in oil and protein levels from four regions of the United States and
found that qualitative differences can occur. For example, oil from the Mississippi
Delta region is usually more unsaturated than oil from Texas (43). The iodine value
of the oils tends to increase in seeds grown farther north (42). Meara and Steiner
(44) examined several American, Indian, African, and West Indian cottonseed oils
and concluded that the geographic location had a major effect on the degree of
unsaturation. The iodine values of 25 cotton varieties ranged from about 97 to 112.
Certain varieties appear to be better suited for certain geographical regions. During
the 19781982 Regional Cotton Variety Tests, the variety Acala SJ-5 had higher oil
contents than Stoneville 213, but this difference was larger in the western regions,
particularly in the San Joaquin Valley (41).
Fertilizers. Nitrogen, phosphorus, and potassium in various percentages are the
components of most crop fertilizers. Other elements are vital to plant growth, but
their effects on oil content and composition have not been studied as closely. High
nitrogen levels favor increased protein content in seeds along with decreased lint
and higher number of seeds per boll (42). In a comparison of two levels of nitrogen
application, lower levels produced seeds with higher oil content and viability
(42, 45, 46). Phosphorus has no clear effect on oil and protein levels and potassium
seems to cause a slight increase in oil content. Potassium can increase the oil result-
ing from a given land area by both increasing the seed cotton produced and the oil
percentage, the latter perhaps at the expense of seed protein (47). Interactions
among all three nutrients and any growth regulators can occur and a balance is
important (46).
Seed Handling and Storage. Cottonseed oil from improperly stored cottonseeds
will develop a dark color that requires additional processing. Boatner (48) noted
that the following conditions favor the development of colored or reverted oil:
harvest of immature (bollie) seeds, extremes of moisture or temperature, or other
damage to the cotton plant during cultivation or harvest. Oxidation of gossypol
and other pigments has been proposed as the chemical cause of the color (49).
To avoid this color change, processors avoid intermingling seed known to be
damaged or immature with good-quality seeds. Seeds with higher moisture or higher
than normal free fatty acid (FFA) content are usually processed first. Only seeds
with good quality are stored to be processed throughout the crop year before the
next crop season begins. The moisture, temperature, and FFA level of seeds in
storage are monitored periodically and serve as the basis for further processing
and handling decisions.
Varieties. Glandless cotton varieties tend to contain more oil in the seed than do
glanded varieties (50), and the glandless oil is slightly more unsaturated (51).
Although both types of cottonseed contain approximately equal levels of cyclopro-
penoid fatty acids, the higher gossypol content of glanded seeds causes higher
refined and bleached oil colors compared to glandless. Despite these problems,
glanded seed types are by far the most common. Glanded varieties of cottonseed
normally contain numerous tiny glands (3070 micrometers) which hold the natural
pigment, gossypol, and other gossypol-like compounds. Gossypol, a polyphenolic
pigment (sesquiterpene), is a known antinutrient to livestock and thought to be a
source of color problems to cottonseed oil, has been the focus of both scientific
and practical research (23). However, gossypol is not detectable in RBD cottonseed
oil using the most sensitive analytical techniques currently known. Research efforts
have been devoted to removing pigment glands from cottonseed by processing or
breeding. Screw press, expeller, liquid cyclone (LCP), and air classification pro-
cesses have been developed to reduce the free gossypol content or to remove the
188 COTTONSEED OIL
gossypol pigment glands from cottonseed meal (52). All these processes are able to
produce cottonseed meal with a low enough free-gossypol content (< 450 ppm) for
food applications. However, except for a small amount of screw pressed meal,
neither LCP nor air classified cottonseed meals have been made commercially
available. Glandless cottonseed, a new variety seed without the pigment glands,
was developed with traditional breeding research and introduced in the 1960s
(53). Lusas and Jividen (54) have provided a thorough review of the discovery
and history of glandless cottonseed.
Like other agricultural crops, cotton has been the subject of traditional breeding
and genetic modification programs to develop new varieties with improved charac-
teristics. Cotton is one of the leading crops to be genetically engineered (55) and
since its introduction in 1996, transgenic cotton has been one of the most rapidly
adopted technologies ever (56). In 2002/2003, transgenic cotton varieties comprised
about 20% of the harvested cotton acres and about 27% of the cotton produced in
the world (57). Presently, it is being grown in the United States, China (Mainland),
Australia, South Africa, Argentina, Mexico, India, and Indonesia. In the United
States, in 2002, about 78% of the planted cotton acreage was transgenic cotton.
Genetic engineering is being used to produce transgenic cottons with insect resis-
tance (e.g., Bollgard1; Bt cottons incorporating genes from Bacillus thuringiensis
for boll worm/bud worm resistance), herbicide tolerance [e.g., bromoxynil
(Buctril1; BXN cotton), and glyphosate (Roundup1; Roundup Ready1;
cottons (59)) tolerant cottons] (57, 58). A second wave of transgenic cottons is
imminent, with plants containing two Monsanto Bt genes in a stack debuted
in 2003, followed by glufosinate (Ignite1) herbicide-resistant cottons and plants
with better tolerance to glyphosate about 2006. Other countries, e.g., Egypt and
Pakistan, are evaluating the performance of transgenic cottons (60) and some are
preparing to plant them on a commercial scale (6163).
Research is underway to produce transgenic cottons with other improved agro-
nomic traits as well as improved seed and fiber quality properties (56). Transgenic
technology can provide a means for modifying the lipid profile of cottonseed oil to
improve it nutritionally (e.g., high oleic) and provide the functional properties for
various food and industrial applications (e.g., high stearic) (64). Elimination of gos-
sypol from cottonseed would both enhance the feed value of the meal and could
reduce the processing cost of cottonseed oil (65). Genetic engineering may allow
the reduction or elimination of gossypol only in the seed without affecting its levels
in other mature parts of the plant where it has a beneficial function (66). Efforts
to eliminate or reduce gossypol in cottonseed are in progress through sense and
antisense transgenic approaches (67, 68).
Cottonseed oil, with the minimum amount of processing for a food product, is
refined, bleached, and deodorized (RBD) similarly to all vegetable oils that are used
for human consumption (69). Due to the processing, RBD cottonseed oil should not
contain any detectable aflatoxin, pesticides, or gossypol using current analytical
techniques. In addition, RBD cottonseed oil contains no detectable protein or
DNA, so oil from the transgenic plant should be identical to that from nontrans-
genic plants (55). The current focus of concern about transgenic plants is on
food products (70), which have already impacted the use of cottonseed oil, meal,
and cake in the European Union. However, in 2002, the use of cottonseed oil
derived from plants containing the Bollgard1 gene or Roundup Ready1 gene
were approved by the European Commission for human consumption (71).
3.1. Cottonseed Oil Glyceride Composition

Chemically, all fats and oils are esters of glycerol and fatty acids; nevertheless, the
physical properties of natural fats and oils vary widely. This is because; (1) the pro-
portion of the fatty acids vary over wide ranges and (2) the triglyceride structures
vary for each individual oil and fat. Fats and oils are commonly referred to as tri-
glycerides because the glycerin molecule has three separate points where a fatty
acid can be attached. All triglycerides have identical glycerol components that
leave the fatty acids to contribute the different properties. Three aspects can differ-
entiate the fatty acid components: (1) chain length, (2) the number and position of
the double bonds, and (3) the position of the fatty acids with regard to the glycerol.
Variations in these characteristics cause a large portion of the chemical and physical
differences experienced with edible fats and oils. Fats and oils, for all practical pur-
poses, contain fatty acids with carbon chain lengths between 4 and 24 carbon atoms
with zero to three double bonds. The fatty acids occurring in edible fats and oils are
classified according to their degree of saturation:
Saturated fatty acids, which contain only single carbon-to-carbon bonds, are
chemically the least reactive and have a higher melting point than correspond-
ing fatty acids of the same chain length with one or more double bonds. Most
of the natural saturated fatty acids have an unbranched structure with an even
number of carbon atoms. Fatty acids with carbon chain lengths from 2 to 30
have been reported, but the most important saturated fatty acids are butyric
(C4:0), lauric (C12:0), myristic (C14:0), palmitic (C16:0), stearic (C18:0),
arachidic (C20:0), behenic (C22:0), and lignoceric (C24:0). The melting point
of saturated fatty acids increases with chain length. Saturated fatty acids with
10 and longer carbon chains are solids at room temperature.
Unsaturated fatty acids, which contain one or more carbon-to-carbon double
bonds, are the most chemically reactive and those with the most double bonds
are the most reactive. A fatty acid containing only one double bond is called
monounsaturated; the most notable is oleic (C18:1). When a fatty acid
contains more than one double bond, it is identified as polyunsaturated. The
notable polyunsaturated fatty acids are the essential fatty acids linoleic
(C18:2) and linolenic (C18:3). These fatty acids are essential in the sense
that the human body needs them and yet cannot either synthesize at all or in
sufficient quantities. In nature, the double bonds are cis-form, which has both
hydrogen atoms on the same side of the double bond. Trans-form fatty acids,
with the hydrogen atoms on opposite sides of the double bond, are thermo-
dynamically more stable. Trans-fatty acids are cis-form fatty acids that have
been isomerized by oxidation or hydrogenation (72).
190 COTTONSEED OIL
TABLE 5. Summary of Cottonseed Oil Fatty Acid Compositions.
References
c
Fatty Acid, % a b CA TX d e f
Myristic C14:0 0.8 0.9 0.7 0.9 0.8 0.8 0.7

Palmitic C16:0 27.3 24.7 22.7 25.2 23 22.7 21.6
Palmitoleic C16:1 0.8 0.7 0.6 0.8 0.6 0.8 0.6
Stearic C18:0 2.0 2.3 2.3 2.7 2.3 2.3 2.6
Oleic C18:1 18.3 17.6 17.3 17.5 15.6 17 18.6
Linoleic C18:2 50.5 53.3 55.8 52.6 55.6 51.5 54.4
Linolenic C18:3 0.3 0.3 0.2 0.7
Arachidic C20:0 0.1
a. (73). d. (75).
b. Durkee Industrial Foods, Cleveland, Ohio. e. (76).
c. (74). f. Capital City Products Co., Columbus, Ohio.
Cottonseed Oil Fatty Acid Composition. The specific fatty acid profile of the
triglycerides in cottonseed is dependent on the variety of cotton grown, growing
conditions such as temperature and rainfall, and the analytical method used to
determine the profile. Table 5 summarizes the fatty acid composition observations
of several research and commercial groups. Cottonseed oil is typical of the oleic
linoleic group of vegetable oils, because those two fatty acids comprise almost 75%
of the total fatty acids. Although oleic acid makes up 22% and linoleic makes up
52%, less than 1% linolenic acid is present. Palmitic fatty acid makes up about 24%
of the fatty acids. Minor amounts of other saturated fatty acids are also found.
Bailey (23) noted that the composition of American cottonseed oils will rarely
fall outside of these limits: 2328% total saturated fatty acids, 2228% oleic acid,
and 4453% linoleic acid. The Food and Agriculture Organization of the World
Health Organization (FAO/WHO) has determined a range of fatty acid contents
for commercial fats and oils. The acceptable range of fatty acids prescribed by
Codex in 1997 for cottonseed oil is shown in Table 6 (77).
Cherry et al. (78) examined the variation in cottonseed oil content and composi-
tion as part of a study of genetic and location effects on Texas cottonseed. Both
factors were found to have a significant effect on oil quantity. The oil content of
moisture and lint-free seeds ranged from 23.2% to 25.7% depending on location.
The variation in the six key fatty acids of cottonseed oil was significant within
both cultivars and location, and five of them had significant interactions between
cultivar and location. Linoleic acid varied from 49.07% to 57.64% with a mean
of 54.54%. Palmitic acid varied from 21.63% for an Acala variety grown in
Lubbock to 26.18% for a Lockett variety grown in Corpus Christi, whereas the
mean was 23.68%. The authors also indicated that agronomic practices as well
as weather conditions at the specific location may play a part in the observed
variation. Such variations in oil quantity are not well understood and have not
TABLE 6. Codex Fatty Acid Ranges for Cottonseed Oil.
Fatty Acids, % Typical Range
Lauric C12:0 00.2

Myristic C14:0 0.7 0.61.0
Palmitic C16:0 21.6 21.426.4
Palmitoleic C16:1 0.6 01.2
Stearic C18:0 2.6 2.13.3
Oleic C18:1 18.6 14.721.7
Linoleic C18:2 54.4 46.758.3
Linolenic C18:3 0.7 00.4
Arachidic C20:0 0.3 0.20.5
Gadoleic C20:1 00.1
Eicosadienoic C20:2 00.1
Behenic C22:0 0.2 00.6
Erucic C22:1 00.3
Docosadienoic C22:2 00.1
Lignoceric C24:0 00.1
been carefully studied in cotton, and more work needs to be done. Cherry et al. (79)
have provided a good review of the existing information.
Cottonseed Oil Triglyceride Composition. The triglyceride structure of an edible
fat or oil is affected by which carbon atom of the glycerol has the fatty acid linked,
whether the three fatty acids are the same or different, and the position of each.
Triglycerides with three identical fatty acids are called simple or monoacid trigly-
cerides. Triglycerides containing more than one type of fatty acid are called com-
plex or mixed triglycerides. A mixed triglyceride containing three different fatty
acids has three isomeric forms, depending on which fatty acid is in the middle,
2, or beta position of the glycerol portion of the molecule and which fatty acids
are in the alpha or outside positions. Therefore, both the chemical and physical
properties of fats and oils are largely determined by the fatty acids that they contain
and their position within the triglyceride structure.
As linoleic, oleic, and palmitic fatty acid account for over 90% of the fatty acids
in cottonseed oil, most of the triglycerides contain some combination of these fatty
acids. Table 7 lists the possible combinations of composition and position of satu-
rated, oleic, and linoleic acids in cottonseed oil triglycerides. These ten types of
triglycerides account for 92% of the total triglycerides found (73). The predominant
type is SLL (saturated and linoleic fatty acid in the 1, 2, and 3 positions, respec-
tively), which accounts for over 22 mol% of the triglyceride molecules.
The predominant pair of these ten types includes palmitic acid as the saturated
acid in acyl positions 1 and 3, whereas position 2 is occupied by oleic or linoleic
acid. This is illustrated in position distribution data from Jurriens and Kroesen (73)
that indicates the middle acyl position was occupied by linoleic acid 64.3% of the
time.
The amounts and the types of fatty acids and the interpositional and intraposi-
tional distribution result in various triglyceride forms that contribute to the various
192 COTTONSEED OIL
TABLE 7. Triglyceride Composition of Cottonseed Oil.
Fatty Acid Number of

Pattern Double Bonds Mol %
SOS 1 4.5
SOO 2 4.8
SLS 2 12.4
SOL 3 9.4
SLO 3 8.4
OOL 4 4.1
SLL 4 22.5
OLL 5 6.4
LOL 5 6.5
LLL 6 13.0
Other 8.0
Total 100.0
S saturated; O oleic; L linoleic.
functional properties of cottonseed oil. Bland et al. (80) used more recent separation
techniques to identify the tricacylglycerides and positional isomers of a sample of
cottonseed oil. Their results are presented in Table 8.
Cyclopropenoid Fatty Acids. Cotton, and other plants in the Malvaceae family,
contain a pair of unique cyclopropene fatty acids (CPFA). These two fatty acids,
sterculic and malvalic acid, are generally referred to collectively as cyclopropenoid
fatty acids. Sterculic acid is the most active of the two fatty acids whose general
action is to inhibit the desaturation of stearic to oleic fatty acid in the animal
body with a resultant alteration in membrane permeability or an increase in the
melting point of fats.
TABLE 8. Cottonseed Oil Triglyceride Structure.
Triacylglyceride GLC, % HPLC, %a
PPL 25.7 27.5

LLL 16.1 19.0
POL 14.0 14.0
OLL 12.9 12.5
PPL 8.7 7.1
OOL 4.4 3.1
POO 3.3 3.1
PPO 2.5 2.2
OOO 2.4 1.6
SLL 2.4 1.4
SPL 2.1 1.5
SOL 1.5 1.3
P palmitic; O oleic; S stearic; L linoleic; GLC gas liquid

chromatography; HPLC high-performance liquid chromatography.
a
Corrected using GLC analysis of fatty acid methyl esters.
TABLE 9. Cottonseed Oil Cyclopropenoid Fatty Acid Levels.
Cyclopropenoid Fatty Acid, %

Cottonseed Oil
Source Malvalic Sterculic Total Comment Reference
Crude Oil 0.52 Expeller Extraction (84)

Crude Oil 0.560.90 Hexane Extraction (84)
Crude Oil 0.580.98 Petroleum Ether (85)
Extraction
Crude Oil 0.71.5 0.30.5 (86)
Crude Oil 0.221.44 0.080.56 0.32.0 (81)
Crude Oil 0.64 Glanless Variety (84)
Refined Oil 0.62 (84)
RBD Oil 0.0150.324 0.0050.126 0.020.45 (81)
RBD Oil 0.04 (87)
Winterized Brand Name
Salad Oil 0.040.42 Salad Oils (84)
Winterized Various (88)
Salad Oil 0.10.23 Commercial Brands
Sterculic and malvalic acids are 18 and 17 carbons long, respectively, and
include one double bond at the site of the propene ring, either at the 9, 10 position
or 8, 9 position. The cyclopropene ring is the physiologically active entity of the
two fatty acids (81, 82). The physiological activity of sterculic acid is reported to
be greater than that of malvalic acid (82). The ratio of malvalic acid to sterculic acid
in cottonseed oil is usually about 3 to 1. The cyclopropenyl structure is highly
strained, which apparently accounts for its reactivity (83). Such activity makes it
highly susceptible to the inactivation that regularly occurs during processing. In
cottonseed oil, the CPFAs are reduced in processing with the result that cottonseed
oil in commercial channels contains a negligible level. Examples of levels of
CPFA found in cottonseed oil products are given in Table 9 (81, 84, 85, 87, 88,
89). Deodorization and hydrogenation are the processing steps responsible for
the greatest inactivation of CPFA.
3.2. Cottonseed Oil Nonglyceride Components

The primary constituents in crude vegetable oils are the triglycerides, but they also
contain varying amounts of nonglyceride materials. Cottonseed oil is unusual for
the amount and variety of nonoil substances in the crude oil. Its content of nongly-
ceride substances, exclusive of free fatty acids, commonly amounts to 2% or more
in the crude state. These minor components, identified as the unsaponifiable frac-
tion, consist of phospholipids, tocopherols, sterols, resins, carbohydrates, pesti-
cides, gossypol, and other pigments. Some, but not all, of the nonglyceride
materials are undesirable. Therefore, the objective in all edible oil processing is
to remove the objectionable impurities with the least possible damage to the desir-
able constituents.
194 COTTONSEED OIL
Gossypol. Cottonseed oil is unique among the commercially important fats and
oils because of the presence of a relatively complex system of pigments. Most of
the pigments are of the gossypol type, a biologically active terpenoid substance
present in discreet glands in the seed, leaf, bract, stem, taproot, bark, and root of
the cotton plant. The adaptive function of the compound is believed to be insect
resistance (90). During seed processing, the glands are ruptured, allowing the
gossypol and other similar substances to mix with the protein and oil. Fortunately
for the oil processor, most of the gossypol is bound to the protein. However,
because gossypol and its chemically related compounds are strong pigments, it is
a major objective during caustic refining and bleaching processes to remove as
much of the pigments as possible. Gossypol compounds give crude cottonseed
oil a red color so dark that it usually appears to be black. The characteristic yellow-
ish amber color of refined, bleached, and deodorized cottonseed oil is primarily
caused by the remaining gossypol after processing.
Recognition of gossypols role as a fertility control agent heightened interest in
this polyphenolic compound. In the 1960s, small batches of cold-pressed cotton-
seed oil consumed without further processing caused infertility in Chinese men.
The cooking step, in industrial extraction processes, binds some of the gossypol
to the protein to keep the level below its physiologically active level. Caustic
refining segregates gossypol into the soapstock to levels high enough to impact
its use in animal feeds. After refining, the active earths employed in bleaching
are effective in gossypol removal. Deodorization, by purging the oil with live steam
under vacuum, removes many impurities and odor-causing compounds. Few foods
are as chemically clean as an oil or fat that has been refined, bleached, and deodor-
ized. Analyses have indicated that alkali refining and bleaching reduced the gossy-
pol content of cottonseed oil to less than 1 ppm from 0.05% to 0.42% in solvent
extracted oil and 0.25% to 0.47% for screw-pressed oil (91, 92).
As a practical matter, the presence of gossypol in commercially available cotton-
seed oil conflicts with the refiners need to sell a light, clear product, so there is
a significant incentive to do a good job of removing the gossypol from the oil. A
variety of cotton bred through traditional techniques for its glandless characteristic
became available in the early 1960s. It is generally devoid of gossypol-containing
glands. Glandless cottonseed produces a light colored oil with less-restrictive pro-
cessing conditions (54). However, this variety has never become popular, because
the fiber-producing characteristics of cotton varieties are more important to growers
than the seed genetics, and the market for vegetable proteins has been readily met
by soybean and other products.
Phospholipids. These components are better known to oil processors as phospha-
tides and are frequently referred to, together with small quantities of carbohydrates
and resins, as gums that have adverse effects on product quality and refined oil
yield. Phosphatides are emulsifiers, and so hinder the separation of oil and water
phases in the caustic-refining process. The phosphatides are broadly separated
into hydratable and nonhydratable types. As the name implies, hydratable phospha-
tides can be removed by treatment with water, and the nonhydratable compounds,
which are salts or coordination compounds of calcium and magnesium primarily
with phosphatidic acid, can only be rendered insoluble in the oil by the use of che-
mical reagents, the most commonly used being phosphoric acid.
Phosphatide contents are normally calculated from the determination of total
phosphorus and the use of a factor relating the molecular weight of phosphorus
to the mean molecular weight of the phosphatides in the oil. Typically, the amount
of phospholipids in cottonseed crude oil varies from about 0.7% to 0.9%. The phos-
phatides can also be beneficial: They act as synergists for the tocopherols to inhibit
the autoxidation of vegetable oils (93). This synergistic effect is partly responsible
for the oxidative stability of crude cottonseed oil.
Tocopherols. Various tocopherol isomers that act as naturally occurring antioxi-
dants are found in cottonseed. Natures fat soluble antioxidants can exist in at least
seven forms with alpha-, beta-, delta-, and gamma-tocopherol predominating in
vegetable oils. Alpha-tocopherol contributes Vitamin E activity and some oxidation
resistance, but the gamma and delta forms are the most effective antioxidants.
Typically, crude cottonseed oil contains about 1000-ppm tocopherols, but up to a
third can be lost during processing. The tocopherol content decreases during
each stage of processing, with the highest reductions occurring during chemical
refining and deodorization. Caustic refining can remove as much as 10% to 20%
of the tocopherols, and 30% to 60% of the remaining natural antioxidants can be
lost during deodorization. Typical tocopherol contents for selected vegetable oils
after processing are compared in Table 10 (94, 95).
Sterols. Sterols are crystalline, neutral, unsaponifiable, high-melting alcohols with
multiple-ring structures. Sterols, minor components of all natural fats and oils, are
the major constituents of the unsaponifiable matter remaining in processed vegeta-
ble oils; the remainder consists essentially of saturated and unsaturated hydrocar-
bons. The sterols are colorless, heat stable, and relatively inert, so they do not
contribute any important property to a fat or oil. Chemical refining removes a
portion of the sterols, but more effective separation requires fractional crystalliza-
tion, molecular distillation, or high-temperature steam distillation. Vegetable oil
soapstocks from caustic refining and deodorizer residue are rich sources for
sterol reclamation. An evaluation, comparing caustic-refined cottonseed oil with
TABLE 10. Typical Tocopherol Contents of Selected Vegetable Oils.
Tocopherols, %
Total
Vegetable Oil Tocopherols, ppm Alpha Beta Gamma Delta
Cottonseed Oil 830900 41 trace 58 1

Canola Oil 690695 35 63 2
Corn Oil 8702500 16 2 79 3
Olive Oil 30300 93 7
Palm Oil 360560 30 trace
Peanut Oil 330480 51 1 44 4
Soybean Oil 9001400 10 3 63 24
Sunflower Oil 630700 96 2 2
196 COTTONSEED OIL
TABLE 11. Sterol Content of Several Vegetable Oils (mg/kg oil).
Soybean Oil
Cottonseed Sunflower
Sterol Oila Crude Refined Oila
Cholesterol 0.5 0.5 ndb 0.5

Brassicasterol 0.5 0.5 0.5
Campesterol 276.0 563.0 470.0 242.0
Stigmasterol 17.3 564.0 470.0 236.0
beta-Sitosterol 3348.0 1317.0 1230.0 1961.0
delta-5-Avenesterol 85.1 46.1 10.0 163.0
delta-7-Stigmasterol 0.7 92.0 10.0 298.0
delta-7-Avenasterol 17.9 63.2 5.0 99.4
24-Methylene-cycloartenol 0.5 53.0 204.0
Total 3746.5 2699.3 2195.0 3204.4
a
Crude or refined oil not specified.
b
nd nondetectable.
deodorized cottonseed salad oil, showed a reduction of total sterols from 0.574 to
0.397 mg/100 g of oil (96). Recovered sterols from oil-processing byproducts are
the starting materials for the synthesis of sex hormones and the preparation of syn-
thetic Vitamin D. The sterols of vegetable oils are mixtures called phytosterols. The
sterols in most vegetable oils include beta-sitosterol, campesterol, stigmasterol, and
others in lesser amounts. The most extensively investigated and consumer-recog-
nized sterol is most likely cholesterol. Trace amounts of cholesterol have been
found in some vegetable oils, but animal fats like tallow are routinely found to
contain 1077 mg/kg (97). The sterol content of several vegetable oils are presented
in Table 11 (75, 98).
Pesticides. Pesticides have been used for increased agriculture production through-
out the world. Studies have shown that the majority of the pesticides applied even-
tually reach the soil surface, where they gradually spread, translocate to other
environments, or degrade eventually. Translocation to oil-bearing plant seeds has
also been demonstrated by studies. Processing studies have shown that neither
solvent extraction nor bleaching affected the pesticide levels in the vegetable
oils. However, it was found that pesticides were removed by volatilization during
hydrogenation and/or deodorization (99101). U.S. government agencies have
recognized that the insecticides are distilled from edible oils during the deodoriza-
tion process.
Trace Metals. Vegetable oils contain varying levels of trace metals, depending on
exposure during the growing season as well as during extraction and processing.
Metals can be encountered throughout processing; these reduce the efficiency of
the process and are harmful to product quality and human health. Trace quantities
of copper, iron, manganese, and nickel substantially reduce the oxidative stability
of oils, whereas calcium, sodium, and magnesium reduce the efficiency of refining,
bleaching, and hydrogenation systems. The effects of the metals can be diminished
by the use of chelating agents at various process points to sequester the trace metals
(102). The most widely used chelating agents are citric and phosphoric acids.
3.3. Cottonseed Oil Physical Characteristics

The physical properties for all fats and oils, including cottonseed oil, are deter-
mined by their individual chemical compositions. Physical properties are of prac-
tical importance because most applications depend on the melting behavior,
solubility, flavor, density, appearance, and the other physical properties to provide
functionality for finished products. Analytical and physical evaluation methods are
used to measure these attributes for identification, trading, and control purposes.
Melting Point. The usual definition for melting point is the temperature at which a
material changes from a solid to a liquid. Determination of a fat or oils melting
point is difficult because natural fats and oils do not have true melting points;
the different components melt at different temperatures. Fats and oils are complex
mixtures of triglycerides that pass through a gradual softening before becoming
completely liquid. The melting point procedure is further complicated by the fact
that fat crystals can exist in several polymorphic modifications, depending on the
specific triglycerides involved and the temperature-time pretreatment or tempering
of the sample. For a melting point value, one point within the melting point range
must be selected. Several methods to determine the melting point have been
standardized by AOCS and other organizations, each providing different values:
capillary melting point, softening point, slipping point, Wiley melting point, Met-
tler dropping point, and others. The temperature range at which cottonseed oil
changes from a solid to a liquid is 50 F to 60 F or 10 C to 16 C.
Solid Fat Index. This analysis has become the most important criterion for the
melting behavior and crystalline structure of fats and oils products. It determines
the proportion of solid and liquid materials at a given temperature. The solid fat
index (SFI) analysis is an empirical measure of the solid fat content. It is calculated
from the specific volume at various temperatures using a dilatometric scale gradu-
ated in units of milliliters times 1000. Values for the solid contents are usually
determined at 50 F, 70 F, 80 F, 92 F, and 104 F or 10 C, 21.1 C, 26.7 C,
33.3 C, and 40 C. Unlike the tropical oils, cottonseed and the other oleic- and lino-
leic-classification oils do not contain any significant quantity of triglycerides made
up of two or three saturated fatty acids; therefore, the solid fat index at the lowest
temperature usually measured would have minimal values. Natural cottonseed oil
can have a solid fat index content at 50 F or 10 C but not at the higher temperature
measurements.
Cold Test. The ability of an oil to withstand refrigerator storage is determined by
the cold test analysis; crystallization resistance is measured as the time in hours
before the oil appears cloudy at 32 F or 0 C. Standardized AOCS Method Cc
11-53 requires that dry filtered oil be placed in a sealed 4-ounce bottle and sub-
merged into an ice bath (103). A go/no-go examination after 5.5 hours for clarity
is stipulated by the Official AOCS Method: however, most laboratories practice the
alternative procedure, which continues the clarity examinations until a cloud
198 COTTONSEED OIL
appears. The cold test procedure was developed to evaluate cottonseed oil for the
production of mayonnaise and salad dressings. An oil that solidifies at refrigerator
temperatures will cause an emulsion break with a resultant separation of the oil and
water phases. Currently, the cold test is also used to assure that bottled oils for retail
sales will not develop an unattractive appearance on the grocery shelf.
Cloud Point. An empirical cloud point analysis is performed by stirring a sam-
ple of fat while it is being cooled until the oil has clouded enough to block a light
beam of known intensity. Both cloud point and congeal point values are more clo-
sely related to consistency than melting points. A definite relationship exists
between the cloud point results and the solid fat index values at 92 F or 33.3 C.
Cottonseed oil that has not been winterized or hydrogenated will have a cloud point
of 30 F to 38 F or 1.1 C to 3.3 C. Winterized cottonseed salad oil, with the hard
fraction removed, will have a cloud point of approximately 22 F to 26 F or 5.6 C
to 3.3 C.
Titer. The titer test, AOCS Method Cc 12-59 (103), measures the solidification
point of the fatty acids. First, a fat sample must be saponified and dried before
determining the titer value. Then, a titer tube is filled to the 57-mm mark with dried
fatty acids and suspended in an air bath, which is surrounded by a water bath at
15 C to 20 C below the expected titer result. The sample is stirred until the
temperature begins to rise or remains constant for 30 seconds, after which the stir-
ring is stopped and the endpoint is determined as the maximum temperature that the
fat starts crystallizing or solidifying. Titer analyses are used predominantly in the
soap and fatty acid industries. For edible oils, titer values are commonly specified
for an oil that has been hydrogenated to almost complete saturation. Cottonseed oil
hydrogenated to a 5 iodine value or less should have a titer value of 60 C .
Pour Point. A vegetable oils pour point is the temperature at which the oil just
remains pourable. Actually, this analysis is another melting point determination.
For crude or natural cottonseed oil, the pour point is between 25 F and 32 F.
The pour point temperature increases as the oil is saturated; for hydrogenated cot-
tonseed, the pour point will be higher than for the unhardened oil and can be as high
as 140 F depending on the degree of saturation.
Refractive Index. The refractive index of fats and oils is an important character-
istic because of the ease and speed with which it can be determined precisely, the
small amount of sample required, and its relationship to structure. It is useful for
source oil identification, for observing progress of reactions rapidly, and for estab-
lishing purity. The general relationship between refractive index and the composi-
tion of an oil product with minor exceptions are as follows (104):
Refractive indices increase as the carbon chain length increases.

Refractive indices increase as the number of double bonds increases.
Refractive indices are higher for glycerides than those of fatty acids.
Refractive indices of mixed glycerides are close to corresponding simple
glycerides.
Refractive indices are higher for monoglycerides than for corresponding
triglycerides.
Refractometers equipped with temperature controls are used for fats and oils.
The measurements are usually made at 25 C for soft oils, and the higher melting
point products require temperature adjustments to 40 C or 60 C, depending on the
melting point of the product. The refractive index values decrease as the tempera-
ture is increased but still increases with the length of the carbon chains and the
number of double bonds present in the sample. By reference to a predetermined
curve relating the refractive index at temperature measured to iodine value, a rapid
estimation of the iodine value may be made. One source of error in this method is
that trans-acids formed during hydrogenation affect refractive index values but not
iodine value.
Viscosity. The physical property of a fluid or semifluid that enables it to develop
and maintain a certain amount of shear stress, dependent on the velocity of flow,
and then to offer continued resistance to flow is defined as viscosity. The viscosity
of an oil is temperature dependent; flow increases as the temperature increases. For
edible fats and oils, viscosity decreases as saturation decreases and with shorter
carbon-chain lengths that have lower molecular weights. Most vegetable oils in
the linoleic category have similar viscosities; however, cottonseed oil should be
slightly less viscous by measurement because of a higher saturation level than other
oils in this category. The lauric oils have an easily detected lower viscosity than
other vegetable oils, caused by their high level of short-chain fatty acids. This dif-
ference in viscosity with the use of coconut and palm-kernel oils may be demon-
strated by coverage from coating using these oils verses the longer chain vegetable
oils.
Specific Gravity. The density of a substance compared with water is called specific
gravity. Density or specific gravity is the ratio of the weight of a volume of an oil to
the weight of an equal volume of water at the same temperature. Cottonseed oils
density measurement results are affected by both temperature and fatty acid com-
position. The oils density decreases about 0.000638 units for each C or 0.000355
for each F increase when heated in the 150500 F range. At lower temperatures,
the change is greater, about 0.00069 units per C for cottonseed oil between 0 and
40 C. The specific gravity results of vegetable oils have an inverse relationship with
molecular weight and a direct relationship with the degree of unsaturation. Lund
(105) developed an equation based on saponification and iodine values to predict
the specific gravity of liquid vegetable oils at 15 C: Specific Gravity
0.8475 0.0003 (Saponification Value) 0.00014 (Iodine Value). Therefore, a liter
of cottonseed oil with a specific gravity of 0.917 will weigh 917 g or 917 kg/m3. A U.S.
gallon of cottonseed oil with this specific gravity would weigh 7.66 lbs.
Smoke Point. As oils or fats are heated, a thin bluish smoke appears. The smoke
point is the lowest temperature, under controlled conditions, that the smoke
becomes visible. Cottonseed oils smoke, fire, and flash points, like other fats
and oils, are almost entirely dependent on the free fatty acid content. Fats and
oils smoke point results decrease when the triglycerides are split during hydrolysis
to form free fatty acids and glycerol. The glycerol portion decomposes to form
acrolein, which is the major portion of the smoke evolved from heated fats and
oils. Like other long-chain fatty acid oils, cottonseed oil with 0.01% free fatty
acid will have a smoke point of approximately 450 F. Additions of monoglycerides
200 COTTONSEED OIL
Figure 1. Relationship between free fatty acid content and smoke, flash, and fire points of
cottonseed and peanut oils. , Refined cottonseed oil; , peanut oil.
and diglycerides lower the smoke point because of its free glycerol content. AOCS
Method Cc 9a-48 measures the temperature at which smoke is first detected in a
laboratory apparatus protected from drafts and equipped with special lighting
(103). The temperature at which smoking will be observed in actual use will be
somewhat higher than the test result. Figure 1 shows the relationship among
free fatty acid content, smoke, flash, and fire points for processed cottonseed and
peanut oils.
Flash and Fire Point. Flash point is the temperature at which the volatile products
are evolved at such a rate that they are capable of being ignited but not supporting
combustion. At the fire point, the accumulated breakdown products are capable of
supporting a flame on their own. A crude cottonseed oil with a fatty acid content of
1.8% was found to have a flash point of 560 F or 293.3 C. Solvent-extracted oils
can have a low flash point because of a solvent residue. A flash point analysis would
identify this crude oil deficiency to prevent an accidental fire or explosion in an
atmosphere that was not explosion proof. Crude vegetable oil shipments received
with a flash point below 250 F are rejectable by most trading rules. Figure 1 shows
the relationship between free fatty acid content, smoke, flash, and fire points of
processed cottonseed and peanut oils.
Color. Crude cottonseed oil has a dark reddish-brown color because of the pre-
sence of highly colored material extracted from the seed. After processing, it typi-
cally has a rich golden yellow color that is lighter than peanut and corn oils but
darker than soybean, sunflower, canola, and safflower oils. Some of the color found
in cottonseed oil comes from carotene, but most of the color is caused by a minimal
residual level of gossypol and gossypol derivatives. Although the carotene pigments
are rendered colorless by heat bleaching and some of the pigments may be removed
by adsorption bleaching, gossypol can only be removed by alkali refining. The color
removal level is dependent on the cottonseed handling and storage conditions prior
to extraction. Cottonseed oil darkens when exposed to high temperatures, which
sets the color and makes it impossible to remove, even with caustic refining. Vege-
table oil trading rules recognize the color removal problems with cottonseed oil,
with premium grades specifying bleach color as well as refining loss. The trading
rules have established price adjustments for higher colors and refining losses. The
best crude cottonseed oil grade is prime crude cottonseed oil, which requires that
the oil be capable of refining to not less than 7.6 red on the Lovibond scale. Max-
imum colors for basis prime crude, off crude, and reddish off crude oils are
12, 20, and 30 red, respectively. No maximum red color is specified for low grade
cottonseed oil.
The Wesson method, which is the principal color method for the U.S. edible oil
industry, has been used for many years primarily because of its simplicity. AOCS
Method Cc 13b-45 (103) determines the color of a melted fat or oil product by com-
parison with red and yellow Lovibond glasses of known characteristics. This
method, originally developed in England for measuring the color of beer, is only
intended to assess the degree of redness. Yellow is necessary for assessment of
redness by allowing the colors to closely match with that of oil sample; the amount
of yellow was considered unimportant for this method, and a fixed yellow ratio of
10 yellow to 1 red was adopted for oils with red colors below 3.5 and higher yellow
settings were specified for the darker oils.
Flavor. One of the most important palatability parameters for edible fats and oils
users is flavor. Generally, the flavor of an edible oil product should be completely
bland, so that it can enhance the food products flavor rather than contribute its own.
Cottonseed oil is well known for its initial bland flavor and the nutty flavor it devel-
ops with oxidation. It has been used as the standard for comparison with other oils
for both flavor and odor. The nutty flavor developed with oxidation is more pleasant
than the oxidized flavor of some of the other oils in the oleic linoleic classifications;
for example, soybean oil reverts to a painty, green, watermelon type flavor with oxi-
dation. Another major cause of off-flavors in food oils is hydrolysis. The free fatty
acids liberated with hydrolysis have a distinct flavor and odor that are more dis-
agreeable when the fatty acid chain length is shorter than 14 carbons. Cottonseed
oil that contains mostly C-16 and C-18 fatty acids does not become unpalatable
until the free fatty acids exceed 1.0%.
Consistency. Fats and oils are polymorphic, which means that with cooling, a ser-
ies of increasingly organized crystal changes occur until a final crystal form is
achieved. Each fat and oil has an inherent crystallization tendency, either beta or
beta-prime. The tiny, uniform tightly knit, needle-like, beta-prime crystals produce
smooth-textured shortening, margarine, and specialty solidified oil products with
good plasticity, heat resistance, and good creaming properties. The large, high-
melting, self-occluding, course, stable beta crystals produce grainy, sandy, brittle
solidified oil products that can experience separation of the liquid oil portion. Crys-
tal habit can be controlled by source oil selection as shown on Table 12. Cottonseed
202 COTTONSEED OIL
TABLE 12. Crystal Habit of Hydrogenated Oils and Fats.
Beta Type Beta-Prime Type
Canola Butter Fat

Cocoa Butter Cottonseed
Coconut Menhaden
Corn Palm
Lard Rapeseed (High Erucic)
Olive Rice Bran
Palm Kernel Tallow
Peanut
Safflower
Sesame
Soybean
Sunflower
oil is on the shorter list of hydrogenated oils that crystallize in the beta-prime
crystal form. Almost all of the other domestic oils in the United States crystallize
in the beta form.
3.4. Cottonseed Oil Chemical Characteristics

Cottonseed oil, like all oils and fats, is made up of glyceridic materials with some
nonglyceridic material in lessor quantities. It is the chemical composition that
defines the chemical and physical properties of all fats and oils, which in turn
will determine the suitability of the oil in various processes and applications.
Free Fatty Acid. Oil chemists learned that the free fatty acid (FFA) content was a
good indicator of crude cottonseed oil quality in the 1880s. Hydrolysis causes the
triglyceride molecule to split at the ester linkage to form FFA, diglycerides and
monoglycerides, and eventually free glycerine. This reaction is normally induced
by the presence of moisture and accelerated by heat, but it can also be caused by
certain enzymes. The liberated free fatty acids have a distinct flavor and odor,
which are more disagreeable when the fatty acid chain length is shorter than 14
carbons. Cottonseed oil, which contains mostly C-16 and C-18 fatty acids, does
not become unpalatable until the FFA level exceeds 1.0%.
Crude oil from the best quality cottonseed will have an FFA content of 0.5% to
0.6%. During a good season, the FFA content will be less than 1.0%; however, dur-
ing unfavorable climatic conditions, the FFA content may average 5% or higher.
Dry weather during cotton picking favors low FFA development. The oil in moist
seeds, either in the field or in storage, undergoes rapid hydrolysis. Extremely poor
oil may contain 15% to 25% FFA. As the refining process neutralizes or removes
free fatty acids to a level of 0.05%, this impurity has a direct relationship with refin-
ing loss, i.e., the amount of usable oil enclosed in the soapstock. For chemical refin-
ing, the quantity of sodium hydroxide for neutralization is based on the FFA level of
the crude cottonseed oil. Refining losses as low as 2.5% to 3.0% are encountered
with cottonseed oils containing 0.5% to 0.6% FFA, but oils with high free fatty acid
contents may have refining losses as high as 40% to 50% (23).
Peroxide Value. Oxidation of oils is a major cause of their deterioration. Hydro-
peroxides are the primary products formed by the reaction between oxygen and the
unsaturated fatty acids. Hydroperoxides have no flavor or odor but break down
rapidly to form aldehydes, which have a strong, disagreeable flavor and odor.
The peroxide concentration, usually expressed as peroxide value (PV), is a measure
of oxidation or rancidity in its early stages. PV measures the concentration of
substances, in terms of milliequivalents of peroxide per l000 grams of sample,
that oxidize potassium iodide to iodine. AOCS Method Cd 8-53 (103) is the official
method for peroxide value determinations.
Peroxide value is one of the most widely used chemical tests for the determina-
tion of fats and oils quality. It has shown good correlation with organoleptic flavor
scores. However, a peroxide value determination does not provide a full and unqua-
lified evaluation of oil quality because of the transitory nature of peroxides and their
breakdown to nonperoxide materials. Although a linear relationship has been
observed between peroxide values and organoleptic flavor scores during the initial
stages of lipid oxidation, this method alone is not a good flavor quality indicator
because peroxide increases to a maximum and then decreases as time increases.
Therefore, a high peroxide value indicates oxidation to produce a poor flavor, but
a low peroxide value is not always an indication of a good flavor.
Anisidine Value. Anisidine value is a measure of secondary oxidation or the past
history of an oil. It is useful in determining the quality of crude oils and the efficiency
of processing procedures, but it is not suitable for the detection of oil oxidation or
the evaluation of an oil that has been hydrogenated. AOCS Method Cd 18-90 has
been standardized for anisidine value analysis (103). The analysis is based on the
color reaction of anisidine and unsaturated aldehydes. An anisidine value of less
than ten has been recommended for oils upon receipt and after processing (94).
Inherent Oxidative Stability. The unsaturated fatty acids in all fats and oils are
subject to oxidation, a chemical reaction that occurs with exposure to air. The even-
tual result is the development of an objectionable flavor and odor. The double bonds
contained in the unsaturated fatty acids are the sites of this chemical activity. An
oils oxidation rate is roughly proportional to the degree of unsaturation; for exam-
ple, linolenic fatty acid (C18:3), with three double bonds, is more susceptible to
oxidation than linoleic (C18:2), with only two double bonds, but it is ten times
as susceptible as oleic (C18:1), with only one double bond. The relative reaction
rates with oxygen for the three most prevelent unsaturated fatty acids in edible
oils are:
Reaction Rate
Fatty Acid With Oxygen
Oleic C18:1 1
Linoleic C18:2 10
Linolenic C18:3 25
204 COTTONSEED OIL
Oxidation deterioration results in the formation of hydroperoxides, which

decompose into carbonyls, dimerization products, and polymerized gums. It is
accelerated by temperature, oxygen pressure, prior oxidation, metal ions, lipoxi-
dases, hematin compounds, antioxidant reductions, absence of metal deactivators,
time, and ultraviolet or visible light. Extensive oxidation will eventually destroy the
beneficial components contained in many fats and oils, such as the carotenoids or
vitamin A, the essential fatty acids (linoleic and linolenic), and the tocopherols or
Vitamin E. Fats and oil oxidative reactions are directly related to the fatty acid com-
position, or more specifically, to the unsaturation type and amount. Oxidative sta-
bility estimates can be made from the iodine value measurement, calculated iodine
value from the fatty acid composition, or an oxidative stability formula. The oxida-
tive stability estimates, determined for the major natural vegetable oils, are pre-
sented in Table 13 (106). Inherent oxidative stability rating is the combined
reaction rates of an oils unsaturated fatty acids. It is determined by multiplying
the decimal fraction of each unsaturated fatty acid by its relative oxidation rate
and then summing the results (107). As indicated in Table 13, deodorized cotton-
seed oil, with approximately 26% saturates, resists oxidation and subsequent rever-
sion better than other less saturated oils containing higher quantities of linolenic
fatty acids (C18:3).
AOM Stability. The active oxygen method (AOM) is the most commonly used
analytical method for measuring oxidative stability of fats and oils products.
AOM employs heat and aeration to accelerate oxidation of the oil by continuously
bubbling air through a heated sample. Periodic peroxide value analyses are per-
formed to determine the time required for the oil to oxidize under the AOM con-
ditions. This method requires close attention to detail to produce reproducible
results, and even then, the variation between laboratories is 25 for a l00-hour
AOM sample.
TABLE 13. Vegetable Oils Oxidative Stability
Inherent Vegetable Calculated Total

Oxidative Oil Iodine Double
Ratinga Stabilityb Source Value Bonds
Best 0.3 Coconut 9.6 11

0.4 Palm Kernel 17.2 20
1.5 Palm 52.2 61
1.6 Olive 82.4 96
3.7 Peanut 97.0 113
5.3 Canola 116.5 135
5.8 Cottonseed 112.4 130
6.5 Corn 128.4 148
7.2 Sunflower 136.2 157
7.5 Soybean 132.5 153
Worst 8.0 Safflower 146.1 169
a
A perfect inherent oxidative stability value is zero.
b
Calculated Inherent Oxidative Stability Sum of the decimal fraction of each
unsaturated fatty acid times its relative oxidation rate.
Oil Stability Index. Two conductivity instruments, Rancimat and The Oxidative
Stability Instrument, have been developed as alternatives to AOM Stability analy-
sis. These instruments measure the increase in deionized water conductivity result-
ing from trapped volatile oxidation products produced when the oil product is
heated under a stream of air. The conductivity increase is related to the oxidative
stability of the products. These instruments provide a more reproducible measure-
ment of oxidation stability with less technician time and attention.
Iodine Value. Iodine value (IV) is a simple and rapidly determined chemical con-
stant that measures the unsaturation of an oil, but it does not define the specific fatty
acids. The iodine value procedure determines the grams of iodine absorbed by 100 g
of oil. A higher iodine value indicates a greater number of double bonds. The
iodine value results for cottonseed oil vary somewhat from year to year, sections
of the country, and by growing season. A cooler growing season provides oil
with a higher than average linoleic fatty acid (C18:2) content with a lower oleic
fatty acid (C-18:1) content; warmer growing seasons reverse this trend. These var-
iations increase or decrease the number of double bonds, which affects the iodine
value. Typically, cottonseed oil iodine values range from 103 in Texas to 112 in
other regions of the United States.
Halphen Reaction. The halphen test is a very sensitive and reliable method for
detecting the presence of cottonseed oil in another oil. A reaction with sulfur in
carbon disulfide mixed with equal amounts of amyl alcohol gives a cherry red color
when cyclopropenoid fatty acids unique to the Malvacae family, which includes
cottonseed and okra, are present. This test is capable of detecting 0.25% or less
cottonseed oil in an oil blend. The oil is no longer responsive to the halphen test
after hydrogenation, which decreases the iodine value 25 units.
Unsaponifiable Matter. Unsaponifiable matter are those substances dissolved in
an oil that cannot be saponified by alkalies but are soluble in nonpolar solvents.
These materials are made up of sterols, hydrocarbons, tocopherols, pigments, and
higher materials that are insoluble in water. The level of unsaponifiable matter in
good-quality cottonseed oil usually ranges from 0.5% to 0.7%. It may decrease
slightly in deodorized oils due to slight reductions of sterols with alkali refining
and high-temperature deodorization.
Saponification Value. Saponification value is useful in predicting the type of gly-
cerides in an oil by measuring the alkali-reactive groups. It is a measure of the aver-
age molecular weight of the glycerides in the oil. Glycerides containing short-chain
fatty acids have higher saponification values than those with longer chain fatty
acids. Cottonseed oil saponification values range from 189 to 198 with an average
of 195. Independent of any other analytical measurement, the saponification value
results overlap too much to identify individual fats or oils; most oleic and linoleic-
classification oils have saponification values in the 180 to 200 range. In edible oil
processing, saponification value analyses have been replaced almost entirely by
fatty acid composition analysis via gas-liquid chromatography.
Fatty Acid Composition. The classical method for determining the fatty acid
composition of an oil used a combination of its iodine value, relative density,
refractive index, and saponification value. This method has been replaced with
206 COTTONSEED OIL
TABLE 14. Fatty Acid Compositions of Oleic/Linoleic Classification Vegetable Oils.
Fatty Acid Olive Canola Peanut Corn Cottonseed Soybean Sunflower
Myristic C14:0 0.1 0.1 0.1 0.8 0.1 0.1

Palmitic C16:0 9.0 4.0 11.1 10.9 23.2 10.6 6.3
Palmitoleic C16:1 0.6 0.3 0.2 0.2 0.7 0.1 0.1
Stearic C18:0 2.7 1.8 2.4 2.0 2.1 4.0 5.1
Oleic C18:1 80.3 58.8 46.7 25.4 16.9 24.0 17.8
Linoleic C18:2 6.3 21.4 32.0 59.6 55.8 52.9 69.1
Linolenic C18:3 0.7 10.3 1.2 0.2 7.7 0.4
Arachidic C20:0 0.4 0.6 1.3 0.4 0.3 0.3 0.4
Gadoleic C20:1 1.5 1.6 0.1
Behenic C22:0 0.3 2.9 0.1 0.3 0.6
Erucic C22:1 0.7
Lignoceric C24:0 0.2 1.5
fatty acid composition analysis determined by gas-liquid chromatographic (GLC)

patterns. The GLC procedure identifies the percentage of each individual fatty
acid with one analytical procedure, which applies equally well to refined and
unrefined oils. The fatty acid composition analysis by GLC provides a rapid
and accurate means of determining the fatty acid distribution of fat and oil pro-
ducts. This information is beneficial for all aspects of product development, pro-
cess control, and marketing because the physical, chemical, and nutritional
characteristics of fats and oils products are determined by the kinds and propor-
tions of the component fatty acids and their position on the glycerol moiety. The
physical characteristics of an oil depend on the degree of unsaturation, the carbon
length, the isomeric fatty acid forms, and the molecular configuration. Each fatty
acid has an individual melting point that increases with chain length and
decreases as the fatty acids become more unsaturated. The unsaturated fatty acids
are chemically more active than the saturates because of the double bonds, and
this reactivity increases as the number of double bonds increase. The double
bonds are subject to oxidation, polymerization, hydrogenation, and isomerization.
Usually, fats and oils products are liquid at room temperature when the unsatu-
rates are high and solid when the level of unsaturates are low. However, this gen-
eralization can be complicated by the trans-isomers formed during hydrogenation
(108). Table 14 compares the typical fatty acid composition of the oleic-linoleic
classification oils.
3.5. Cottonseed Oil Analytical Characteristics

All of the edible fats and oils vary considerably in their chemical structure, which
determines the physical characteristics that provide functionality. The physical,
chemical, and performance analyses are the tools available to the fats and oils pro-
cessor for the evaluation of the products produced, development of new products,
purchase of raw materials, and identification of specific customer requirements.
COTTONSEED HANDLING, OIL EXTRACTION AND PROCESSING 207
TABLE 15. Typical RBD Cottonseed Oil Analytical Characteristics.
Characteristic Typical Range

Specific Gravity at 25/25 C 0.916 to 0.918
Refractive Index at 25 C 1.468 to 1.472
Iodine Value 108.0 98.0 to 118.0
Saponification Number 189 to 198
Unsaponifiable Number < 1.5
Titer, C 34.9 30.0 to 37.0
Melting Point, C 13.0 10 to 16
Solidification Point, C 12.0 to 13.0
Cloud Point, C 3.0 1.0 to 3.0
Cold Test, hours none
AOM stability, hours 16.0 16 to 19
Tocopherol Content, ppm
alpha-tocopherol 355 340 to 369
gamma-tocopherol 502 481 to 522
delta-tocopherol 8 8 to 9
Fatty Acid Composition, %
Myristic C14:0 0.7 0.6 to 1.0
Palmitic C16:0 21.6 21.4 to 26.4
Palmitoleic C16:1 0.6 0 to 1.2
Stearic C18:0 2.6 2.1 to 3.3
Oleic C18:1 18.6 14.7 to 21.7
Linoleic C18:2 54.4 46.7 to 58.2
Linolenic C18:3 0.7 0 to 1.0
Arachidic C20:0 0.3 0.2 to 0.5
Gadoleic C20:1 0 to 0.1
Behenic C22:0 0.2 0 to 0.6
Erucic C22:1 0 to 0.3
Lignoceric C24:0 0 to 0.1
Triglyceride Composition, %
Trisaturated (GS3) 0.1 0 to 0.1
Disaturated (GS2U) 13.2 14.0
Monosaturated (GSU2) 58.4 50.0 to 58.0
Triunsaturated (GU3) 28.3 28.0 to 36.0
Hydrogenated Crystal Habit beta prime
G glycerides; S saturated; U unsaturated.
Table 15 lists the typical cottonseed oil analytical characteristics, including fatty
acid and triglyceride compositions and ranges (43, 80), which allow for the differ-
ent varieties, growing conditions, and analytical error.
4. COTTONSEED HANDLING, OIL EXTRACTION

AND PROCESSING
Nature has provided plants with systems to synthesize, utilize, and store food lipids.
Improper handling and storage of oilseed prior to extraction can have deteriorous
effects on the oil quality. Therefore, control of cottonseed transportation, storage,
208 COTTONSEED OIL
segregation of lots, and moisture are the first processes for processing a desirable
edible oil. Oil extraction has a long development history, while most of the oil pro-
cessing methods were largely introduced during the twentieth century. Until the
recent past, crude oil extraction and oil processing were two separate industries.
However, during the last quarter century, sheer economics and product synergy
have caused both horizontal (merger of similar operations) and vertical integration
(combination of different but related operations) of these activities to occur. Now,
companies continue to increase their crushing capacity and many extract and refine
the oil as a continuous operation. Most of these operations have integrated miscella
refining with sodium hydroxide to produce a prime-bleachable-summer-yellow
(PBSY) cottonseed oil with consistently lighter color. PBSY, the caustic-refined
cottonseed oil, is a trading definition of the National Cottonseed Products
Association with an AOCS official bleach color of no greater than 2.5, no more
than 0.25% free fatty acid, and no more than 0.10% moisture and volatile matter.
The oil processor is guaranteed quality via this trading rule with its 2.5 laboratory
bleach color.
4.1. Cottonseed Handling and Storage

Once a cotton boll opens, the cottonseeds within are susceptible to deterioration.
The living seeds respire, producing carbon dioxide and heat. Other biological pro-
cesses occur in the seeds as well. Triglycerides are split by enzymes via hydrolysis
to release free fatty acids, which the embryo plant uses for energy. The production
of free fatty acids is undesirable for the oil processor because seeds high in FFA
contain low-quality oil, which usually results in a higher refining loss. The rate
of hydrolysis is dependent on temperature and moisture, but the hydrolytic enzymes
may be inactivated by heat. Oxidation of the fatty acids in the oil can result from
heating (109). Altschul (110) comprehensively reviewed the biological processes
that take place in the cottonseed before and after harvest, and more recent reviewers
(111) have considered cottonseed development.
Another factor in cottonseed storage is the control of mold development on the
seeds. Cottonseed is particularly susceptible to the fungus Aspergillus flavus, which
produces aflatoxin. Aflatoxin formation in cottonseed is generally, but not exclu-
sively, initiated in the field, rather than during storage. The rigors of normal
processing, along with the fact that the toxin is associated with the protein segment
of the seed, preclude aflatoxin from the oil, but the residual meal from affected seed
is still contaminated. The use of contaminated meal for animal feed is strictly
controlled by both state and federal regulation. For health and economic reasons,
it is important to minimize conditions that favor the growth of A. flavus and to
monitor for its presence. An excellent review of the subject of aflatoxin control
in cottonseed has been provided by Park et al. (112).
Ideally, cottonseed should be stored at a moisture content of less than 10% (49).
Dehulled seeds should contain no more than 9% moisture and 1% FFA (109). Prior
to storage, cottonseeds must be sampled for moisture analysis. The American Oil
Chemists Society (AOCS) method Aa 3-38 specifically describes a procedure for
determining moisture in cottonseeds, although other methods may be used. Seeds

with 1011% moisture may be stored immediately, but seeds with higher moisture
require additional drying. Drying may take place at either ambient temperatures
or up to 104.5 C (220 F), to 12% or less. There is no benefit to drying the
seeds below 9% moisture. Although dryers may not be needed every season,
they are a necessity when wet seeds are delivered to the mill (113). As the seeds
are hygroscopic, the relative humidity of the air in storage must be monitored to
maintain the desired seed moisture level. Treatment of high-moisture seeds with
propionic acid to retard deterioration has been investigated, although it is not
widely practiced (114).
The most common type of cottonseed storage facility is the seed house, however,
some silos are used; and west of the Mississippi, seed is also stored outside on solid
slabs and covered with canvas. An air-cooling system is vital to the successful
storage of cottonseed. The temperature of the seeds is dependent on the ambient
temperature and degree of ventilation in the storage area (115). As the seeds are
respiring, heat can build up, particularly if the seeds have a high moisture or
FFA level (49). Overheated seeds must be cooled to below 60 F (15.6 C) to prevent
further deterioration (116). Storage for over one year is possible provided that the
seeds are held under adequate conditions.
4.2. Cottonseed Oil Extraction

Cottonseed was one of the earliest oils to be extracted from the seed. The extraction
of cottonseed oil slowly progressed from edgestone to wedge press to hydraulic
press (23). Hydraulic pressing was the predominant means used to separate the
oil from cottonseed for most of the nineteenth century. As improved cotton spin-
ning, weaving, and ginning operations during the eighteenth century made more
cottonseed available for crushing, the labor-intensive hydraulic-press operation
quickly yielded to the continuous screw press in the early 1900s. As edible oil
commanded a respectable market price and because either the hydraulic press
or the screw press still left nearly 20% of the available oil of the seed in the
press-cake, much research activity was initiated to find an acceptable solvent to
extract the remaining oil from the cake. This effort led to the prepress solvent
extraction process in the 1930s. This combination of mechanical press and solvent
extraction of the press-cake was able to recover better than 97% of the available oil
in the cottonseed. Further demand in productivity and oil quality brought expander-
and miscella-refining of cottonseed oil into the crushing business in the 1970s.
Toward the late 1980s, more than 80% of the cottonseed crushed in the United States
was accomplished by the expander solvent extraction operation (3). The flow
sequence for cottonseed oil extraction most practiced in the United States is
illustrated in Figure 2 (69).
Cottonseed Preparation. Most oilseeds require some degree of cleaning and pre-
paration before the oil is separated from the solid portion of the seed. After the cot-
ton fiber is removed from the seed by the ginning operation, the seed still has short
linters with a white appearance. This is called white or fuzzy cottonseed in the
210 COTTONSEED OIL
Figure 2. Cottonseed oil extraction flow sequence.
trade. On a dry basis, white cottonseed is composed of 12.7% linter, 31.8% hull,
and 55.5% kernel (23). As illustrated in Figure 2, cottonseeds are usually stored
uncleaned when received at the oil mill. When the cottonseeds are removed from
storage for extraction, dirt and other trash must be removed. Several seed-cleaning
systems are used, which are all based on some type of screening. Trash that is
lighter or smaller than the seeds will be aspirated in pneumatic systems or sifted out
mechanically. The larger pieces of trash are screened out and magnets are used to
remove ferrous metal. Foreign materials that have the same size and density as the
seed can still be carried on through the process stream (117).
Delinting. This step is unique to cottonseed among all the oilseeds. The short
cellulose linter fibers must be removed from the seeds because leaving them on
the seed would lower the yield of oil due to absorption of the oil by the cellulose
fibers. Linters, are bulky and tend to hold the neutral oil or occupy valuable extrac-
tor space during the extraction. Chemicals, such as sulfuric acid, have been used to
remove the linters, especially when seeds are prepared for replanting. However, all
commercial mills remove the linters mechanically as these short fibers have many
nonfood applications, serving as the starting material for pure cellulose, plastics,
and rayons, and are used in high-quality paper, batting or padding in bedding and
furniture, and automotive uses.
Hulling. Once the lint is removed, the hulls are separated from the seeds. Hulls that
are allowed to remain with the kernels absorb oil during extraction and lower the
quality of the meal produced by lowering the protein level. The hulls cannot be
completely eliminated without a loss of kernel, so an acceptable level of hull reten-
tion must be determined, depending on the desired protein level of the final meal.
Two types of hullers are used in the industry. The bar huller consists of a bar- or
knife-studded cylinder that rotates within another cylinder having similar knives
protruding from its interior. The hulls are cut as the seeds pass around the inner
cylinder. The seed decorticator has two hardened steel rolls, both of which have
longitudinal grooves cut in the surface. The seeds are fed between the rolls and
then cut by the grooves and the difference in speed between the two rolls. The hulls
and uncut seeds are removed from the kernels by screening and the hulls are aspi-
rated so that the seeds may be returned to the huller.
Flaking. After hulling, the meats, or kernel, are reduced in size or flaked to facil-
itate oil removal. This rolling process minimizes the distance through which the
free oil must pass, but it does not necessarily rupture the walls of the oil cells. Pro-
per moisture content of the seeds is essential for flaking, and if the moisture level is
too low, the seeds must be conditioned to raise the moisture to about 11% (116).
Cottonseeds may be flaked by passing between two rolls mounted side-by-side;
however, they are more often flaked in a series of five stacked-crushing rolls
because a thinner flake may be achieved with the vertical rolls. The ultimate thick-
ness of the flake is determined by the method of extraction used. For mechanical
pressing, a thickness of 0.1270.254 mm (0.0050.010 inch) is common, and for
solvent extraction, flakes of not less than 0.2300.254 mm (0.0080.010 inch)
are common (49). Thinner flakes tend to disintegrate during the solvent process.
Cooking. Prior to extraction, the flakes are heated or cooked. Ward (118) summar-
ized the purpose of cooking the flakes as follows: (1) cell walls are broken down
allowing the oil to escape; (2) oil viscosity is reduced; (3) moisture content is con-
trolled; (4) protein is coagulated; (5) enzymes are inactivated and micro-organisms
are killed; (6) gossypol is bound to protein, to some extent, by the action of heat
in combination with moisture and physical treatment and thus some portion of it
212 COTTONSEED OIL
is detoxified; and (7) certain phosphatides are fixed in the cake, which helps to
maintain subsequent refining losses.
Cottonseed flakes are usually cooked in stack cookers that are 48 kettles high.
The sides and floors of each kettle are steam-jacketed to heat the flakes. The flakes
are fed into the top kettle, heated for a specific time, and then swept into the kettle
below. The temperature of lower kettles are usually maintained at higher tempera-
tures than the top kettle. If the flakes are relatively dry, moisture may be added to
the top kettle to reach a level of 1112%. As the flakes progress toward the bottom
kettle, water is evaporated and removed by vents in each of the lower kettles until
the final moisture level is reached. The desirable level is 56% moisture for seeds to
be hydraulically pressed and about 3% for seeds to be expeller or screw pressed
(49). Cooking seeds at low temperatures and moisture content may result in less
protein binding of gossypol. Higher gossypol levels will then develop in the oil
and affect the color of the crude oil (49).
The flakes are heated to over 190 F (87.8 C) in the upper kettle. Flakes with
high phosphatide content may benefit from being cooked at slightly lower tempera-
tures to avoid elevating refining losses. The temperature of the flakes is raised to
230270 F (110132.2 C) in the lower kettles. The seeds are cooked for up to
120 minutes and, depending on the size of the cooker, 81136 metric tons (90
150 short tons) of meats may be cooked in a 24-hour period.
Overcooking lowers the nutritional quality of the meal and darkens both the oil
and the meal. Poor-quality seeds with high levels of free fatty acids cannot be
cooked for as long a period as high-quality seeds because of darkening. Darker
oil requires additional refining to achieve the desired bleach color.
Oil Extraction. Four types of processing systems are used to extract oil from oil-
bearing materials: (1) hydraulic press, (2) expeller or screw press, (3) prepress
solvent extraction, and (4) direct solvent extraction. These systems employ the
two techniques in common practice for the extraction of cottonseed oil. These
are mechanical by means of a press or the solvent process with the use of hexane.
Mechanical pressing is normally applied to oilseeds that are relatively high in
extractable oil. Hull-free cottonseed kernels contain as much as 34% oil and are
suitable for the mechanical extraction process (3). The prepress solvent system
employs a combination of the two techniques, where seeds are lightly screw-
pressed to reduce the oil by one-half to two-thirds of its original level before solvent
extraction completes the job. After 1980, more than 80% of the cottonseed crushed
in the United States was accomplished by the expander solvent extraction proce-
dure. In most cases, solvent extraction and refining processes are coupled together
for quality and efficiencies as reviewed in Miscella Refining of Section 4.3. On a
worldwide basis, due to available transportation infrastructure, hardware, solvent,
and skilled labor, cottonseed is still being processed with all four extraction
systems.
Hydraulic Pressing. Batch pressing was the earliest commercial method of oil
extraction. Hydraulic equipment replaced the mechanical operations and the meth-
od became known as hydraulic pressing (49). In open presses, oilseed meals
were wrapped in cloths and placed between plates, which were then gradually
TABLE 16. Comparative Yield of Cottonseed Oil

from Different Extraction Processes.
Kg of Oil
per MetricTon
Method of Cottonseed
Hydraulic Press 154

Continuous Screw Press 163.5
Solvent Extraction Including
Prepress Solvent Extraction 180.5188
compressed to squeeze the oil from the seeds. Box-type presses were most often
used for cottonseed, and this method was fairly labor intensive. Today, worldwide,
very little cottonseed is hydraulically pressed and none in the United States. The
relative efficiency of this type of extraction to more modern methods is shown in
Table 16 (49). Wrenn (5) presented the historical context of hydraulic pressing of
cottonseed oil and included pictures of the equipment in operation.
Screw Pressing. The screw-press process has been used since the early 1900s in
place of the hydraulic press to separate oil from cottonseed meats. With this system,
pressure is gradually applied to the flakes as a screw conveys them from the feed
end to the discharge end of the expeller barrel. A plug of the compressed meal
develops at the discharge end and a drainage barrel surrounds the press to collect
the oil expressed during the passage of the flakes. About 34% oil remains in the
cake that results from screw pressing. Anderson (Cleveland, Ohio) expellers have
both vertical and horizontal presses to maximize pressure, and French (Piqua, Ohio)
screw-presses generally consist of a horizontal, water-cooled cage. Both types of
presses exert 6801089 atm (58 tons per square inch) pressure on the flakes
(49). A disadvantage of the screw press method compared with the now outdated
hydraulic pressing was the tendency of screw-pressed cottonseed oil to have higher
color due to the lower moisture content of the cooked meats prior to pressing (49).
Direct Solvent. This process is based on the use of a nonpolar solvent, specifically
hexane, to dissolve the oil without removing proteins and other compounds. The
flakes are mixed with hexane in a batch or continuous operation. The resulting
oil-solvent micelle and the residual meal are heated to evaporate the solvent, which
is collected and reused. Solvent extraction yields about 11.5% more oil than does
the screw-press method (49), and 1% or less oil remains in the meal. However,
direct solvent extraction is problematic for the cottonseed industry because the
high oil content of cottonseed flakes causes them to break into fires during extrac-
tion after the oil is removed; occasional overheating of the oil-solvent miscella will
cause irreversible color changes in the oil; the meal requires additional heating to
bind gossypol, if it is destined for use as poultry or swine feed; the solvent poses fire
problems and is expensive; hexane poses environmental pressures as a volatile
organic compound (VOC); and the main component of hexane, n-hexane, is also
classified as a Hazardous Air Pollutant (HAP) by the U.S. Environmental Protection
Agency and is strictly regulated in the United States (55, 119).
214 COTTONSEED OIL
Prepress Solvent Extraction. Solvent extraction is relatively costly and is not well
suited for the high-oil content of cottonseed. Mechanical pressing leaves about 5%
oil in the press-cake, and it is desirable to recover as much oil as possible. A logical
processing step was to combine the two extraction techniques. With prepress sol-
vent extraction, cottonseeds are pressed to remove most of the oil and then the oil
remaining in the press-cake is extracted with solvent. This solvent extraction oper-
ates on a reduced volume of feed stock (i.e., press-cake, as opposed to full-fat
flakes) and, therefore, requires a modest size extractor with modest amounts of
desolventizer and solvent.
Expander-Solvent Extraction. The most recent development in cottonseed
extraction is the use of expanders. The expander is a low-shear extruder that heats,
homogenizes, and shapes oilseeds into porous collets or pellets with a high-bulk
density. Steam is injected into the oilseed flakes or cake in the expander while under
pressure, and then this mixture is extruded through plates to the atmosphere. The
collets expand when released to the atmosphere, hence the name expander. Some
expanders have a drainage cage to reduce the oil content of high-oil seeds to less
than 30%, thus enabling the production of intact collets for direct solvent extrac-
tion, instead of the prepress extraction process (120). Watkins (121) reported that
after hexane extraction, the collets had 2545% less solvent holdup and 15 times
less oil than did traditionally prepared flakes. Solvent capacity is enhanced because
the extruded cottonseed requires solvent-feed ratios of 1:1, compared with 1.8:1 for
direct extraction of cottonseed meats (122). After extraction, the meal contains
0.100.20% free gossypol, which is about half the amount found in flakes. As of
2000, essentially all commercial cottonseed extraction in the United States employs
the expander solvent extraction process.
4.3. Cottonseed Oil Processing

Processing flow sequences for six different product groups are illustrated in
Figure 3. One finished product that is still used in some parts of the world was
omitted from Figure 3: extracted and filtered cottonseed oil. Cottonseed oils that
did not have any processing after extraction were purchased at local bazzars in
Central Asia in 1998. These oils would be unacceptable in the west, where the con-
sumers have been conditioned to prefer edible oils that are light in color, bland
flavored, have a high smoke point, maintain a clear appearance both on the grocery
shelf and under refrigeration, contain additives to prolong flavor and frying stabi-
lity, are modified to provide a specific performance characteristic, and attractively
packaged for convenient handling. The processes responsible for these and other oil
product qualities are presented in this section.
Refining. As used here, the term refining refers to any purification treatment
designed to remove FFA, phosphatides, gossypol, and other gross impurities in cot-
tonseed oil; it excludes any other process, such as bleaching or deodorization. The
refining process probably has more impact on a vegetable oils quality and econom-
ic performance than any of the other processes during the conversion to a finished
product. Inadequately refined oils will affect the operation of all succeeding pro-
215
Figure 3. Cottonseed oil processingflow sequence.

216 COTTONSEED OIL
cesses and the quality of the finished product. Additional processing and handling
required because of poorly refined oils will increase the costs of a suspect quality
finished product beyond that of one produced with a properly refined oil that has a
good quality.
Two different refining systems are currently used to refine vegetable oils; chemi-
cal and physical refining. Some shortcomings experienced with physical refining
have maintained alkali refining as the preferred vegetable oil purification technique
in the United States. Some oils, like cottonseed, contain nonglyceride materials that
cannot be removed adequately by the processes employed with physical refining.
Gossypol and related pigments in cottonseed oil readily combine with caustic
soda and, thus, are removed most effectively by alkali-refining. Several different
versions of the alkali-refining process are practiced in the United States and other
countries: long-mix, short-mix, miscella, and the Zenith process. This discussion
will be limited to the conventional, or long-mix, caustic-refining process, which
evolved to provide extended reaction times with caustic soda for the effective
removal of gossypol and phosphatides from cottonseed oil and the miscella
process.
Conventional or Long-Mix Caustic-Soda Refining. The conventional caustic-
soda refining process is the most widely used and best known refining system.
The addition of an alkali solution to a crude oil brings about a number of chemical
and physical reactions. The alkali combines with the free fatty acid present to form
soaps. The phosphatides and gums absorb alkali and are coagulated through hydra-
tion or degradation. Much of the coloring material is degraded, absorbed by the
gums, or made water soluble by the alkali, and the insoluble matter is entrained
with the other coagulable material. With heat and time, the excess caustic can
also bring about the saponification of the neutral oil. Therefore, selection of the
NaOH strength, mixing time, mixing energy, temperature, and the quantity of
excess caustic all have an important part in making the alkali-refining process oper-
ate effectively and efficiently.
The current alkali-refining techniques are a result of the gradual application of
science to the basic art of batch refining originally performed in open-top, cone-
shaped kettles. Efficient separation of soapstock from the neutralized oil is the sig-
nificant factor in alkali refining, and the technique of using centrifugal separators
materially improved the yield from 1.5% to 2.5%. The caustic soda continuous sys-
tem that evolved has the flexibility to efficiently refine all crude edible oils presently
used in the world. The system may be outlined as follows (106).
Crude Oil ConditioningCrude oils with high levels of phosphatides and

trace metals are usually treated with food-grade phosphoric acid for 4 to 8
hours before refining. The purpose of the acid pretreatment is to: (1) help
precipitate phosphatidic materials; (2) precipitate natural calcium and
magnesium as insoluble phosphate salts; (3) inactivate trace metals, such as
iron, copper, and others that may be present in the oil; (4) reduce the neutral
oil losses; and (5) improve the color and flavor stability of the finished
deodorized oil.
Caustic TreatmentThe crude oil is continuously mixed with a proportioned

stream of dilute caustic-soda solution and heated to break the emulsion
formed. Selection of the caustic treatment is determined by the type of crude
oil, FFA content, past refining experience with similar oils, and the refining
equipment available. In general, the minimum amount of the weakest strength
necessary to achieve the desired endpoint should be used to minimize
saponification of the neutral oil and prevent emulsions during separation.
Usually, the best results are obtained with relatively weak caustic solutions for
low-FFA oils and with stronger concentrations for high-FFA oils. Phosphatide
reduction during refining is determined largely by the amount of water present
in the caustic solution. Higher excess caustic treatments remove more
phosphatides, but the increase in removal is caused more by the increased
water than the sodium hydroxide (NaOH). The strength of the caustic solution
is measured in terms of specific gravity expressed in degrees Baume (Be).
Therefore, the caustic treat selected for the crude oil will vary with the FFA
content and the level of caustic excess over the theoretical quantity deter-
mined for each oil. The theoretical quantity of caustic is based on the ratio of
molecular weights of NaOH to oleic fatty acid. This factor is determined as
follows:
NaOH Molecular Weight 40

Factor 0:142:
Oleic Fatty Acid Molecular Weight 282
Thus, the formula for caustic treatment is:
%FFA X 0:142 %Excess Acid Addition

% Treat 100:
% NaOH in caustic
The refining conditions for cottonseed oil are chosen more for the improve-
ment of color because of the presence of gossypol. This pigment is sensitive
to heat and oxidation, forming colored compounds that are difficult to remove
from the oil other than by reaction with caustic. Therefore, a larger excess of a
more concentrated NaOH solution treat is used for cottonseed oil than most
other vegetable oils; good quality cottonseed oils generally require 0.2%
excess, while darker colored oils may require up to 0.4% excess treat (29).
Caustic Oil MixingAfter the caustic reagent has been proportioned into the
crude oil, it must be adequately blended to ensure sufficient contact with the
FFA, phosphatides, and color pigments. The gums are hydrolyzed by the water
in the caustic solution and become oil-insoluble. The caustic solution and oil
are mixed at 3035 or 8695 F in a dwell mixer with a 515-minute
residence time. High-oil temperatures during the caustic addition must be
avoided because they can increase the neutral oil saponification and reduce the
refined oil yield. After the caustic mixing phase is complete, the mixture is
delivered to the centrifuges at a temperature suitable for optimum separation,
218 COTTONSEED OIL
normally the oils are heated to 74 C or 165 F to provide the thermal shock
necessary to break the emulsion.
Soap-Oil SeparationFrom the caustic oil mixer, the resultant soap-in-oil
suspension is fed to the centrifuges for separation into light- and heavy-
density phases. There are various types of centrifuges used in vegetable oil
refining, however, most of them contain a bowl or hollow cylinder that turns
on its axis. The flow of material enters the rotating bowl and is forced outward
to a disc stack. The heavier density soapstock is forced to the outside of the
bowl and flows over the top disc and out the discharge port. The lighter neutral
oil phase moves to the center of the bowl for discharge from the neck of the
top disc. Refining yield efficiency is dependent on this primary separation
step.
Water WashingSodium soaps remaining from the primary centrifugation
phase are readily washable and easily removed from the oils with either a
single or double wash. The refined oil from the primary centrifuge is mixed
with softened water heated 5 C to 8 C or 10 F to 15 F above the oil
temperature at a rate of 10% to 20% of the oil flow. Softened water must
be used to avoid the formation of insoluble soaps. The water-oil mixture
passes through a high-speed, in-line mixer for maximum soap transfer from
the oil to the water phase. The soapy water-oil mixture is centrifuged to
separate the two phases.
Vacuum DryingRefined oil from water washing is usually dried if it is to be
stored before bleaching to remove the traces of water remaining after water
washing to avoid hydrolysis. Typically, washed oil at approximately 85 C or
185 F is sprayed through nozzles into a vacuum chamber controlled at 70 cm
of mercury. This weak vacuum reduces the moisture content to below 0.1%,
most often in the range of 0.05%. Oils that are vacuum bleached immediately
after water washing can bypass the drying step.
Re-RefiningIn handling dark oils, it frequently happens that no matter how
efficient the original caustic refining of a crude cottonseed oil may be, the
bleached color is less than desired. Under these circumstances, it becomes
necessary to re-refine the oil for additional color removal. Usually, 1% or 2%
of a 16 Be caustic treat is sufficient to provide a bleach color of 2.0 to 2.5
Lovibond red color (123).
Miscella Refining. Since 1960, miscella refining has become the most common
method for the treatment of cottonseed oil in the United States. Facilities with an
existing oilseed solvent extraction system found miscella refining to be advanta-
geous by using the same solvent recovery unit for both purposes. Miscella is the
solution or mixture that contains the extracted oil. Both continuous and batch
miscella-refining processes are suitable for most fats and oils. Miscella refining
is especially beneficial for cottonseed oil to provide an oil with a lighter red color
and a high-neutral-oil yield. This type of refining should be done at a solvent extrac-
tion plant as soon as possible, preferably within 6 hours after the oil is extracted
from the oilseed or animal tissue. The advantages for miscella refining, as com-
pared with conventional continuous caustic-soda refining, are (1) higher oil yield,
(2) lighter color oil without bleaching, (3) elimination of the water wash step,
and (4) extraction of the color pigments before solvent stripping has set the
color (124).
For this purification process, the crude miscella source may be from (1) the pre-
evaporator of a direct-solvent extraction plant, (2) a blend of prepressed crude oil
and solvent-extracted miscella from the press-cake, or (3) a reconstituted blend of
crude oil with solvent. In the process, a mixture of approximately 40% to 58% oil
in solvent is heated or cooled to 104 F (40 C) and filtered to remove meal, scale,
and other insoluble impurities. Two solvents that have been used commercially for
miscella refining are hexane and acetone.
Hydrolysis of phosphatides and pigments in the crude oil miscella requires an
acid pretreatment, which usually varies between 100 ppm and 500 ppm by weight
of the oil, depending on the quality of the crude oil. An acid such as phosphoric or
glacial acetic has been found effective in improving oil quality and reducing refin-
ing losses. Phosphoric acid is used more commonly because of its less corrosive
properties and its availability. The acid is mixed with the miscella in a static mixer
to provide an intimately dispersed acid phase that immediately reacts with the crude
miscella.
The pretreated crude miscella is then alkali refined using dilute caustic soda with
a 1624 Be and a 0.20.5% NaOH excess over the theoretical amount required to
neutralize the free fatty acids. The reaction of the caustic soda with the free fatty
acids proceeds rapidly at 130135 F (5457 C), using homogenizers with a shear-
mixing intensity capable of homogenizing milk, hydrolyzing the phosphatides and
pigments with the caustic soda to produce a two-phase mixture. The miscella tem-
perature is adjusted to 135 F (57 C) to obtain the best separation of the heavy phase
or soapstock from the oil or the light phase with the centrifuge. The neutral oil is
then filtered through a diatomaceous earth precoated pressure-leaf filter. At this
point, the refined and filtered miscella can be stripped of the solvent to produce
a neutral yellow oil, or it can be further processed as miscella to dewax, fractionate,
or hydrogenate the oil (125, 126).
PreBleaching. Bleaching is popularly and correctly regarded as the partial or com-
plete removal of color; however, it is also a purification process to prepare the oil
for further processing. Bleaching is relied on to clean up the traces of soap, phos-
phatides, and pro-oxidant metals remaining after caustic neutralization and water
washing that hinder filtration, poison hydrogenation catalysts, darken the oil, and
adversely affect the flavor of the finished oil. Another function, considered primary
by many processors, is the removal of peroxides and secondary oxidation products.
These impurities compete for space on the adsorbent surface of the filter media. The
key process parameters for bleaching include (1) procedure, (2) bleaching media,
(3) temperature, (4) time, (5) moisture, and (6) filtration (106).
ProcedureThe three most common types of contact bleaching methods are

batch atmospheric, batch vacuum, and continuous vacuum. Although vacuum
220 COTTONSEED OIL
bleaching is preferred, atmospheric bleaching can produce quality bleached

oils. Vacuum bleaching, either batch or continuous, is more effective than
atmospheric bleaching because it can use less clay, operates at lower
bleaching temperatures, effects quicker moisture evacuation for less FFA
development, and does not expose the oil to oxidation at high temperatures.
Batch bleaching is preferred to continuous operations when a variety of source
oils are processed in the same system. However, continuous systems are more
efficient and effective for systems dedicated to single source oils.
Bleaching AgentsChemical agents have been proposed and some used, but
practically all edible oil decoloration and purification is accomplished with
adsorptive earths or carbons. The basic kinds of adsorbents used in edible oil
bleaching are neutral clays, activated earths, and activated carbon. Particle
size is a major physical parameter affecting bleaching earth performance
because adsorption theory considers adsorption as a surface phenomena. In
general, the finest particle size earths have the best performance; however, too
small particles create severe filtration problems and oil retention is increased.
The natural bleaching earths, usually referred to as fullers earth, are
bentonite clays that exhibit adsorptive properties in the natural state with
only physical processing. Activated bleaching earths have been treated with
organic or inorganic acids to enlarge the surface and pore volume to make it
selectively attractive to the detrimental components in refined oils. Also, the
activated bleaching earths normally contain 10% to 18% moisture, which
supports the montmorillonite layers in the clays. Activated carbon is effective
in adsorbing certain impurities not affected by earths, but it is used sparingly
due to problems with filtration, relatively high cost, and a high oil retention.
Bleaching Media DosageThe amount of bleaching material used depends
on the type of adsorbent used and the impurities to be removed. Bleaching
earth requirements vary in wide range from 0.15% to 3.0%. On the basis of
adsorbent activity, the acid-activated earths are generally 1.5 to 2 times more
effective than the natural earths. Carbon is rarely used alone but sometimes
employed in admixture with a bleaching earth in a ratio of 1020 parts
bleaching earth to 1 part carbon (49).
TemperatureBleaching earth activity increases as the temperature is
increased by reducing the viscosity of the oil, but decoloration declines after
the optimum temperature is reached. Temperature also affects other properties
of the oil, which dictate that it should be kept as low as possible to minimize
product abuse, but high enough for adequate adsorbance of the impurities and
pigments. The optimum bleaching temperature for nearly all edible fats and
oils ranges between 70 C and 110 C or 160 F to 230 F. Fewer problems are
encountered when the bleaching temperature is maintained below 110 C or
230 F with vacuum bleaching. Secondary oxidation products begin to develop
at temperatures above 110 C or 230 F. Low temperatures favor the retention
of the adsorbed pigment on the bleaching earth surface. Although higher
temperatures favor the movement of pigment molecules into the pores where
chemisorption occurs, it will also promote structural changes of the unsatu-
rated fatty acid groups. Under extremely high temperatures, isomerization of

the unsaturated fatty acid groups and excessive FFA could develop.
TimeIn theory, adsorption should be practically instantaneous; however, in
practice, this is not the case. The rate of decoloration is very rapid during the
first few minutes after the adsorbent comes in contact with the oil and then
decreases to a point where equilibrium is reached and no more color is
adsorbed. Usually 15 to 20 minutes contact time is adequate at a bleaching
temperature above the boiling point of water. Contact time is made up of two
time periods: (1) the time in the bleaching vessel and (2) the time in the filter
during recirculation or final filtering.
FiltrationAfter an adsorbent has selectively captured the impurities, it must
be removed from the oil before it becomes a catalyst for color development or
other undesirable reactions. Filtration, the separation method most often used
for spent bleaching media removal, is the process of passing a fluid through a
permeable filter material to separate particles from the fluid. Examples of the
filtration materials used are filter paper, filter cloth, filter screen, and
membranes. Filter aid, such as diatomite, perlite, or cellulose, are usually
used in conjunction with the permeable filters for surface protection. Tradi-
tionally, either plate and frame or pressure-leaf filters have been used for spent
bleaching media removal. Currently, self-cleaning, closed filters that operate
on an automated cycle are available.
Winterization. When cottonseed oil is designed for use as salad oil, it must be
winterized, that is, a considerable portion of the more saturated glycerides must
be removed so that the material will remain clear when exposed to reduced tem-
peratures, such as those likely to be encountered with refrigeration. If the saturated
glycerides in cottonseed oil are not removed, it will solidify at temperatures
encountered in a refrigerator; 45 F or 7.2 C. The composition of the products
resulting from winterization of cottonseed oil are presented in Table 17 (127). The
solid-fat fraction that settles out is referred to as the stearine.
Winterization is a narrow form of fractionation. Both fractionation and winter-
ization processing operations for edible oils basically consist of the separation of
TABLE 17. Characteristics of Products from Cottonseed Oil Winterization.
Cottonseed Oil Unwinterized Winterized Winterized

Product Cooking Oil Salad Oil Stearine
Fatty acid composition, %

Myristic C14:0 0.8 0.7 0.6
Palmitic C16:0 24.2 22.6 32.4
Palmitoleic C16:1 0.6 0.4 0.3
Stearic C18:0 1.6 2.8 2.4
Oleic C18:1 21.0 19.8 17.2
Linoleic C18:2 51.8 53.7 47.1
Iodine value 107.8 111.4 98.0
Cold test, hours 0 5 plus 0
222 COTTONSEED OIL
oils into two or more fractions with different melting points. In the winterization
process, the oils are cooled in a simple way, kept at the low temperatures for
some time to crystallize solid-fat fractions that would normally cloud when the
oil is held at refrigerator temperatures, and generally separated by filtration. With
fractionation processes, cooling of the oil and the separation of the fractions are
performed with more sophisticated techniques and controlled conditions to provide
substances with unique properties.
The descriptive term of winterization evolved from the observation that refined
and bleached cottonseed oil stored in outside tanks during the winter months phy-
sically separated into clear and hard fractions. Topping or decanting the clear oil
from the top of the tanks provided an oil that remained liquid without clouding
for long periods at cool temperatures. A need for a liquid oil with these character-
istics was created by the introduction of the refrigerator for home use and the
requirements of the mayonnaise and salad dressing industry. The indoors process
developed to simulate the natural winter process consisted of a chilled room held
at 42 F, or 5.6 C, with deep, narrow, rectangular tanks to provide the maximum
surface exposure to cooling. Warm, dry, refined, and bleached cottonseed oil
pumped into the chill room tanks began to cool and crystallize out stearine imme-
diately but slowly. Convection heat transfer simulated the outside storage condi-
tions. Agitation was avoided because it fractured the crystal, causing formation
of small, soft crystals that were difficult to filter. Cooling with the 42 F, or
5.6 C, room temperature, which simulated mild winter conditions closely, required
3 days to produce the desired large crystals for filtering. After the oil temperature
equated with the room temperature, it was held for several hours to allow the stear-
ine or hard fraction to precipitate more fully. The stearine was separated from the
liquid oil by filtering with plate and frame presses. Normally, the oil was gravity fed
to the filters to avoid breaking up the crystals. Winterization is still performed with
the classic technique described above. However, most processors have made equip-
ment and process modifications to improve efficiency; such as jacketed, enclosed
tanks equipped with programmable cooling and agitation, better filtration,
improved pumping methods, and so forth.
Fractionation. Cottonseed oil has melting points spanning a range from 13.3 C,
or 8 F, to 35 C, or 95 F, due to its triglyceride composition. Cottonseed oils trigly-
ceride composition, as determined by high-performance liquid chromatography
(80), liquidity zones (106), triglyceride melting points, and functionality (128) of
each are compared in Table 18 (3). This range of melting points limits the applica-
tions for cottonseed oil, as well as all other edible fats and oils. Application poten-
tial can be increased with fractionation, a thermomechanical process by which an
oil is separated into two or more portions. Thermomechanical separation processes
include distillation and crystallization. Distillation is commercially unsuited for
the separation of triglyceride mixtures because of their low vapor pressure and
relatively low stability at high temperatures. However, separation can be effected
by crystallization.
All edible fats and oils are polymorphic in their crystalline behavior. The three
forms of importance are alpha, beta-prime, and beta, which are of increasing
TABLE 18. Cottonseed Oil Triglyceride Functionality.

Triglyceride HPLC Melting Point

Liquidity Composition Analysis, % F C Functionality
PPO 2.2 95.0 35.0 Structure
Heated SPL 1.5 86.0 30.0 and
PPL 7.1 81.0 27.2 Moisture Barrier
POO 3.1 60.0 15.6 Room
Room SOL 1.3 43.0 6.1 Temperature
Temperature OOO 1.6 42.0 5.6 Clarity
SLL 1.4 34.0 1.1
OOL 3.1 30.0 1.1
POL 14.0 27.0 2.8 Refrigerated
Cool PLL 27.5 22.0 5.6 Clarity
OLL 12.5 20.0 6.7
LLL 19.0 8.0 13.3
P Palmitic; S Stearic; O Oleic; L Linoleic.
stability in that order. The rate of crystallization of the alpha form is greater than
that of the beta-prime, which is greater that of the beta polymorph. If supercooling
is carried out too rapidly, crystallization of the alpha form occurs resulting in a mass
of very small crystals. To obtain good separation of cottonseed oil fractions, crystal-
lization is required in the beta-prime form. This is because the beta-prime crystals
agglomerate into large aggregates that are firm and of uniform spherical size that
separate well with filtration.
Fractional crystallization is a thermomechanical separation process wherein
component triglycerides of oils are separated, usually as a mixture, by partial crys-
tallization in a liquid phase. In the fractionation processes, three successive stages
are recognized:
1. Cooling of the liquid oil to supersaturation, resulting in the formation of

nuclei for crystallization;
2. Growth of the crystals by gradual cooling to a shape and size that permits
efficient separation; and
3. Separation, isolation, and purification of the resultant crystalline and liquid
phases.
In the fractionation process, the minor components of the original oil become
concentrated in the separated fractions. This concentration has a considerable effect
on the oxidative stability of the individual fractions. Relative to the starting oil, the
liquid or soft fraction is enriched in tocopherols and depleted of trace metals. The
reverse occurs with the hard or stearine fraction, which becomes appreciably more
susceptible to oxidation despite its lower content of unsaturates. The stearine frac-
tion is also the recipient of other impurities remaining in the oil after refining and
bleaching, such as phosphatides and soap.
224 COTTONSEED OIL
There are three processes in commercial use for the fractionation of edible fats
and oils:
Dry Fractionation - The principal of this fractionation process is based on the

cooling of oil under controlled conditions without the addition of chemicals or
solvents. The liquid and solid phases are separated by filtration.
Detergent Fractionation - The oil is crystallized on its own similar to the dry
fraction technique, but separation is affected by employing an aqueous
detergent solution and centrifugation.
Solvent Fractionation - The separation of component triglycerides that differ
in solubility is accomplished by fractional crystallization of a solution of oil in
an organic solvent, followed by separation of the solids from the liquid by
filtration, and finally, removal of the solvent from the separated fractions by
steam stripping. This process is the most versatile of the fractionation
techniques presented.
Some processors have employed solvent fractionation systems to produce salad
oil, which has demonstrated three major advantages over the traditional winteriza-
tion process: (1) a considerably lower viscosity, which allows a faster crystal
growth for more rapid stearine separation; (2) the cottonseed salad oil produced
has a better resistance to clouding at cool temperatures for longer cold tests; and
(3) less liquid oil is trapped in the stearine component for higher salad oil yields.
An operational continuous solvent process was described by Cavanagh (129) for
winterization of cottonseed oil. A 50% solution of oil in hexane is cooled rapidly
with a heat exchanger to 20 F to 26 F, or 6.6 C to 3.3 C. After cooling, it
passes through a continuous winterization column that cools with a series of agi-
tated trays over a 4060 minute period to temperatures as low as 40 F or 40 C.
The stearine or solid stream is separated from the liquid miscella stream with a con-
tinuous solids discharge centrifuge. The solvent is removed from the salad oil por-
tion with an evaporator system before deodorization. The comparison in Table 19 of
TABLE 19. Cottonseed Salad Oil Stearine Analysis.
Winterization Process Conventional Solvent
Iodine value 95.5 71.6

Solids fat index, % at:
10 or 50 F 21.6 52.3
21.1 C or 70 F 1.3 33.7
26.7 C or 80 F 1.2
33.3 C or 92 F 0.1
Fatty acid composition, %
Myristic C14:0 0.7 0.6
Palmitic C16:0 34.6 52.1
Palmitoleic C16:1 0.6 0.8
Stearic C18:0 2.1 1.9
Oleic C18:1 15.8 9.1
Linoleic C18:2 46.2 35.5
conventional verses solvent process produced stearine confirms that the solvent sys-
tem produces a harder, firmer, more compact stearine crystal with less entrapped oil
than the conventional process.
Fractionation technology, in particular solvent fractionation, has been utilized to
produce some very highly specialized edible oil products. High-stability liquid oils,
with AOM stability results of 350 hours minimum without the benefit of added
antioxidants, and cocoa butter equivalents are two examples of products that can
be produced with fractionation technology. Fractionation technologies may also
be used to produce basestocks for utilization as components in finished products
for various applications.
Hydrogenation. The hydrogenation process is an important tool for the edible fats
and oils processors. With hydrogenation, cottonseed oil can be converted from a
liquid oil into a plastic or solid fat more suitable for numerous applications. There
are two reasons to hydrogenate fats and oils: (1) to convert naturally occurring oils
into physical forms with melting and handling characteristics more suited to the
desired product functionality; and (2) to improve oxidative stability, which provides
prolonged organoleptic acceptability. A wide range of edible oils products can be
produced with the hydrogenation process, as shown previously (see Figure 3),
depending on the conditions used, the starting oils, and the degree of saturation
or isomerization.
During the hydrogenation process, hydrogen is chemically attached at the
double bond sites on the carbon chain of the unsaturated fatty acids. This reac-
tion eliminates a double bond and converts an unsaturated fatty acid to a more satu-
rated fatty acid. Isomerization during the hydrogenation process can also create
trans-isomers. Both of these chemical changes increase the melting point of the
reacted oil.
Typically, an unhardened cottonseed oil is composed of 2.5% stearic fatty acid
and 2223% palmitic fatty acid, which are the principal saturated fatty acids. The
unsaturated fatty acids are composed of approximate 18% oleic, 54% linoleic, and
less than 1% each of linolenic and palmitoleic fatty acids. If the hardening reaction
was completely selective, before any hydrogen was absorbed by oleic fatty acid the
linoleic fatty acid would have to be completely converted to oleic. On the disap-
pearance of the linoleic fatty acid, the oleic would next absorb hydrogen to be con-
verted to the fully saturated stearic fatty acid. This degree of selectivity is never
attained in practice but how closely it is approximated depends on the catalyst
type and dosage, temperature, and pressure.
Hydrogenation can take place only when the three reactants have been brought
together: unsaturated oil, catalyst, and hydrogen gas. The hydrogen gas must be dis-
solved in the liquid before it can diffuse through the liquid to the solid catalyst sur-
face. Each absorbed unsaturated fatty acid can then react with a hydrogen atom to
complete the saturation to the double bond, shift it to a new position, or twist it to a
higher melting trans-form. When the unsaturated oil to be hydrogenated contains
mono-, di-, and tri-unsaturates, there is competition for the catalyst surface. The di-
and tri-unsaturates are preferentially absorbed from the oil to the catalyst surface
and partially isomerized or hydrogenated to a mono-unsaturate until their
226 COTTONSEED OIL
concentration is very low, permitting the mono-unsaturate to be absorbed and

reacted. The variables that can affect the hydrogenation reaction are temperature,
agitation, hydrogen pressure, type and quantity of catalysts, feedstock quality,
and the fatty acid composition of the source oil. The effects of these variables
are (106) described below.
Temperature - Hydrogenation is operated at temperatures ranging from 150

230 C or 300450 F. Hydrogenation, like most chemical reactions, proceeds
at a faster rate at higher temperatures. The hydrogenation rate is increased as
the temperature is elevated resulting from an increase in hydrogen gas
solubility in the liquid oil. Higher reaction temperatures increase selectivity
and trans-isomer formation and hydrogenation rate to provide steep solid-fat
index curves. Hydrogenation is an exothermic reaction; it creates heat as long
as the reaction is active. The reaction temperature is increased by 1.61.7 C,
or 2.93.1 F, for each iodine value drop.
Agitation - The function of agitation is to supply dissolved hydrogen gas to
the catalyst surface. Agitation decreases selectivity and isomerization by
keeping the catalyst supplied with sufficient hydrogen to increase the reaction
rate.
Hydrogen Pressure - Most edible oil hydrogenations are performed at
hydrogen pressures ranging from 0.84.0 atmospheres. At low pressure, the
hydrogen gas dissolved in the oil does not cover the catalyst surface, while at
high pressure, hydrogen is readily available for saturation of the double bonds.
Higher pressure increases the saturation rate, which results in a decrease in
trans-formation and selectivity to produce flatter SFI curves.
Catalyst Level - Hydrogenation reaction, selectivity and trans-isomer forma-
tion increase as the catalyst concentration is increased, up to a point where it
levels off. The rate increase is caused by increased activity at the catalyst
surface. The maximum is reached when hydrogen will not dissolve quickly
enough to supply the catalyst levels.
Catalyst Type - Nickel metal catalyst, sometimes promoted with copper,
aluminum oxide, or sulfur, are commonly used in commercial hydrogenation.
These catalysts are prepared by a variety of techniques, some proprietary to
the catalyst supplier, to provide the surface activity necessary for the desired
selectivity. Precious metals have been found to be effective hydrogenation
catalysts, which are more active at lower temperatures and produce less trans-
isomers. However, their use has been deterred by the initial cost and recovery
problems associated with the minute quantities required.
Source Oil - Hydrogenation selectivity also depends on the type of unsatu-
rated fatty acids available and the number of unsaturated fatty acids per
triglyceride. Those oils with high-linolenic and high-linoleic fatty acid levels
hydrogenate more rapidly and to higher melting points than oils with high-
oleic fatty acid levels. The relative hydrogenation reactivity for the 18-carbon
fatty acids was described by Hastert (130),
Fatty Acid Relative Reactivity
Linolenic C18:3 40
Linoleic C18:2 20
Oleic C18:1 1
Cottonseed oil with a high-linoleic fatty acid content and a saturated fat level higher
than any of the other linoleic-classification oils requires less hydrogenation to attain
the same degree of hardness.
Hydrogenation Systems. Batch hydrogenation is most commonly used for
edible oil processing because of the simplicity and flexibility for use with different
source oils. Essentially all that is required is a reaction vessel, usually referred to as
a converter, that can withstand 7- to 10-bar pressure, with an agitator, heating and
cooling coils, a hydrogen gas inlet, piping and pumps to move the oil in and out,
and a sample port for process control of the reaction. The converter must also be
provided with the means to control the three reaction variables of pressure, tem-
perature, and rate of reaction.
Continuous hydrogenation systems have been available for quite some time, but
their commercial usage has been limited for several reasons. The maximum value
for any continuous operation is realized when it is used to produce large quantities
of the same product. As most edible oil processors produce a variety of products,
several different basestocks are routinely required that can be produced more effi-
ciently, with better uniformity using batch hydrogenation systems.
Basestock Systems. Most prepared foods are formulated with ingredients
designed for their application or, in many cases, specifically for the particular pro-
duct or processing technique employed by the food processor. These customer-
tailored products have expanded the product base for edible oil processors from
a few basic products to literally hundreds. Each of these products could be formu-
lated to require a slightly different hydrogenated oil. This practice, with the ever
increasing number of finished edible oil products, could be an inventory nightmare
with a large number of product heels remaining from overproduction waiting for
the next production run. Basestock systems with a limited number of hydrogenated
stock products for blending to meet the finished product requirements are used by
most processors for better control and efficiency.
Basestock requirements will vary with each processor, depending on the custo-
mer requirements, which dictate the finished products produced. A basestock sys-
tem can include several source oils, or be limited to a single source oil if that oils
composition can provide all of the properties necessary for the finished products
functionality. Cottonseed oil is the closest to meeting these qualifications for
most product applications: It has a good oxidative stability because of the higher
level of natural saturated fatty acids; it has a high-palmitic fatty acid content, which
helps to make it stable in the beta-prime crystal form; and it is a rich source of lino-
leic acid, an essential fatty acid that the human body cannot synthesize, when not
hydrogenated to a great extent. Table 20 (131) presents a basestock system that uses
only cottonseed oil. Use of this basestock system should enable a processor to
228 COTTONSEED OIL
TABLE 20. Cottonseed Oil Basestock System.
Basestock Type RB Flat Steep Saturated

Iodine Value 110 80 75 70 65 58 >5
Mettler dropping point, C 32.0 34.0 36.0 38.0 43.5 Too Hard
Solids fat index, % at:
50 F or 10 C 20.5 37.0 47.0 63.0 Too Hard
70 F or 21.1 C 8.0 10.5 21.5 30.5 51.0 Too Hard
80 F or 26.6 C 5.0 7.0 14.0 18.0 40.0 Too Hard
92 F or 33.3 C 1.5 2.5 3.5 10.5 25.0 Too Hard
104 F or 40 C 10.5 Too Hard
Titer, C 60.0
Fatty Acid Composition, %
Myristic C14:0 0.7 0.7 0.7 0.7 0.7 0.7 0.7
Palmitic C16:0 22.5 22.6 22.7 22.7 22.6 22.9 23.2
Palmitoleic C16:1 0.7 0.6 0.6 0.5 0.5 0.5
Stearic C18:0 2.5 5.7 6.8 3.9 7.0 7.4 71.3
Oleic C18:1 18.2 48.2 51.0 61.4 62.6 65.0 4.3
Linoleic C18:2 54.1 21.5 17.7 10.0 6.2 2.4
Linolenic C18:3 0.8 0.2 0.3
Arachidic C20:0 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Behenic C22:0 0.2 0.2 0.2 0.2 0.2 0.2 0.2
Hydrogenation conditions: None Nonselective Selective Nonselect

Gassing temp., F 300 300 300
Hydrogenation temp., F 350 440 450
Pressure, bar 2.0 to 3.0 0.7 to 1.0 5.0
Catalyst, % Nickel 0.02 0.04 to 0.08 0.4 to 0.8
Agitation Fixed Fixed Fixed
Stands for a range of values.
produce most formulated edible oil products by blending two or more of the base-
stocks, except for some specialty products that can only be made with special
hydrogenation conditions or a lauric oil. The basestocks represent saturation levels,
starting with natural cottonseed oil followed by six hydrogenated basestocks
ranging from an iodine value of 80 to almost complete saturation. The basestock
types are as follows:
RBRB is an abbreviation for refined and bleached. Oils are refined and
bleached before hydrogenation to remove impurities and catalyst poisons.
FlatMany food products require a shortening, margarine, or tailored
formulated oil product that has an extended plastic range with a good
oxidative stability. These products must be relatively soft, with a plastic or
workable consistency at room temperature while still possessing some body at
temperatures of 37.8 C or 100 F, and with a melting point only slightly above
body temperature. So-called nonselective hydrogenation conditions are used
to produce a flat SFI curve basestock for these products. A flat SFI curve are
those SFI values that provide east to west (x-axis) directional slopes when
plotted on a graph.
SteepPoor plasticity but good mouthfeel and oxidative stability are indi-
cated by a north-to-south (y-axis) directional SFI slope, usually referred to as
steep. A narrower span of temperatures with higher SFI values provides
sharper melting tendencies for an improved palatability and a good oxidative
stability. The preferred bland flavor is retained longer because of a lower
unsaturated fatty acid content. However, steep SFI curves produced with
hydrogenation usually indicate a higher trans-fatty acid content.
SaturatedTechnically, an oil should have to have a zero iodine value to be
designated as fully saturated. In theory, it is possible to attain a zero IV, but it
is impractical in production. These basestocks are also referred to as fully
hydrogenated fats, low iodine value hardfats, stearines, or hardfats. Beta prime
hardfats are added to softer basestocks to extend the plastic or workable range,
to improve the products tolerance to high temperatures, and to establish the
finished products crystal habit. Hydrogenation conditions to produce hardfats
is the least critical of all hydrogenation operations because neither selectivity
nor isomerization are factors, and unsaturation is very minor because the oil is
almost totally saturated.
Post Bleaching. A separate bleaching operation immediately after the hydro-

genation process has three purposes: (1) to ensure that all traces of the prooxidant
hydrogenation catalyst that may have escaped the filtration system after hydro-
genation have been captured, (2) to remove undesirable colors that may have
been accentuated during hydrogenation, and (3) to remove peroxide and secondary
oxidation products. Postbleach systems can be exact duplicates of the prebleach
process. One difference is that a chelating acid, either citric or phosphoric, should
be used to ensure that any residual nickel content is reduced to the lowest possible
level. Batch systems are preferred for postbleaching over continuous for the
same reasons as for hydrogenation systemsproduction of a wide variety of
hydrogenated basestocks (132). Postbleaching systems are also used to remove
undesirable impurities formed during other modification processes, such as interes-
terification, fractionation, and blending.
Interesterification. The least known and practiced processing technique available
to the fats and oils processor for modification of the physical properties of an oil is
interesterification, often referred to as rearrangement. The ability to modify the
melting point and functional crystallization characteristics without changing the
fatty acid composition makes interesterification a process with a number of unique
possibilities. The benefits of the random interesterification processes are summar-
ized in Table 21 (133).
TABLE 21. Random Interesterification Process Benefits.
Modified Characteristics Unaffected Characteristics
Melting point Iodine value

Solids fat index Fatty acid composition
Crystal habit Trans-isomer content
Oxidative stability Nutritional aspects
230 COTTONSEED OIL
TABLE 22. Interesterification Process Comparison.
Process Random Directed
Type reaction Nonselective Selective

Reaction time 13 to 30 minutes 6 to 24 hours
Catalyst usage, % 0.05 to 0.1 0.1 to 0.3
Temperature, F 90 to 140 20 to 60
Filtration required none yes
One of the most important types of interesterification with respect to applications

is chemically catalyzed random-ester interchange. Natural fats and oils have speci-
fic fatty acid distributions and positional arrangements that determine their physical
and functional characteristics such as melting point and oxidative stability. During
the interesterification process, the triglyceride ester bonds are broken, the resulting
fatty acids mix together, and eventually they reattach. If the fatty acids reattach to
the same glycerol molecules, the reaction is called intraesterification. When the
fatty acids attach to different glycerol molecules, it is called interesterification. A
random equilibrium arrangement of the fatty acids is reached if the reaction is
allowed to continue long enough. The primary benefits are modification of the phy-
sical properties, such as, melting point and solid fat index values, and modification
of the crystal form tendencies. Similar types of physical changes are possible with
hydrogenation or fractionation, but both of these processes change the fatty acid
distribution of the final product.
Two basic types of chemical interesterification are practiced: random and direc-
ted. Both involve the use of transition metals, such as sodium, or more commonly
derivatives, such as sodium methoxide, as a catalyst. The differences between these
two interesterification reactions are summarized in Table 22.
Random chemical rearrangement of fats and oils can be accomplished with
either a batch or continuous process. Both perform the three important rearrange-
ment steps: (1) pretreatment of the oil, (2) reaction with the catalyst, and (3) deac-
tivation of the catalyst. A typical batch rearrangement reaction vessel is equipped
with an agitator, coils for heating and cooling, nitrogen sparging, and vacuum cap-
abilities. The process steps are as follows:
1. Heat the oil to 250 F to 300 F or 120 C to 150 C in the reaction vessel under
a vacuum to dry the oil. Drying is critical because moisture deactivates the
catalyst.
2. After drying, cool the oil to the reaction temperature, suck in the catalyst with
the vacuum, and agitate for 30 to 60 minutes or until the distinctive brown
color is formed indicating randomization. Analyze the reaction mixture with a
refractometer to determine if the reaction is complete or requires additional
catalyst and/or time to reach the predetermined analytical endpoint.
3. When reaction completion is confirmed, the catalyst is neutralized either with

the addition of phosphoric acid or water washing.
4. The neutralized interesterified oil must be postbleached to remove the brown
color formed during the reaction before blending and/or deodorization.
In directed rearrangement processes, one or more of the trisaturated glycerides

are selectively removed from the ongoing reaction when the mixture is cooled
below its melting point. This selective crystallization upsets the equilibrium, for-
cing the reaction to produce more trisaturated glycerides to reestablish equilibrium.
Continuous processes are normally used for directed interesterification because the
batch process is difficult to control and a number of reaction vessels would be
required. Scraped-wall heat exchangers cool the mixture to initiate crystallization
and later remove the heat liberated when the trisaturates crystallize. The trisaturates
formed can be separated by filtration, or the reaction can be controlled to produce
the trisaturate level desired for a compound-type shortening formulation. In either
case, the reaction is usually terminated by adding water to inactivate the catalyst.
After neutralization of the catalyst, the liquid soap phase is normally removed with
centrifugation and vacuum drying followed by bleaching to remove the undesirable
impurities formed.
Table 8 lists the triglyceride composition for cottonseed oil as determined by
HPLC. The triglyceride distribution for cottonseed oil is considered even or nonran-
dom because the saturated fatty acids are positioned predominately in the sn-1 or -3
positions with the unsaturated fatty acids in the sn-2 position. Random rearrange-
ment shuffles the fatty acid distribution to effect a melting point increase from
50.9 F or 10.5 C to 93.2 F or 34 C (134). With directed interesterification, a solid
fat may be produced using natural cottonseed oil as the feedstock. The trisaturated
triglycerides are crystallized and precipitate as they are formed. The unsaturated
portion that remains may be used as a salad oil, as most of the saturated fatty acids
have been removed (29). Interesterification can be used to produce basestocks
similar to those produced with hydrogenation or fractionation for blending to
formulate shortenings, margarines, high-stability liquid oil products, and other
specialty products for the food processor.
Blending. Various basestocks are blended to produce the specified composition,
consistency, and oxidative stability for edible fats and oils products, such as short-
enings, frying fats, margarine oils, specialty products, and even some salad oils.
The basestocks may be composed of natural source oils and/or modified oils pro-
duced with hydrogenation, fractionation, or interesterification processes. Blends are
made to meet both the composition and analytical consistency controls identified by
the product developers and quality assurance. The consistency controls can include
specific limits for SFI, IV, melting point, fatty acid composition, or other analysis
specific to the physical characteristics of the particular product. The blending pro-
cess requires storage tanks to inventory the basestocks and scale tanks and/or
meters to proportion the basestocks accurately for each different product. The blend
tanks should be equipped with agitators and heating coils to assure a uniform
blend (106).
232 COTTONSEED OIL
Deodorization. Edible fats and oils retain undesirable flavors and odors after refin-
ing, and they develop other organoleptic undesirables during bleaching, hydrogena-
tion, fractionation, and interesterification processing. Deodorization is basically a
vacuum-steam distillation process operated at elevated temperatures to remove
FFA and other volatile odoriferous components that cause the undesirable flavors
and odors. Additional deodorization benefits include heat bleaching to destroy car-
otenoid pigments, pesticide removal, and cyclopropenoid fatty acid reduction to a
negligible level, all of which ensure oil purity. Deodorization is the last major pro-
cessing step where flavor, odor, and many of the other qualities of an edible fat and
oil product can be controlled. From this point forward, all efforts are directed
toward retaining the quality of the deodorized product.
Experience has shown that edible fats and oils flavor and odor removal correlates
well with the reduction of FFA. The odor and flavor of an oil with a 0.1% FFA will
be eliminated when the FFA is reduced to 0.01% to 0.03%, assuming a zero per-
oxide value. Therefore, all commercial deodorization consists of steam stripping
the oil for FFA removal. Typical conditions practiced in the United States for the
three deodorizer system types are shown in Table 23. The four interrelated operat-
ing variables that influence deodorizer design are vacuum, temperature, stripping
rate, and retention time at deodorization temperatures.
TABLE 23. Typical Deodorization Conditions.
Deodorization Conditions Range
Vacuum, absolute pressure, mbar 2 to 4

Deodorization temperature, F 410 to 500
Deodorization temperature, C 210 to 260
Retention time at deodorization temperature:
Batch deodorizer, hours 3 to 8
Continuous deodorizer, minutes 15 to 120
Semicontinuous deodorizer, minutes 15 to 120
Stripping steam, weight percent of oil
Batch deodorizer 5 to 15
Continuous deodorizer 1 to 5
Semicontinuous deodorizer 1 to 5
Drop temperature
Liquid oils, F 100 to 120
Liquid oils, C 37.8 to 48.9
Higher melting products:

F above melting point 10 to 15

C above melting point 5.5 to 8.5
Product free fatty acid, %
Feedstock 0.05 to 6.0
Deodorized Product 0.02 to 0.03
Product peroxide value, meq/kg
Feedstock 2.0 max
Deodorized product zero
Deodorized product flavor bland
VacuumDeodorization must be carried out at an extremely low absolute

pressure to promote boiling of fatty acids like palmitic, stearic, and oleic to
permit distillation of the odoriferous substances.
TemperatureIncreases in temperature combine with decreases in pressure to
accelerate deodorization. However, an excessive increase in temperature
promotes polymerization, trans-isomer development, excessive removal of
tocopherols and sterols, color reversion, and the formation of odoriferous low-
boiling products.
Stripping SteamInjection of steam into the oil modifies the vapor pressure
of the materials to be distilled to effect deodorization. Adequate stripping
steam, consistent with the temperature and pressure, is required; however, too
much live steam may cause hydrolysis, recreating FFA.
Retention TimeRetention time is the period during which the oil is at
deodorization temperature and subjected to stripping steam. Factors such as
temperature, vacuum, depth of the oil layer, and product type can vary the
retention time required to properly deodorize an oil.
Deodorization is a multistep process that includes deaeration, heating, deodori-

zation/deacidification, cooling, metal chelating, and polish filtration. Deodorization
equipment in current use can be classified into three principal types: batch, contin-
uous, and semicontinuous. Batch deodorization, the original method used for edible
oils, consists of a vacuum vessel containing internal heating and cooling coils, strip-
ping steam injection apparatus, steam-jet ejector vacuum equipment, high-tempera-
ture heating media source, and polish filtration. These units are now less popular
because of (1) high stripping steam requirements, (2) high vacuum ejector motive
steam consumption (3) long deodorization cycle time, and (4) the finished product
quality is not consistent. Batch systems are still used in circumstances where low
construction costs, small capacities, and specialized or gourmet end products are
involved.
Continuous deodorization systems are used when the feedstock is constant and
the same material is processed for several days. These systems generally have a
modest investment cost with the greatest energy efficiency and, therefore, the low-
est operating cost. The continuous flow allows uniform temperatures during heating
and cooling, permitting smaller ancillary equipment. The general approaches for
continuous deodorizer design are to carry out heating, cooling, and heat recovery
in exchangers external to the deodorizer, or to do the heating, cooling, and heat
recovery within the deodorizer unit. The internal approach is less efficient for
heat recovery but affords a more efficient, more reliable method for product
changes than the external heat exchange. Continuous deodorization benefits are
lost with as few as two or three stock changes in a 24-hour period because of
loss of production time and likelihood of commingling products.
Semicontinuous systems operate on the basis of handling finite batches of oil in a
timed sequence of deaeration, heating, holding-steam stripping, and cooling so that
all of the oil is completely subjected to each condition before proceeding to the next
234 COTTONSEED OIL
step. Most semicontinuous deodorizers consist principally of a tall cylindrical shell

of carbon steel construction with five or more type-304 stainless steel trays stacked
inside of the outer shell. Each tray is fitted with a steam sparge and is capable of
holding a measured batch of oil. By means of a measuring tank, oil is charged to the
top tray, where it is deaerated while being heated with steam to about 320 F to
330 F or 160 C to 166 C. At the end of the heating period, the charge is automa-
tically dropped to the second tray, and the top tray is refilled with oil. In the second
tray, the oil is heated to the operating temperature and, again after a timed period, is
automatically dropped to the tray below. When the oil reaches the bottom tray, it is
cooled to 100 F to 130 F or 38 C to 54 C and discharged to a drop tank from which
it is pumped through a polishing filter to storage.
Deodorization Process Control. Deodorization is the last processing step, in
which flavor, odor, and many of the stability qualities of an edible fat and oil pro-
duct can be controlled. To produce quality deodorized products, attention must be
focused on all of the factors involved with the process. The deodorization physical
process removes the volatile, odoriferous materials present in the oils. The
other factors that influence the quality of the deodorized oil products are as follows
(106):
Undeodorized oil preparationThe first process control requirement is to

assure that the processing of the oil prior to deodorization has been preformed
properly. Preparation of the oil before deodorization has a significant effect on
the product after deodorization. For example, deodorization will remove the
hydroperoxides from abused oils, but the secondary oxidation products
formed will accelerate the rate of oxidation during storage to compromise
the flavor and odor. With proper process control, the abused oil would have
been bleached prior to deodorization to remove the aldehydes and ketones that
make up the secondary oxidation products. Two other deleterious impurities
that deodorization will not remove are soap and phosphatides, which must be
eliminated in the up-stream processes.
Air eliminationOils must be protected from air before, throughout, and after
the deodorization process. At deodorization temperature, the oil reacts
instantly with oxygen to become oxidized, which, in turn, causes flavor
reversion. Following are recommended procedures to eliminate potential air
sources before and during deodorization: (1) Deaeration of the feedstock is
essential to remove oxygen from previous exposure; (2) air leaks at fittings
below the oil level and in external pumps, heaters, and coolers should be
checked and stopped; and (3) the stripping steam must be generated from
deaerated water to be oxygen-free. After deodorization, the oils must be
protected from air to preserve the deodorized oil quality. The usual procedure
is to replace air with nitrogen. Oxygen contact can be reduced considerably by
keeping the entire handling system after deodorization protected with an inert
gas like nitrogen. Gas-free finished oil can be delivered from the deodorizer to
a storage tank under a complete nitrogen blanket, or the oil can be sparged
with nitrogen at the deodorizer exit. Thereafter, effective protection against
REGULATORY CONSIDERATIONS: COTTONSEED OIL EXTRACTION 235
oxidation requires the product to be protected by nitrogen gas in the storage

tanks and bulk transports, as well as in packaging.
Metal chelatingTrace metals may be absorbed from the soil by the plants
during the growing season, and later in the oil from contact during crushing,
storage, and the other processes and transfers. Many of the trace metals
promote autoxidation, which results in off-flavors and odors accompanied by
color development. Studies have identified copper as the most harmful metal,
with iron, manganese, chromium, and nickel following. The effects of
prooxidants can be diminished by using chelating agents before and after
deodorization. The most commonly used chelating agents are citric acid,
phosphoric acid, and lecithin. Deodorized oils are usually treated with citric
acid during the cooling cycle at 50 to 100 ppm.
Oil polishingThe final stage of deodorization is filtration of the oil to
remove any fine particles of soaps, metallic salts, rust, filter aid, polymerized
oil, or any other solid impurities.
Antioxidant additionVegetable oils contain tocopherols that are natural
antioxidants. Tocopherols are removed in the refining, bleaching, and deodor-
ization processes, but enough survive to provide the optimum stability
available from the natural antioxidants. Several phenolic compounds have
been identified that can also provide oxidative stability and longer shelf life
for fats and oils by delaying the onset of oxidative rancidity. Phenolic
substances or antioxidants function as free radical acceptor, thereby terminat-
ing oxidation at the initial step. Several antioxidant compounds are available,
but tertiary-butylhydroquinone (TBHQ) is the most effective antioxidant for
vegetable oils.
5. REGULATORY CONSIDERATIONS: COTTONSEED OIL

EXTRACTION AND PROCESSING
Many workplace, environmental, food safety, and other regulations apply to oilseed
and oil processors (55). Some of the regulations required in the United States are
discussed. Many other countries have similar requirements, but if they do not, it
would be prudent for oilseed solvent extraction operations and vegetable oil proces-
sors to consider meeting these regulations and for these industries to have environ-
mental, health and safety, and quality management programs (135, 136).
5.1. Workplace Regulations (OSHA)

Workplace regulations are promulgated and enforced in the U.S. by the Occupa-
tional Safety and Health Administration (OSHA), which is part of the U.S. Depart-
ment of Labor, under the Occupational Safety and Health Act (OSH Act; PL 91-596
as amended by PL 101 552; 29 U.S. Code 651 et. seq.). OSHA general industry
health and safety standards (29 CFR 1910) apply to oilseed extraction and oil
236 COTTONSEED OIL
refining. In addition, even if there is not a specific standard, OSHA can site a
facility under the general duty clause [Sec. 5(a)(1) of the OSH Act], because
the OSH Act requires the employer to maintain a safe and healthful workplace.
Some health and safety standards that affect cottonseed oil extraction and
processing are as follows:
1. OSHA Health Standards:

a. Air Contaminants Rule, 29 CFR 1910.1000 (the permissible exposure
limit [PEL] for n-hexane is 500 ppm [1800 mg/m3], 8-hr time-weighted-
average; there are PELs for sodium hydroxide, sulfuric acid, vegetable oil
mist, nuisance dust, etc.)
b. Hazard Communication Standard, 29 CFR 1910.1200
c. Cotton Dust Standard, 29 CFR 1910.1043 (cottonseed oil mills have
medical surveillance, recordkeeping, and reporting requirements)
d. Bloodborne Pathogens, 29 CFR 1910.1030
e. Occupational Exposure to Hazardous Chemicals in Laboratories, 29 CFR
1910.1450
2. OSHA Safety Standards:
a. Process Safety Management, 29 CFR 1910.119 (for n-hexane)
b. Emergency Action Plan, 29 CFR 1910.38(a)(1)
c. Fire Prevention Plan, 29 CFR 1910.38(b)(1)
d. Fire Brigades, 29 CFR 1910.156
e. Permit-required Confined Space, 29 CFR 1910.146
f. Lockout-Tagout, 29 CFR 1910.147
g. Occupational Noise Exposure, 29 CFR 1910.95 and Hearing Conservation
Program, 29 CFR 1910.95(c)
h. Personal Protection Equipment:
General Requirements, 29 CFR 1910.132
Eye and Face Protection, 29 CFR 1910.133
Respiratory Protection, 29 CFR 1910.134
Head Protection, 29 CFR 1910.135
Foot Protection, 29 CFR 1910.136
5.2. Environmental Regulations (EPA)

The U.S. Environmental Protection Agency (EPA) administers all regulations
affecting the environment and chemicals in commerce. EPA regulations are
intended to protect human health and welfare and the environment. The individual
states and state environmental regulatory control boards implement and enforce
most of the regulations.
The legislation that serves as the basis for the regulations can be divided into:
1. Statutes that are media-specific [Clean Air Act (CAA) and Clean Water Act
(CWA)]
2. Statutes that manage solid and hazardous waste [Resources Conservation and
Recovery Act (RCRA) and Comprehensive Environmental Response, Com-
pensation and Liability Act (CERCLA; Superfund)]
3. Statutes that directly limit the production rather than the release of chemical
substances [Toxic Substances Control Act (TSCA) and Federal Insecticide,
Fungicide and Rodenicide Act (FIFRA)]
Some of the more important environmental regulations that affect oilseed extraction
and processing are as follows:
1. Clean Air Act (CAA; 42 U.S. Code 7401 et seq.). States and state air control
boards are required to implement regulations and develop state implementa-
tion plans (SIP) (137). Hazardous air pollutants (HAP), such as n-hexane, are
regulated with National Emissions Standards for Hazardous Air Pollutants
(NESHAP) and criteria pollutants [e.g., ozone (O3), particulate matter (PM),
nitrogen oxides (NOx), sulfur oxides (SOx), carbon monoxide (CO), and lead
(Pb)] are regulated with National Ambient Air Quality Standards (NAAQS).
n-Hexane is a regulated HAP but isohexane and acetone are not. Regulated criteria
pollutants, such as O3, PM, CO, and NOx, also are emitted during the extraction and
refining of cottonseed oil.
a. Hazardous Air Pollutants (HAP) or Air Toxics (40 CFR 61): If a facility is a
major emitter of n-hexane, the EPA requires sources to meet national
emissions standards (119). The air toxic control measures for source cate-
gories are technology-based emission standards (not health based) established
for major sources (10 tons/yr of one HAP or 25 tons/yr of total HAP) that
require the maximum degree of reduction emissions, taking costs, other
health and environmental impacts, and energy requirements into account.
Compliance with a NESHAP involves the installation of Maximum Achiev-
able Control Technology (MACT), which essentially is maximum achievable
emission reduction. The NESHAP for Solvent Extraction for Vegetable Oil
Production (4/12/01, 65 FR 34252; 40 CFR 63 subpart GGGG) requires all
existing and new solvent-extraction processes that are major sources to meet
HAP emission standards, as a 12-month rolling average based on a 64%
n-hexane content. Solvent-extraction facilities covered are those that produce
crude vegetable oil and meal products by removing crude oil from listed
oilseeds (corn germ, cottonseed, flax, peanuts, rapeseed, safflower, soybeans,
and sunflower) through direct contact with solvent. HAP emission standards
(solvent loss factor) for cottonseed oil production are: for cottonseed, large
238 COTTONSEED OIL
(process>120,000 tpy) 0.5 gal/t and for cottonseed, small (<120,000 t/yr)
0.7 gpt. Facilities have until 4/12/04 to get into compliance. As the emission
loss factor values are 12-month rolling averages, the first compliance report
would be due no sooner than 48 months after the standard was promulgated
(i.e., 4/12/05).
b. NAAQS: The NAAQS are set at levels sufficient to protect public health,
including the health of sensitive populations (primary air quality standards)
and public welfare (secondary air quality standards) from any known or
anticipated adverse effect of the pollutant with an adequate (appropriate)
margin of safety.
Volatile organic compounds (VOC) are essentially considered the same as
the criteria pollutant ozone (119). n-Hexane and hexane isomers are VOCs.
Most U.S. cottonseed oil extracting facilities would be major sources of
VOCs and would be covered by the requirements for ozone emissions and
attainment, unless they used a solvent that was not classified as a VOC
(e.g., acetone).
Most vegetable oil production facilities are major sources of particulate
matters (PM). Depending on the oilseed processed, PM emissions can be
0.10.3 lbs of total suspended particulate (TSP), which is about 50% PM10
(PM smaller than 10 microns) and less than 2% PM2.5 (PM smaller than
2.5 microns), per ton of seed processed. PM controls would also have to be
part of a facilitys federal and state permits. Cottonseed oil production facil-
ities probably also have to include NOx, SOx, and CO emissions in their
federal and state permits.
Any new or significantly modified facility would have to comply with the
new source review (NSR) requirements. NSR is a preconstruction permitting
program. If new construction or making a major modification will increase
emissions by an amount large enough to trigger NSR requirements, the source
must obtain a permit before it can begin construction.
c. Odor: There are no specific federal regulations for odor. States can, however,
regulate odor if they choose to.
d. Federal Permits (40 CFR 70): All major sources of regulated solvents are
required to have federally enforceable operating permits (FOP) (137) (also
referred to as Title V permits).
e. State Permits: Most states require state permits for facilities that emit listed
air pollutants (119). In some states, federal permits and state permits are
combined, while in other states, facilities are required to have both a state or
county (air district) permit and a federal permit. As part of annual emission
inventory reporting requirements, many states already require reporting of
HAP and VOC because of their state implementation plan (SIP).
2. Clean Water Act (CWA; 33 U.S. Code 1251 et seq.). Under the CWA, the U.S.
EPA establishes water quality criteria used to develop water quality standards,
technology-based effluent limitation guidelines, and pretreatment standards
and has established a national permit program [National Pollution Discharge

Elimination System (NPDES) permits; 40 CFR 122] to regulate the discharge
of pollutants. The states have responsibility to develop water quality manage-
ment programs.
Oilseed processing and oil refining are covered by the following (138):
Basic discharge effluent limitations that require NPDES permits (40 CFR
122)
Stormwater regulations that require a stormwater permit (40 CFR 122 and
123)
Oil spill prevention and response plans (40 CFR 112)
3. Resource Conservation and Recovery Act (RCRA; 42 U.S. Code 6901 et seq.).
RCRA Subtitle D covers nonhazardous wastes. Subtitle C (40 CFR 261) is a
federal cradle-to-grave system to manage hazardous waste (including
provisions for cleaning up releases and setting statutory and regulatory
requirements). Materials or items are hazardous wastes if and when they
are discarded or intended to be discarded. Hazardous wastes are either listed
wastes (40 CFR 261.30-.33) or characteristic wastes (40 CFR 261.21-.24).
The U.S. EPA defines four characteristics for hazardous waste: ignitability
(40 CFR 260.21); corrosivity (40 CFR 260.22); reactivity (40 CFR 260.23);
and toxicity (40 CFR 260.24).
a. Spent bleaching clay is not a RCRA hazardous waste (40 CFR 302). It is
usually disposed of by taking it to a regular landfill. There sometimes can be a
spontaneous combustion (oxidation of unsaturated fatty acids in the retained
oil causing self heating leading to combustion) problem when it is taken to
the landfill. The potential for spontaneous combustion in bleaching earth
depends on the type and amount of oil retained and rises with increasing
unsaturation of the fatty acids in the retained oil. U.S. DOT classifies
materials liable to spontaneous combustion as Class 4.2 hazardous materials
[49 CFR 173.124 (b) and Appendix E 3]. Spent bleaching clay can be finely
ground and put in small quantities into the animal meal in operations that do
oil extraction. With the increased concern about dioxin in food and feed
product by FDA and EPA, this is discouraged.
b. Spent nickel catalyst is not considered a RCRA hazardous waste (40 CFR
table 302.4). No reporting of release of this substance is required if the
diameter of the solid metal released is equal to or exceeds 100 mm. The RQ
for particles greater than 100 mm is 100 lbs. Most, if not all, spent nickel
catalyst is recycled.
4. Emergency Planning and Community Right-to-Know Act (EPCRA; 42 U.S.
Code 11001 et seq.) EPCRA requires states to establish emergency planning
districts with local committees to devise plans for preventing and responding
to chemical spills and releases.
240 COTTONSEED OIL
Section 313 (40 CFR 372), Toxic Release Inventory (TRI): Businesses are required
to file annual reports with federal and state authorities of releases to air, water, and
land above a certain threshold for chemicals on the TRI/Section 313 list (40 CFR
372.65) by July 1 each year for the previous years releases (139). TRI requirements
are triggered if a facility is involved in manufacturing with 10 or more full-time
employees, manufactures, processes, or otherwise uses with one or more listed sub-
stances in a quantity above the statutory reporting threshold of 25,000 lbs./yr (man-
ufactured or processed) or 10,000 lbs./yr (otherwise used). Beginning with the 1991
reporting year, such facilities also must report pollution prevention and recycling
data for such chemicals pursuant to Section 6607 of the Pollution Prevention Act
(42 U.S. Code 13106). n-Hexane was added to the TRI list in 1994 with reporting
for 1995 emissions (135). Isohexane is not on the TRI list.
5. Toxic Substances Control Act (TSCA; 15 U.S. Code 2600 et seq.) If a
chemicals manufacture, processing, distribution, use, or disposal would
create unreasonable risks, the U.S. EPA, under the TSCA (40 CFR section
700, et seq.), can regulate it, ban it, or require additional testing.
Inventory Update Rule (IUR) (40 CFR 710). The IUR was established in 1986 to
require manufacturers and importers of chemicals listed on the master TCSA Inven-
tory to report current data every four years on the production volume of chemicals
imported or produced. Food and feed products produced from natural agricultural
product, such as oilseeds, are not required to be reported but all oil and meal
products obtained by solvent extraction that is sold for other than food or feed
use (e.g., oils as chemical raw materials and meal as fertilizer) are. Cottonseed
oil, soap stocks, acidulated soap stocks, deodorized distillates, hydrogenated
cottonseed oil are some of the substances reported by extraction and refining
operations under IUR. EPA amended this rule in 2003 (1/9/03; 68 FR 848). Cotton-
seed oil is on the list of partially exempt substances, which are not subject to the
new reporting requirements for processing and use data but continue to have
to report the current IUR information as well as manufacturing exposure-related
information.
5.3. Food Safety (FDA)

The U.S. Food and Drug Administration (FDA) regulates all aspects of food,
including food ingredients and labeling in the United States. An extraction solvent
or other processing substances that can be in food are subject to premarket approval
by FDA unless its use is generally recognized as safe (GRAS) by qualified experts.
Oilseed extraction solvents and food processing substances, to be legally used in the
United States, must have been subject to an approval by the U.S. FDA or the U.S.
Department of Agriculture (USDA) during 19381958 for this use (prior sanc-
tion); be GRAS for this use by FDA [GRAS affirmation (21 CFR 170.35); sub-
stances dont have to be specifically listed and there are several ways to determine
GRAS]; or be used in accordance with food additive regulations promulgated by the
U.S. FDA [21 CFR 170.3(h)(I)]. Food additives generally fall into two broad
categories: 1) those added directly to food (21 CFR 172), and 2) those that are
added indirectly to food through contact with packaging materials, processing
equipment, or other food-contact materials (21 CFR 174-178).
Many prior sanctions and early GRAS determinations are not codified in the
U.S. FDA regulations. Extracting solvents used in food manufacturing, such as
n-hexane or isohexane, have been labeled as food additives or incidental additives
(i.e., additives that are present in a food at insignificant levels and do not have any
technical or functional effect in that food). Incidental additives can be processing
aids, (i.e., substances that are added to a food during processing but removed
from the food before it is packaged). Most food-processing substances, including
solvents, can be regarded as incidental additives and, thus, are exempt from label
declaration in the finished food product.
As vegetable oil and other human food grade oils undergo refining, bleaching,
deodorization, (steam distillation) and sometimes other purification processes as
part of the manufacturing process prior to being used as a food product, they should
not contain any (< 100 ppb) of the extraction solvent, if proper manufacturing prac-
tices are followed.
Commercial hexane, containing about 5085% n-hexane (the rest is the hex-
ane isomers that make up commercial isohexane), has been mostly used since
the 1940s as an oilseed-extraction solvent on the determination that it is GRAS
and it may also be subject to a prior sanction. Like many other food-processing
substances, there is no U.S. FDA regulation specifically listing n-hexane as
GRAS or prior sanctioned. However, under FDA regulations, hexane has been
cleared as a solvent (residue not more that 5 ppm) for use in many products
(140). Isohexane/hexane isomers also are not specifically listed as GRAS or prior
sanctioned.
In Europe, the maximum residue limit (MRL) in vegetable oils has been
established as 5 ppm n-hexane (141). There is no MRL for isohexane/hexane
isomers.
In summary, extraction solvents can be considered incidental additives/proces-
sing aids that are exempt from label declaration. GRAS status may be determined
by a company or an industry (GRAS self-determination (142) or GRAS noti-
fication (April 17, 1997; 62 FR 18938)), an independent scientific organization
(e.g., FEMA GRAS (143), or the U.S. FDA (GRAS affirmation (21 CFR
170.35)). The Federal Food, Drug and Cosmetic Act (FFDCA; 21 U.S. Code 321
et seq.) does not provide for the U.S. FDA to approve all ingredients used in food,
and the U.S. FDA explicitly recognizes that its published GRAS list is not meant to
be a complete listing of all substances that are, in fact, GRAS food substances.
Although there is no requirement to inform the U.S. FDA of a GRAS self-determi-
nation or to request FDA review or approval on the matter, the U.S. FDA has estab-
lished a voluntary GRAS affirmation program under which such advice will be
provided by the agency. Solvents that do not have prior sanction, a GRAS determi-
nation of somekind, or a tolerance set probably should be evaluated for compliance
under food safety requirements if a facility is considering changing its extracting
solvent.
242 COTTONSEED OIL
6. COTTONSEED OIL PRODUCTS FINISHING TREATMENT
Not too many years ago, fats and oils finished products were only available pack-
aged in containers no larger than 55-gallon drums. Then, large industrial users of
cooking and salad oils that could pump oils at ambient temperatures installed bulk
handling facilities. These installations were followed by fats and oils handling,
which required equipment in addition to storage facilities; i.e., specifically equip-
ment for chilling and plasticization of shortening and margarine-type products. Ser-
ious consideration must be given to the means selected to deliver the product to the
consumer or ingredient user, be it shipped by rail or truck, or packaged before ship-
ment in smaller quantities. Quality, cost, and handling by both the fats and oils sup-
plier and the consumer are dramatically affected by the means selected. An oil
product that has been carefully processed to attain optimum palatability, nutrition,
and performance can be maintained or lost by the selection of packaging, crystal-
lization, cooling, filling, or bulk shipment choices.
6.1. Cooking and Salad Oil Filling and Packaging

The salad and cooking oil package contains and protects the oil on its journey from
the processor through usage by the consumer. Salad and cooking oils are usually
packaged shortly after deodorization in containers for home, restaurant, or food
processor use. It is not customary to store large quantities of deodorized oils; com-
mon practice is to package these oils as soon as possible. The preservation measures
necessary are nitrogen protection, temperature control, light avoidance, and the
addition of any additives required by product or customer specification. A cost-
effective package is one that delivers the product to the satisfaction of the user at
the least cost.
Early in the twentieth century, glass bottles replaced tin-plated cans in Europe
and the United States for cottonseed and other salad oils packaged for the house-
hold consumer. Glass packaging was well established as the efficient package for
retail consumer oils until the mid-1980s. Then, plastic replaced glass due to con-
sumer preference for lightweight, unbreakable, and easily handled containers with
added benefits of improved economics effected by lower package costs, reduced
transportation costs, and equalivent product protection. Consumer liquid oils in
the United States are currently packaged in 8-, 16-, 24-, 32-, 38-, 48-, and 64-ounce
clear, rectangular shaped, stretch-blown polyethylene terephthalates (PET) con-
tainers, as well as 1-gallon opaque plastic containers. The same oils and some
additional products with additives, such as antifoamers and antioxidants, are
packaged in 35-pound or 5-gallon plastic jugs and 425-pound closed head drums
for food service and food processor customers. High-speed automatic filling,
capping, labeling, and casing machines are used for packaging the smaller con-
tainers. Slower speed twin spout automatic fillers are normally used for the steel
drums.
COTTONSEED OIL PRODUCTS FINISHING TREATMENT 243
TABLE 24. Crystallization Process Conditions Influence

on Shortening Consistency.
Plastic Process Brittle
13 1% Gas incorporation None

Cold Chilling Warm
More Mechanical working Less
High Pressure Low
85 5 F Tempering Cold
6.2. Shortening Plasticization, Packaging, and Tempering

Plasticized shortening products can be defined as fats with a consistency than can
be readily spread, mixed, or worked. Considerably more is involved in the plasti-
cization of shortening and margarine than merely lowering the temperature to cause
solidification. Slow cooling of these products produces a grainy, pasty, nonuniform
mushy product that lacks the appearance, texture, and functional characteristics
associated with plasticized products. Development of these characteristics is a func-
tion of controlled crystallization or plasticization. The final consistency of a short-
ening is the culmination of all the factors influencing crystallization and
plasticization: chilling, working, tempering, pressure, and gas incorporation. Table 24
summarizes the effects of the crystallization processes upon shortening consistency
(144).
The plasticization process involves the rapid chilling and homogenization of the
shortening mixture. Most shortenings are quick-chilled in closed thin-film scraped-
wall heat exchangers with extrusion valves to deliver a smooth homogenous pro-
duct to the package at 1727 atm pressure. Nitrogen is injected at 13 1% into
most shortenings to increase the products workability and provide a white, creamy
appearance. After packaging, many processors temper shortenings at temperatures
slightly above the packaging temperature to allow the crystal structure of the hard
fraction to reach equilibrium and form a stable matrix. After tempering, shortenings
are usually stored and shipped at controlled temperatures of 7080 F or 21.1
26.7 C to avoid crystal change and loss of the plastic properties (145). Figure 4
depicts a typical shortening plasticization process flowsheet in the United States,
beginning when the deodorized shortening blend has been transferred to the chilling
unit supply tank and all of the specified additives have been incorporated (106).
Packaging of Shortening. Plasticized shortenings for commercial use are usually
packaged in plastic film-lined 50-pound corrugated cartons, commonly referred to
as a cube. The 50- and 110-pound tinplated cans and the 380400-pound open head
drums, both steel and composite, that were favorite packages for many years have
almost completely disappeared. For home consumption, vegetable shortening is
packed in 1- and 3-pound composite or plastic cans, and occasionally in larger
8-pound tinplated lithographed cans with a recloseable pressure-applied lid.
244 COTTONSEED OIL
Figure 4. Typical process flow for shortening plasticization.
The 50-pound cube containers are generally filled directly from the texturing
valve on scales and automatically filled to the package declared weight. Household
shortenings are usually packed in the smaller containers using volumetric type fill-
ing machines. The seamed lid applied on tinplated household shortening cans has
been replaced with a foil seal and a plastic overcap on the composite and plastic
COTTONSEED OIL PRODUCTS FINISHING TREATMENT 245
containers. After filling, all of the shortening products are tempered to stabilize the
crystal structure in the desired form.
6.3. Margarine Mixing, Chilling, and Packaging

Margarine is a flavored food product containing 80% fat, made by blending selected
basestocks with other ingredients, fortified with vitamin A, to produce a table,
cooking or baking fat product that serves the purpose of dairy butter but is different
in composition and can be varied for different applications (146). Margarine was
developed and continues to be a butter substitute but now, spreads have been devel-
oped as margarine substitutes. The major difference between spreads and margarine
is that spreads are not legally required to contain a minimum of 80% fat.
Processing for margarines and spreads begins with the preparation of an emul-
sion of the ingredients. Emulsions are prepared by adding the oil-soluble ingredi-
ents to a heated margarine oil formulation in an agitated emulsion tank.
Concurrently, a pasteurized aqueous phase is prepared by mixing all of the
water-soluble ingredients together in another vat. The water phase is then added
to the oil phase to make the emulsion. The emulsion is rapidly chilled with
scraped-wall heat exchangers similar to those used for shortening products. The
plasticized products are then formed into convenient sizes and shapes wrapped
with paper, which sometimes are called prints, or filled into the various containers
for consumer, restaurant, or food processor use. Most margarine and spread pro-
ducts are stored at refrigerator temperatures immediately after packaging, except
for some specialized baking products (106).
6.4. Flaking and Spray Chilling

Fat flakes describe the higher melting fat and oil products solidified in a thin flake
form for ease of handling, quick remelting, or for a specific function in a food pro-
duct. Flakes are solidified on a chill roll, which has been described as an endless
moving chilling surface held at a temperature below the crystallization point of the
applied fat or oil product to form a congealed film on the outer surface. Chill rolls
and processed oil formulations have been adapted to produce several different
flaked products that can provide distinctive performance characteristics in specialty
formulated foods. The flaked products, produced almost exclusively for restaurant
and food processor consumers, are hardfats or stearines, shortening chips, icing sta-
bilizers, confectioners fats, hard emulsifiers, and other customer specific products.
The flake products are packaged in kraft bags, corrugated cartons with vinyl liners,
or other suitable containers for storage and shipment (145).
Spray chilled or powdered fats are specialized products developed for ease of
incorporation, handling, melting efficiency, uniform delivery with addition systems,
encapsulation, and other special-purpose uses. The spray chilling process consists
of atomizing a molten fat in a crystallization zone or tower, maintained under tem-
perature conditions where a very fine mist of the melted fat is contacted with cooled
air or gas to cause crystallization without marked supercooling (106).
246 COTTONSEED OIL
6.5. Bulk Fats and Oils Shipments

Food processors that use fats and oils in large quantities often have the facilities to
handle this ingredient liquid in bulk. All of the products packaged for shipment and
use can be provided to the customers in tank cars or tank trucks, except margarine
and spread mixes that contain milk and salt. The customers for these bulk products
must have fats and oils bulk handling systems to receive, store, and handle the
liquid products. Chilling equipment is also necessary for the shortening or margar-
ine type products to develop the crystal structure necessary for functionality.
7. COTTONSEED OIL UTILIZATION
Edible fats and oils usage can be separated into four product categories: (1) liquid
oils, (2) shortenings, (3) margarines, and (4) specialty products. Until World War II,
cottonseed oil had been the major oil source for all four food product categories in
the United States. The U.S. vegetable oil industry was developed with cottonseed
oil as the original source oil, and it dominated this market for almost 100 years.
Many of the prepared food products available today were developed with a short-
ening, margarine, or an oil product containing cottonseed oil. Soybean oil became
the dominate vegetable oil in the world as a result of availability and economics, not
overall performance. Unlike other source oils, cottonseed oil is a more universal
source oil to provide the desired functionality for most products; in fact, cottonseed
oil or another beta-prime crystal former is necessary in most product formula-
tions with soybean, sunflower, canola, and corn oils to effect a smooth, plastic
consistency.
Cottonseed oil is still used in all four of the fats and oils product categories, as
shown in Figure 5 (28). However, competition from other vegetable oils grown
domestically and imported has decreased the popularity of cottonseed oil from a
dominant position to that of a specialty oil, as indicated previously in Table 2. In
the United States, soybean oil is the undisputed volume leader for all of the fats and
oil products with 61.0% of the consumption in 2000. Collectively, the animal fats
still have the second highest consumption, but individually as lard, tallow, and but-
terfat, their usage is less than three individual vegetable oils; soybean, canola, and
corn oils. A reduced usage of animal fats along with an increased usage of liquid
vegetable oils indicates an effort on the part of the U.S. consumers to eat a healthier
diet. Consumers have reduced their usage of animal fats to avoid cholesterol and are
decreasing their shortening usage in favor of liquid oils to reduce saturated fats.
Margarine usage has also suffered due to consumers efforts to reduce visible fat
consumption. Consumer table spread popularity has shifted to low-fat spreads while
this product category is experiencing a volume decline.
7.1. Oil and Fat Products Requirements

Fats and oils are very versatile raw materials. Processors have developed methods to
make them even more useful to the food industry and analytical chemists have
COTTONSEED OIL UTILIZATION 247
Other
6.7%
Shortening
26.4%
Export
16.3%
Margarine
2.1
Liquid Oils
48.5%
Figure 5. U.S. Cottonseed oil utilization.
devised methods to qualify the products produced. Satisfactory performance of oil

products is dependent on several important elements that determine suitability. For-
mulators must identify the attributes that are important for each individual finished
product and then effectively use process modification capabilities to satisfy the pre-
pared foods requirements. Successful production of these products relies on the
manipulation of the basestock blends to produce suitable physical properties and
to prevent undesirable changes during and after processing. The important vegeta-
ble oil performance characteristics that must be considered and modified to fit the
finished products requirements are as follows:
Flavor. Generally, the flavor of a cottonseed or a substitute oil product should

be completely bland to enhance the food products flavor rather than
contribute a flavor.
Flavor Stability. The oil ingredient must have the identified degree of resis-
tance to oxidative and lipolytic flavor degradation to maintain a bland flavor
and odor throughout the shelf life and use life of the prepared food product.
Cottonseed oil reverts to a nutty flavor that is not as objectionable as the rever-
sion flavors of the other oils with high levels of linoleic and oleic fatty acids.
Physical Characteristics. The proper blend of basestocks and hardstocks must
be identified that provide the physical, functional, and organolepic properties
required by the finished product. These characteristics are usually identified
analytically with solids fat index, melting point, and iodine value determina-
tions.
Crystallization. Cottonseed oil crystallizes in the beta-prime crystal form,
which is preferred for plasticity, heat resistance, and creaming properties in
248 COTTONSEED OIL
shortening, margarine, and most of the specialized product formulations. This

characteristic provides a more universal functionality for cottonseed oil; it can
be the sole source oil in a product or provide the impetus to force the crystal
habit to the desired beta-prime form in blends with other oils that crystallize in
the beta form.
Nutritional Concerns. All fats and oils are recognized as important nutrients
for both humans and animals because they provide a concentrated source of
energy, contain essential fatty acids, and serve as carriers for fat-soluble
vitamins. However, significant research has been done on the relationship
between fats and the incidence of coronary heart disease. Diet modifications,
including reductions in fat consumption, saturated fats, cholesterol, and trans-
isomers, have been recommended. Cottonseed oil is a good source of the
essential fatty acids because it contains 53% to 55% linoleic fatty acid. Like
other vegetable oils, it also has the advantage of being essentially cholesterol
free while providing the highest saturated fatty acid content (26%) of the other
vegetable oils in its classification. Unhardened cottonseed oil with a typical
iodine value of 111 can be used in some formulations to provide saturation
and oxidative stability without the trans-isomers developed in other oils
during hydrogenation to reach this point.
Current chemical and physical processing techniques provide the processor with
the capability of modifying one or more of an oils properties. It is possible to
change the functionality of an oil product to provide the ability to formulate tailor-
made products to suit a particular product or process. Further, the processing
techniques provide the processor with a wider range of alternative raw material
sources to improve commercial viability. The main objectives for applying the
available modification or processing techniques are as follows:
1. Produce an oil product to meet certain performance characteristics not

possible with natural oils.
2. Production of basestocks with lower processing costs to duplicate the
functionality of a more costly alternative.
3. Oxidation stability improvements through the elimination of the reaction
sites.
4. Palatability improvement.
5. More nutritionally acceptable products; i.e., reduce saturates and trans-acids
while increasing polyunsaturates.
8. LIQUID OILS
Clarity at or below ambient room temperature is the primary characteristic of a

liquid oil. Natural vegetable oils that are liquid at room temperatures in temperate
climates, 75 5 F or 23.9 2.8 C, contain high levels of unsaturated fatty acids
LIQUID OILS 249
with low melting points. Fatty acids with one or more double bonds and 18 carbon
atoms are the most important unsaturated fatty acids for liquid oils. Oleic (C-18:1),
a monounsaturated fatty acid, is the most widely distributed and most stable 18 car-
bon chain-length unsaturated fatty acid. Linoleic (C-18:2) and linolenic (C-18:3)
are the most widely distributed di- and triunsaturated fatty acids. Both of these
polyunsaturated fatty acids are termed essential because they cannot be synthesized
by the human body and must be supplied in the diet. Complete exclusion of the
essential fatty acids results in scaly skin, loss of weight, kidney lesions, and death.
The physical state of a fats and oils product may be natural or the result of pro-
cessing. Changes to either a solid or a liquid can be effected by processes that
change the melting or solidification points. These physical properties can be mod-
ified by several different processes available to the fats and oils industry: (1) hydro-
genation, which saturates the unsaturated fatty acids to increase both the oxidative
stability and melting point; (2) interesterification, which rearranges the fatty acids
among the triglycerides to change the melting, oxidative stability, and crystalliza-
tion tendencies of the oil; (3) fractionation, which physically separates hard and soft
fractions to provide different functional properties; and (4) genetic engineering
procedures, which modify the oils fatty acid composition during the development
of the seed.
Three major oil types have been developed that maintain different degrees of
clarity or oxidative stability at and below room temperature in temperate climates.
These liquid oils are identified by their functionality traits; cooking, salad, and high
stability. The definition for each of these classifications is as follows:
Cooking Oil. An edible oil that is liquid and clear at room temperature, 75 F or
23.9 C, that may be used for cooking. Cooking oils are typically used for pan fry-
ing, deep-fat frying, sauces, gravies, marinates, and other nonrefrigerated food pre-
parations where a clear liquid oil has application. Cooking oils usually congeal or
solidify at refrigerator temperatures. Refined, bleached, and deodorized cottonseed
oil is considered a cooking oil because it contains saturated fatty acids that cause
the oil to cloud and solidify slightly below room temperature.
Salad Oil. An edible oil that is suitable for the production of a mayonnaise or salad
dressing emulsion, which will remain liquid at refrigerated temperatures; 40 F or
4.4 C. This requirement has been refined to require a minimum cold test of 5 and
one-half hours; this measurement requires that a sample of the oil remain clear and
brilliant while submerged in an ice bath. Cottonseed oil must have a portion of the
saturated triglycerides removed to meet the requirements of a salad oil; winteriza-
tion is the process normally used to produce a cottonseed salad oil.
High Stability Oils. An edible oil that is clear at room temperature and possesses
an exceptional oxidative and flavor stability. The usual measure for oxidative
stability is the Active Oxygen Method (AOM), AOCS Method Cd 12-57. High-
stability oils will withstand the AOM abuse for periods in excess of 75 hours,
and some longer than 300 hours, as opposed to a 15-hour AOM result for cotton-
seed cooking oils. High-stability oils can be produced by hydrogenation followed
by fractionation or by genetic engineering. Most high-stability oils are considered
specialty products.
250 COTTONSEED OIL
TABLE 25. U.S. Source Oil Consumption as Salad and Cooking Oils.
Year (Metric Ton)
Oil Source 1960 1970 1980 1990 2000
Canola W 233.6
Corn 112.0 112.0 158.8 288.5 227.7
Cottonseed 341.1 239.0 208.7 208.7 137.9
Olive 23.1 28.1 26.3 96.6 208.2
Peanut 12.7 63.5 67.1 63.0 W
Safflower 5.4
Soybean 402.3 1120.8 1833.4 2114.7 3338.9
Sunflower NR NR NR
Unidentified 0.5 2.3 49.4 15.0 23.1
Total 891.8 1571.3 2343.7 2786.4 4169.5
kg per person 4.2 7.8 9.7 11.0 15.3
W withheld; NR not recorded.
8.1. Cooking and Salad Oil Sources

A steady growth in the consumption of cooking and salad oils is evident from the
USDA Economic Research Service Oil Crops Situation and Outlook Reports for
domestic consumption of salad and cooking oils in the United States. The consump-
tion data by source oil is summarized in Table 25 (25, 33). Deodorized cooking and
salad oils are principally prepared from soybean, cottonseed, corn, canola, sun-
flower, and peanut oils. Olive oil is technically a cooking oil and is considered a
gourmet product by many due to its distinctive flavor and odor, which would be
destroyed by deodorization, considered mandatory for the other liquid oils.
Cottonseed oil contains only minor amounts of linolenic fatty acid and a fair
amount of saturates, approximately 26% to 27%, to make it a relatively stable cook-
ing oil. Cottonseed salad oil production requires winterization to remove the stear-
ine fraction caused by the high level of saturated fatty acids. The unsaturates are
53.3% linoleic fatty acid and 18.3% oleic fatty acid. The natural nutty or buttery
flavor is an additional benefit that can mask less desirable flavors in food products.
Cottonseed oil was the principal cooking and salad oil used in the United States
until the late 1950s. As a source oil, it had lost its dominance for usage in short-
enings and margarines in the 19401950 era, but remained the preferred liquid oil
due to the flavor problems associated with soybean oil. Demand for cottonseed salad
and cooking oil decreased both in use percentage and amount after technology
was developed to overcome the soybean oil flavor problems. The actual pounds of
cottonseed oil used domestically for salad and cooking oils have decreased 60%
since 1960, and the market portion has decreased from 38.25% in 1960 to 3.3%
in 2000. One of the factors that led to the decreased consumption in the United
States was the lucrative export market for cottonseed oil. Export demand provided
an outlet for a high percentage of the available cottonseed oil at good income levels.
These exports helped to maintain the premium pricing for cottonseed oil marketed
domestically. Currently, a major usage for liquid cottonseed oils is as a cooking oil
LIQUID OILS 251
for snack foods. The unique cottonseed oil flavor imparts a pleasant, stable flavor to
potato chips and other salty snacks unattainable with other frying oils that lack the
nutty fried flavor note.
8.2. Liquid Oil Markets

Liquid oils enter into three major areas of food preparation: (1) retail, (2) foodser-
vice, and (3) food processor. For some uses, there is a similarity in product require-
ments, but there are also major differences in performance requirements in the three
areas. Package sizes are obviously different, with the smallest designed for home
use and the largest being the food processor shipments of 150180 thousand pound
tank cars. The retail market consists of bottled oils sold in grocery stores or other
retail outlets for home use. The foodservice industry is composed of restaurants,
hotels, institutions, and other mass feeding operations. In many cases, foodservice
oils are specialized products designed for the intended use; i.e., frying, salads,
sauces, etc. The food processor market is made up of prepared food manufactures
of products sold through retail outlets or used by the foodservice industry. The food
processor oils are normally specifically designed for the individual operation to
provide the desired performance for the product as well as for the process.
Retail Consumer Oils. None of the retail consumer oils currently available have
been chemically modified to change their physical characteristics. Processing for
the majority of the household bottled oils consists of refining, bleaching, and deo-
dorization with additional dewaxing required for canola, corn, safflower, and sun-
flower oils and winterization for cottonseed oil to meet the cold test requirements
for a salad oil. Soybean oil is a natural winter oil that meets the requirements of a
salad oil and requires only refining, bleaching, and deodorization processing. Pea-
nut oil gels, when chilled to the extent that it cannot be separated, into an oil and
hard fraction with winterization to produce a salad oil. Therefore, both peanut and
olive consumer bottled oils will become a semisolid at refrigerator temperatures.
The broad dietary shift from animal fats and then saturated fats has favored
liquid oil products. The U.S. consumer has become increasingly aware of the
fats and oils role in coronary heart disease. As a result, consumers have replaced
solid shortenings with liquid oils. The liquid oils favored by consumers tend to
reflect the findings of the most recent publicized study. Initially, when it was
thought that only polyunsaturated oils were useful in lowering serum cholesterol
corn oil began its rise in popularity. Reports that monounsaturates were equal to
polyunsaturates in lowering serum cholesterol appeared to help peanut and olive
oil sales. Later, the Mediterranean diet helped to improve the sales of olive
oil. Canola oil introductions capitalized on the low saturate level; i.e., 94% saturates
free. In 1991, an oil composed of soybean, sunflower, and canola oils blended to
one-third the saturate level of soybean oil was introduced (147). Blended retail
oils are still marketed by the major branded suppliers but with different composi-
tions. One blended oil is composed of soybean and canola oils with 7% saturates
and another is a blend of corn and canola oils with 15% saturates. The advent of trans-
fatty-acid labeling requirements is shifting consumer attention to that parameter.
252 COTTONSEED OIL
Throughout all these changes, cottonseed oil has not fared well. It has dropped from
the leading consumer bottled oil to almost become a gourmet type oil in most retail
grocery stores sharing space with exotic oils like safflower, avocado, walnut, and grape
seed oils. The major reasons for this decreased presence are probably the high level of
saturates, the winterization requirement for a clear and brilliant salad oil, and the high-
er cost that cottonseed oil commands.
Foodservice Liquid Oils. The foodservice industry consists of restaurants, hotels,
and institutions where food is prepared for in-house service or delivery to the con-
sumer. This industry had an inflation adjusted growth rate of 2.6% per year in con-
trast to a retail food sales growth rate of 0.7%. The fast-food segment of this
industry continues to represent the largest and fastest rising share of sales among
separate eating places (148). Edible fats and oils are major foodservice ingredients,
which includes all of the liquid oil types: cooking, salad, and high-stability oils.
Foodservice kitchens may use liquid oils for a number of different applications,
such as deep-fat frying, pan frying, grilling, pan release, seasoning, preparation of
salad dressings, sauces, gravies, some types of baking, and other food applications.
Vegetable oils became the leading U.S. foodservice frying media in 1990 when
the major fast-food restaurants switched from tallow to vegetable oils for frying
french fries. The vegetable oil products used for deep-fat frying range from hydro-
genated plasticized frying shortenings, to cooking or RBD oils. RBD liquid oils, in
most cases, have the lowest initial cost but also have the poorest frying stability of
all the available products. In general, cooking oils will perform satisfactorily when
the fried food volume is extremely high and an oily product appearance is desired.
However, nature did not create all of the vegetable oils alike; some source oils have
a better frying stability than others and, in some cases, a particular oil type is cho-
sen for the distinctive flavor that the slightly reverted oil contributes.
The major criteria for comparing the frying stability of different oils is the level
and type of unsaturates and the fried flavor characteristics. Oxygen combines with
the unsaturated fatty acids, causing oxidation and a flavor reversion back to the ori-
ginal characteristics. The rate that an oil oxidizes is dependent on the amount and
type of the unsaturated fatty acids. Frying temperatures accellerate the oxidation
reaction, but in comparison testing, all of the cooking oils would be treated the
same. In general, the more unsaturated the fatty acid, the faster its rate of oxidation.
Natural vegetable oils that are liquid at room temperature generally contain high
levels of polyunsaturated fatty acids with low melting points. These polyunsatu-
rated fatty acids, predominately linoleic (C18:2) and linolenic (C18:3), tend to be
very liquid but are also highly susceptible to oxidation and, therefore exhibit a poor
frying life in foodservice applications. Oxidation precedes polymerization in the
frying oil deterioration cycle; polymerization increases an oils viscosity, which
eventually results in persistent foaming after the oil thickens to the point that it
will not allow the moisture from the food to escape. Table 26 compares the levels
of unsaturates, the inherent oxidative stability, and the oxidized or reverted flavors
for the most common cooking oils used in the United States. Cottonseed oil has a
median stability rating but has the advantage of a somewhat pleasant, nutty oxi-
dized flavor. Soybean cooking oil has the poorest stability rating and develops an
LIQUID OILS 253
TABLE 26. Frying Performance Considerations for Cooking Oils.
Fatty Acid Composition, % Inherent

Oxidative Fried
Oil Source Saturates Oleic Linoleic Linolenic Stability Flavor
Peanut oil 19.4 46.7 32.0 3.7 Peanut

Canola oil 7.3 60.9 21.0 8.8 4.9 Earthy
Cottonseed oil 25.7 18.6 54.4 0.7 5.8 Nutty
Corn oil 13.6 25.4 59.6 1.2 6.5 Musty
Sunflower oil 12.8 18.7 67.5 0.8 7.1 Nutty
Soybean oil 15.3 23.3 53.7 7.6 7.5 Fishy
objectionable oxidized flavor and odor; even with these characteristics, it can be
utilized in high-use situations that do not require a long shelf life.
Another clear liquid frying oil type available to the foodservice operator is the
high-stability frying oils. The term high-stability oils describes triglyceride mix-
tures that are both liquid and stable in the temperature range of 6085 F, or
15.629.4 C. In nature, this combination rarely occurs. Natural oils that are liquid
under these conditions are generally much less stable with respect to frying stability
than solid fats. The oxidative/frying stability of vegetable oils, as referred to above,
depends on a combination of factors, including the following: (1) degree of unsa-
turation or number of double bonds; (2) nature of unsaturation, or the position of
the double bonds; (3) pro-oxidant or trace metals content; (4) antioxidant content;
and (5) exposure to oxygen. Two entirely different techniques have been developed
to enhance stability while maintaining liquidity, these being processing or plant
breeding. Using a combination of hydrogenation and fractionation techniques, it
is possible to produce oils that are both stable and liquid from otherwise unstable
polyunsaturated raw materials. Plant breeding technology has developed geneti-
cally superior varieties of oilseeds that yield oils more-suitable for high temperature
frying than the conventional vegetable oils. The high-stability oils, both those pro-
duced with processing or plant breeding, have excellent frying stabilities, which
approach the stability of a hydrogenated plasticized shortening. However, high-
stability oils produced with either technology have not become as popular as the
solid or liquid frying shortenings, probably due to higher product costs.
Most of the foodservice liquid cooking and high-stability oils designed for frying
are stabilized with dimethylpolysiloxane, a combination antifoaming and antioxi-
dant agent added to extend the frying stability. It can extend frying life by three
to ten times, with the degree of increase dependent on the original frying life with-
out the antifoamer; the more stable products initially show greater increases in fry-
ing life than the less stable products.
Another primary use for a salad and even cooking oil in most foodservice opera-
tions is for making salad dressings. Cooking oils are generally not satisfactory for
salad dressing use in industrial preparations, but can be used in foodservice opera-
tions where the salad dressing is prepared fresh and is not expected to withstand
prolonged storage periods at refrigerated temperatures. In making salad dressings,
254 COTTONSEED OIL
ease of blending with other ingredients is quite often an advantage. Most salad
dressings depend on the oil to coat the salad ingredients and to hold the flavors
of herbs, spices, and vinegar.
Other uses for salad oils include pan frying, griddling, sauces, gravies, dips, and
some types of baking. When liquid oils are used in baking, the general use is to
make chiffon or sponge type cakes and other all-purpose baking applications where
a lubricant is beneficial.
Food Processor Liquid Oils. Food processors often purchase their fats and oils
ingredients in bulk quantities to obtain a cost advantage generated by elimination
of packaging, lower shipping rates, and reduced labor costs. Liquid oils offer a
definite advantage for bulk handling as a result of their absence of solid fractions at
ambient temperatures and below. A major concern with fats and oils bulk handling
is that the product may deteriorate before use. This is particularly true of shorten-
ings that must be held in a molten state. Melted products are generally more sus-
ceptible to deterioration than packaged products or liquid oils, which are normally
fluid and pumpable at ordinary ambient temperatures. Autoxidation increases mark-
edly with heating. It has been determined that the rate of oxidation doubles for each
25 F or 15 C increase in temperature within the range of 70 F to 140 F or 20 C to
60 C. Exclusion of oxygen during storage represents an effective method for limit-
ing quality deterioration for all types of fats and oils. The usual procedures replace
oxygen with nitrogen during transit and storage before use.
All three of the liquid oil types (salad, cooking, and high-stability) are utilized
by food processors as ingredients for specific product applications. The choice of an
oil is the result of a synthesis of many parameters, which explain the variety of cho-
sen solutions. Liquid oils offer a choice of cooking, salad, or high-stability oil types
and functional variations occur within each regarding flavor, degree of unsaturation,
fatty acid distribution, essential fatty acids, oxidative stability, frying stability, cost,
availability, etc. (132).
Food Processors utilize cooking oils predominately for frying snack foods, nuts,
fish, poultry, meats, potatoes, and other food products for dry, refrigerated, and fro-
zen distribution to retail and foodservice consumers. Food processors must evaluate
the contributions of each cooking fat or oil to their product and identify the most
suitable ingredient for use. One role of fat in cooking or frying is essentially to pro-
vide an efficient heat-transfer medium, transmitting heat rapidly and uniformly to
the surface of the food being cooked. Additionally, the oil contributes flavor and
palatability to the food fried. Every fat and oil product will experience reversion
to characteristic flavor, which may or may not complement the food product.
Many potato chip manufacturers use cottonseed cooking oil for frying their product
because of its stability compared with other oils and the characteristic nutty flavor
developed with oxidation.
Cooking and frying oils are used at high temperatures, often in the presence of
hydrolyzing conditions, namely water and steam. Hydrolysis causes free fatty acid
development, which results in more acidic flavors. Some products that are eaten
shortly after preparation do not need as stable a frying oil as other foods that are
packaged and require an atmospheric shelf life of several weeks. Products that
LIQUID OILS 255
remain frozen until the consumer thaws them do not require a high degree of
oxidative stability. Surface appearance of the fried food is also affected by the fry-
ing media; liquid oils provide a shiny, wet, soft appearance, whereas solid fats
impart a dry, somewhat dull appearance. Polymerization, either thermal or oxida-
tive, causes a thicker viscosity, which increases the oil absorption giving a greasier
product.
Successful use of liquid oils in baked products has been a technological advance-
ment for the baking industry. This technology involved the use of emulsifiers to pro-
vide the functionality lacking with the liquid oils. Baked products prepared with
liquid oils alone had low volume, fair softness, and poor grain; plastic fats provided
dough strength, gas retention, aeration, volume, symmetry, fine grain, and even
texture. It was found that the use of proper emulsifier systems with liquid oils pro-
vided these functions with some added benefits. The liquid oils allowed the bakers
to merchandise their products with all-vegetable and polyunsaturated label-
ing. Cooking oils are appropriate for these products because refrigeration is not
an issue and the shelf life of baked products is well within the oxidative stability
of polyunsaturated oils.
Pourable and spoonable salad dressings, prepared for retail sale or foodservice
outlets, are a major use for food processor salad oils. Pourable salad dressings may
be prepared as an emulsified form or a two-phase system of oil and water such as
Italian style. As most dressings are stored in the refrigerator after being opened, it is
important to use a salad oil to resist clouding at these temperature conditions. Aside
from the unattractiveness of a cloudy product when refrigerated, it is necessary that
no solid crystals be present that would give a waxy, tallowy tasting sensation in the
mouth. Homogenized, pourable salad dressings require the use of noncrystallizing
salad oils to prevent separation during refrigeration.
Mayonnaise and spoonable salad dressings are another major food processor use
for salad oils. Oil constitutes 80% of most mayonnaise formulations and is respon-
sible for the body and viscosity of the product. Spoonable salad dressings have only
35% to 50% oil whose function is to modify the mouthfeel of the starch paste that
imparts the body. In both cases, a smooth, creamy, nonoily mouthfeel is desired that
will not occur if crystallization occurs. These emulsions are very unstable and the
presence of fat crystals will break the emulsion, rapidly causing oil pockets to form
(128).
Food Processor High Stability Oils. The primary characteristics of a high-
stability oil are liquidity at ambient temperatures and resistance to oxidation. Most
oils that are liquid at room temperature contain high polyunsaturated fatty acid
levels, which are the most susceptible to oxidation. Two technologies have been
developed to enhance the stability of liquid oils while retaining the functional
and nutritional properties. The first technology involves processing with hydroge-
nation to change the polyunsaturates to monounsaturated fatty acids followed by
fractionation to separate the stearine, or hard fractions, formed during hydrogena-
tion from the olein, or soft fraction. The olein fraction, high in oleic fatty acids,
becomes the high-stability oil. The alternate technology for producing high-
stability oil is the use of biotechnology and plant breeding techniques to produce
256 COTTONSEED OIL
oilseeds with low levels of polyunsaturates and saturates but high levels of
monounsaturates.
High-stability oils produced by either technology can be used by food processors
wherever liquidity and oxidative stability influence the finished products quality or
handling conditions. The identified functionalities of the monounsaturated oils in
specific applications are as follows (106):
Frying oilsThe high proportion of oleic fatty acid increases frying stability
by limiting the opportunities for oxidation and polymerization. High-stability
oils have exhibited frying stability results close to the selectively hydroge-
nated heavy-duty frying shortenings.
Spray oilsWhen applied to the surface of food products, the high-stability
oils protect the product from moisture and oxygen invasion, prevent clumping,
and impart a glossy appearance. Specific applications include raisins and other
fruits, breakfast cereals, nut meats, croutons, bread crumbs, spices, season-
ings, crackers, and others.
Pan release agentsAerosol and brushed lubricants for cooking skillets,
baking pans, confectionery products, and other food products where liquidity
and oxidative stability are beneficial.
Product carrierColors, spices, flavors, vitamins, and other additives may be
dispersed in high-stability oils to preserve the flavor, color, or activity during
extended shelf life periods with fluidity.
Product compatibilityHigh-stability oils are compatible with all other fats
and oils products because crystal type is not a concern. Liquid oils do not have
a crystal structure.
9. SHORTENINGS
Shortening is an American invention developed with cottonseed oil to replace lard,

the solid fat of choice. Cottonseed oil was plentiful and inexpensive in the latter
decades of the nineteenth century as a direct result of the growth of the cotton
industry in the United States. As the shortening product category developed, it
expanded from a limited application to include baked and other product types. In
1948, H. C. Black defined shortening as a semisolid plastic material made wholly
from fats and oils for use in cooking, baking, and frying (17). Today, shortening has
become virtually synonymous with fat to include many other edible fats or oils pro-
ducts designed for all prepared foods.
Initially, shortenings were produced to resemble the consistency and plasticity of
lard. Now, shortenings are designed to satisfy individual specific requirements for
all of the food industry as well as offering products with broad general appeal. With
this broader application, shortening consistency varies from wide workable ranges
to brittle products with sharp melting characteristics, from very firm consistencies
to liquid or pumpable products, or from creamy, smooth textures to grainy struc-
tures depending on the requirements of the application.
SHORTENINGS 257
In most cases, products identified as shortening are 100% fat; however, there are
exceptions, such as the roll-in shortenings, which may contain moisture. In some
cases, a fat or oil system may be identified as a shortening to distinguish it from
a margarine product. Generally, in the United States, if a fat product contains at
least 80% fat and has the required Vitamin A content, it is a margarine. Products
that resemble margarine but contain less than 80% and not more than 40% fat are
required to be labeled as spreads. Products that do not meet these criteria have been
identified as shortening since it does not have identity standards. Currently, a
description for shortening would be: processed edible fats and oils that affect flavor,
oxidative stability, shelf life quality, eating characteristics, nutrition, and the eye
appeal of prepared foods by providing emulsification, lubricity, structure, aeration,
moisture barrier, flavor medium, or heat transfer (20, 106, 149).
9.1. Shortening Attributes

Shortenings are functional ingredients that contribute heavily to the success of the
food product prepared. Shortenings can be tailor made for a specific food product
and process or designed as a general purpose product that must perform in varying
conditions, product types, processes, and formulations. Adequate performance of a
shortening for a food application is dependent on a number of factors. These
requirements differ for each customer or application dependent on the product for-
mulation, equipment, processing, preferences, and other considerations. Therefore,
performance is dependent on the interrelated elements that determine acceptability.
Flavor. Generally, the flavor of a shortening should be completely bland, so that it
can enhance the food products flavor rather than contribute its own flavor. The
bleaching and deodorization processes remove primary and secondary oxidation
products from shortenings. In some cases, shortening is the carrier for a flavor
desired in the finished product. Flavor additives to shortening products are usually
butter-like flavors; for example, butter flavors are incorporated into most pan and
grill shortenings. Also, the flavor of the source oil may be desired for a specific food
preparation; for example, the nutty fried-in flavor contributed by cottonseed oil.
Flavor Stability. The bland, typical, or formulated flavor must remain stable
throughout the shelf life and use life of the prepared food product. Reverted or oxi-
dized and hydrolyzed flavors and odors of most fats and oils are objectionable.
Shortenings must possess the identified degree of resistance to both oxidative and
lipolytic flavor degradation. Flavor stability is built into shortenings by selection of
saturated or monounsaturated components or processing to decrease the unsaturated
fatty acid content. The processes that influence flavor stability are hydrogenation,
fractionation, or interesterification. Flavor stability can be preserved after it has
been established with the use of antioxidants.
Physical Characteristics. Shortenings are pictured as solid materials but, in rea-
lity, are predominantly fluids. A plastic shortening consists of approximately one-
quarter crystalline-solid triglycerides suspended in liquid triglycerides. The ratio of
these two phases determines the consistency of a shortening as it relates to firmness,
softness, and spreadibility. Fats and oils processors change the consistency of a
258 COTTONSEED OIL
80.0
Liquid Bread
70.0 Non Dairy
Bakery All-Purpose
Solid Fat Index (SFI), %
60.0 Frying
Puff Pastery
50.0
Shortening Chips
40.0
30.0
20.0
10.0
0.0
10.0 15.6 21.1 26.7 33.3 40.0 43.3
Temperature, C
Figure 6. Shortening solid fat index profiles.
shortening by manipulating the solid-to-liquid: ratio over a range of temperatures.

Three processes capable of altering the consistency of a shortening component are
hydrogenation, fractionation, and interesterification. Fractionation processes make
physical separations of the hard and soft fractions. Interesterification rearranges the
position of the fatty acids in the triglycerides to affect melting characteristics.
Hydrogenation, the most popular process for modifying solid-to-liquid ratios, satu-
rates the unsaturated fatty acids to affect melting characteristics. The solid-to-liquid
proportions of the shortening components can be somewhat selectively altered with
these procedures to provide the physical characteristics most suited to the desired
functionality. Figure 6 compares the solids fat index curves for six shortenings,
designed for different applications, to illustrate changes in physical characteristics
that can be affected for shortenings. Solid fat index (SFI) analysis is an empirical
determination of when a fat passes from a solid to a liquid at the measurement tem-
peratures. A plot of the results produces a curve that illustrates the changes in solids
or liquid content to indicate the following:
ConsistencyThe factor most directly influencing the consistency of a

shortening is the proportion of the product in the solid phase. The consistency
of a shortening at use, preparation, and storage temperature conditions
materially affects the performance of the prepared food product. SFI curves
characterize the firmness of the shortening over a range of temperatures and
liquidity at mouth temperature for development, selection, and control purposes.
Plastic RangeShortenings are normally plastic and workable for SFI values
between 15 and 25. At an SFI value above 25, shortenings start to become
brittle and below 15, they become too fluid. The range of temperatures where
the SFI values remain within the 15% to 25% solids is usually referred to as
the plastic range.
SHORTENINGS 259
TABLE 27. Characteristics of the Polymorphic Forms of Monoacid Triacylglycerols.
Characteristics Alpha-Form Beta-Prime Form Beta Form
Chain packing hexagonal orthorhombic triclinic

Short spacing, (A) 4.2 3.8 and 4.2 4.6
Characteristic single band double at 727 single band
infrared spectrum at 720/cm and 719/cm at 717/cm
Density least dense intermediate most dense
Melting point lowest medium highest
Flat Solids CurvesSFI values that provide east to west (x-axis) directional
slopes generally have a better plasticity because the product is maintained at
the ideal consistency over a greater temperature range. This same attribute
contributes a slower get away in the mouth due to melting points above
body temperature caused by higher SFI contents over a greater temperature span.
Steep Solids CurvesPoor plasticity but good mouthfeel and oxidative
stability are usually indicated by a north to south (y-axis) directional SFI
slope. A narrower span of temperatures for the higher solids values provides
sharper melting tendencies for better palatability and good oxidative stability.
The preferred bland flavor is retained longer due to a lower unsaturated fatty
acid content. However, steep SFI curves from hydrogenated products usually
indicate a higher trans-fatty acid content.
Crystalline Behavior. Fats and oils are polymorphic, which means that, with cool-
ing, a series of increasingly organized crystal changes occur until a final crystal
form is achieved with no change in chemical structure. Typically, triglycerides exhi-
bit three major crystalline forms: alpha, beta-prime, and beta. Characteristics of
each are shown in Table 27 (150). The alpha crystal has a short life and readily
transforms into the beta-prime crystal form. Thereafter, each source oil has an
inherent crystallization tendency of either beta or beta-prime. The tiny, uniform,
tightly knit, needle-like beta-prime crystals produce smooth textured shortenings
with good plasticity, heat resistance, and good creaming properties. The large,
high-melting, self-occluding, course, stable beta crystals produce visibly grainy,
sandy, brittle shortenings that can experience separation of the liquid-oil portion.
Both of these crystal habits provide physical conditions desirable for particular
functionalities; for example, beta-prime for cake and icing shortenings requiring
good creaming or air entrapment capabilities and beta for pie crust shortenings
where a grainy consistency helps provide a flaky texture. The affinity of a fat
and oil product to form beta-prime crystals is dependent on several factors: (1)
the amount of palmitic fatty acid in the product, (2) the distribution and position
of both palmitic and stearic fatty acids on the triglyceride molecule, (3) the degree
of hydrogenation (beta-prime crystallization increases with the degree of hydroge-
nation and trans-isomers development), and (4) the degree of randomization of the
triglyceride (151). Crystal development is complemented by the plasticization con-
ditions employed: (1) chilling, which initiates the crystallization process, and (2)
tempering, where the desirable crystal nuclei are developed and stabilized.
260 COTTONSEED OIL
Appearance. Appearance is one of the collective results of all the processes used
to convert crude edible fats and oils into shortening: (1) refining, bleaching, and
deodorization are all involved in the removal of the color pigments; (2) hydrogena-
tion, fractionation, or interesterification affect the solid fat index (SFI), which helps
determine product consistency; (3) formulation determines crystal structure and SFI
content for texture and consistency control; (4) chilling and tempering initiate,
develop, and stabilize the crystal nuclei; and (5) storage and transportation under
proper conditions prevent product damage.
Nearly all plasticized shortenings contain 12% to 14% nitrogen by volume added
during the chilling process. The gas is finely dispersed to enhance creaming proper-
ties, control texture, and improve the appearance of the shortening. Shortenings
containing creaming gas are creamy white in contrast to a yellow or green cast
characteristic of petroleum jelly when no nitrogen is present.
Emulsification. Shortening emulsifying properties are accomplished by adjust-
ment of the fat structure or the addition of surface-active agents like mono- and
diglycerides. Food emulsifiers supplement and improve the functionality of a prop-
erly developed shortening to act as a lubricant, emulsify fat in batters, build struc-
ture, aerate, improve eating qualities, extend shelf life, modify crystal development,
improve antisticking properties, act as a dispersant, alter moisture retention, and
more. Obviously, no single surfactant can perform all of these different functions.
Emulsifier selection requires the same attention to functionality used to identify the
other components for shortenings.
Antioxidants. The oxidation rate for a shortening depends primarily on the num-
ber of double bonds and their arrangement in the triglyceride. However, the oxida-
tion process can be slowed down with the preservation of the natural antioxidants or
the addition of synthetic antioxidants. Antioxidants are chemical compounds that
delay the onset of oxidation. They function by inhibiting or interrupting the free
radical mechanism of autoxidation; they function as a free radical acceptor, thereby
terminating oxidation at the initial step. Tocopherols are the natural antioxidants
contained in most vegetable oils. Phenolic compounds that provide antioxidant
activities are BHA (Butylated Hydroxyanisole), BHT (Butylated Hydroxytoluene),
and TBHQ (Tertial Butyl Hydroxyquinone). Much of the success of the antioxi-
dants depends on their being in chemical contact with the product they are protect-
ing; therefore, various combinations of different antioxidants and chelating agents
are generally combined for use. These combinations provide a synergistic effect to
increase the antioxidants effectiveness and allow the incorporation of a chelating
agent to sequester prooxidant trace metals.
Trace Metals. Vegetable oils pick up metals from the soil. Many of the metals
picked up from where the plants are grown and later from contact during crushing,
processing, and storage promote oxidation, which results in off flavors, odors, and
off-color development. Studies have identified copper as the most harmful with
nickel, manganese, chromium, and iron following. Metal scavengers, added at
low levels during processing prior to filtration, facilitate removal of the harmful
metals. The most widely used chelating agent is citric acid, which is usually added
at a rate of 50100 parts per million (ppm).
SHORTENINGS 261
Foaming. Polymerization occurs during frying to produce three dimensional poly-

mers, which result in an increased frying shortening viscosity. Foam develops on
the surface when the frying shortening will no longer release moisture but keeps
it trapped as steam vapor. Polymerization can be inhibited with the use of dimethyl-
polysiloxane, which has been labeled as an antifoam agent. Dimethylpolysiloxane
may be safely used in processed foods at levels not exceeding 10 ppm. The effective
usage range to inhibit foaming for frying shortening has been identified as 0.5
2.0 ppm, which must be strictly controlled because concentrations above 10 ppm
will promote foaming. Additional control is also necessary to segregate this addi-
tives use for frying shortenings only; unintentional additions to bakery shortenings
can cause cake and icing failures, glazes may not adhere to donuts, and snack foods
may lack crispness.
Nutrition. Fats and oils are recognized as important nutrients for both humans
and animals because they provide a concentrated source of energy, contain essential
fatty acids, and serve as carriers for fat-soluble vitamins. Research studies have
also indicated a relationship between saturated fats, cholesterol, and trans-isomers
and the incidence of coronary heart disease. In many cases, shortening function-
ality can be maintained with formulations limiting cholesterol, the identified satu-
rated fatty acids, and trans-isomers. Shortening formulation can also aid in
reductions of fat consumption by development of more effective products to
reduce the levels required to produce the desired functionality and finished product
quality.
9.2. Shortening Formulation

Most shortenings are identified and formulated according to usage. Figure 6 illus-
trates the diverse SFI or solids-liquids relationships among different fats and oils
products. The SFI slopes indicate the differences in plastic range necessary to per-
form the desired function in the finished food products. Shortenings with the flattest
SFI slopes have the widest plastic range for workability at cool temperatures as well
as elevated temperatures. The all purpose and puff pastry shortenings in the illus-
tration have the widest plastic range of the plasticized or solidified products. The
frying and nondairy shortenings illustrated have relatively steep SFI slopes, which
will provide a firm, brittle consistency at room temperature but are fluid at only
slightly elevated temperatures. The very flat SFI slope represents a fluid opaque
liquid or pumpable shortening that has become popular due to the convenience
offered, handling cost savings, and lower saturated fatty acid levels. Liquid short-
ening systems can be produced with cottonseed oil but the most successful products
have been prepared with beta crystal forming hardfats, which produce a stable fluid
product with lesser amounts of processing. The most recent addition to the short-
ening classification is shortening chips. These specialty products are modifications
of fat flakes, which were previously limited to hardfats or almost fully saturated fats
and oils. Shortening chips are formulated with steep SFI contents to provide solid-
to-liquid ratios high enough to flake but low enough for good eating characteristics
after melting into the food product (106).
262 COTTONSEED OIL
Wide Plastic Range Shortenings. The basic all-purpose formulation has been the
building block for shortenings where creaming properties, a wide working range,
and heat tolerance are important. The functionality of an all-purpose product at
any temperature is largely a function of the solids content at that temperature.
The all-purpose products are formulated to be not too firm at 5060 F, or 10
15.6 C, and not too soft at 90100 F, or 32.237.8 C. Initially, a liquid oil was
blended with a hardfat to make a compound shortening that had a very flat SFI
curve. This blend provided an excellent plastic range similar to but slightly firmer
than the liquid shortening SFI curve illustrated on Figure 6. However, the low oxi-
dative stability of these shortenings precluded their use for most products requiring
a long shelf life. Currently, most of these products are formulated with a nonselec-
tively hydrogenated basestock and a low iodine value (IV) cottonseed oil hardstock.
Beta-prime (b0 ) hardfats are added to beta (b) crystal tending basestocks both to
extend the plastic range, which improves the tolerance to high temperatures, and
for crystal type and stability. The beta-prime (b0 ) crystal forming cottonseed oil
hardstock functions as a plasticizer for improved creaming properties, texture,
and consistency.
Hydrogenation of a fat and oil basestock increases the oxidative stability. As a
rule, the lower the base IV the better the oxidative stability. However, as base hard-
ness is increased, the level of hardstock required to reach a desired consistency
must decrease. Hardstock reduction reduces the plastic range and heat tolerance.
Therefore, oxidative stability improvements are achieved at the expense of plasti-
city and a wide plastic range can be at the expense of oxidative stability. The extent
that one attribute can be compromised to improve another must be determined by
the requirements of the intended food product. Oxidative stability is directly related
to the level and type of unsaturated fatty acids present; therefore, oxidative stability
results do not average. For example, a 50:50 blend of a 40-hour AOM basestock and
a 100-hour AOM basestock will not have an AOM stability of 70 hours, but rather
will be closer to the component with the lower AOM stability (152).
Plastic range is important for bakery use fats and oils products intended for roll-
in as well as creaming applications because of the consistency changes with tem-
perature. Fats and oils products become brittle above the plastic range and soft
below the range; both conditions adversely affect creaming and workability alike.
Shortenings are normally plastic and workable at SFI values between 15 and 25.
Therefore, products with flatter SFI slopes fall within the plasticity window for a
much greater temperature range than those with steep SFI slopes. The all-purpose
shortening in Figure 6 has a 40.7 F, or 22.6 C, plastic range from less than 50 F to
almost 92 F, or less than 10 C to almost 33.3 C; the frying shortening charted
would have an equalivent workability if it was used within the 5.8 F, or 3.2 C,
range from 84.290 F, or 2932.2 C. The frying shortenings use as a roll-in would
require very strict controlled temperature use probably not available in bakeries and
at a temperature detrimental to the laminated baked product. The 40.7 F, or 22.6 C,
working range for the all-purpose product is decidedly more practical.
The use of a partially hydrogenated base plus hardfat to produce a wide plastic
range with good creaming properties has been expanded into a whole family of
SHORTENINGS 263
TABLE 28. Wide Plastic Range Applications.
Nonemulsified Emulsified
All-purpose Household
Danish roll-In Cake & icing
Puff pastry Icing & filling
Cookie Cake mixes
Donut frying Specialty cakes
Yeast raised Yeast raised
specialized shortenings. The development of these products has involved the selec-
tion of the most suitable hydrogenated basestock and hardfat to produce the desired
plastic range and AOM stability. These developments have taken two directions: (1)
the addition of an emulsifier or an emulsifier system to an all-purpose fat and oil
base or (2) formulating nonemulsified products for a specific functionality. Table 28
outlines some of the current applications for these two categories (106).
Narrow Plastic Range Shortenings. Plasticity is of minor importance and can be
a detriment for products requiring sharp melting characteristics or a high oxidative
stability. Fat and oil products designed for specific frying situations, nondairy sys-
tems, cookie fillers, and confectionery fats require an eating character and flavor
stability not possible with blends of nonselectively hydrogenated oils with hardfats.
These products require as low an iodine value as possible for oxidative stability
with a steep SFI slope to provide a melting point lower than body temperature
for good eating characteristics. There are two alternative routes to obtain a steep
SFI curve: (1) nature has provided the lauric oils with a steep SFI and a sharp
melting point or (2) most liquid oils can be selectively hydrogenated to provide
the desired solids melting relationships. Selective hydrogenation is a progressive
diminution of the most unsaturated fatty acid groups. When the overall hydrogena-
tion effect is that the fatty acids with three double bonds are nearly all reduced to
two double bonds, before those with two are nearly all reduced to one, before the
monounsaturates are saturated, then good or high selectivity exists (153).
The frying shortening illustrated in Figure 6 meets the restaurant industry
requirements for a stable heat-transfer media that becomes a part of the food to sup-
ply texture, mouthfeel, and an enhancement to food flavor. Solid frying shortenings
are usually composed of a single hydrogenated base or possibly two selectively
hydrogenated bases chosen for a slightly different slope than what is available
with a single base. In the case of two bases, both have high solids at the lower tem-
perature readings but fall off rapidly to a low melting point, which will result in the
desired eating characteristics or mouthfeel with an excellent oxidative and frying
stability. For example, the frying shortening on Figure 6 is a blend of the 58 and
70 iodine value cottonseed oil basestocks detailed previously on Table 20. It could
also be produced by selectively hydrogenating cottonseed oil to a 65 iodine value.
Nondairy shortenings are used to replace butter fat in applications such as imita-
tion cheese, mellorine, an ice cream substitute, coffee whiteners, vegetable-
oil-based whipped toppings, dip bases, milk analogs, etc. The solid fat index values
264 COTTONSEED OIL
and melting characteristics of these shortenings usually closely approximate those

of butter. The solids contents are relatively high at the cooler temperatures to pro-
vide structure for the food products but drop off rapidly to melt completely about
body temperature for good eating characteristics (106).
Liquid Shortenings. Prior to the introduction of emulsifiers, antioxidants, and
antifoamers, shortenings relied on fatty acid composition, crystal habit, plasticiza-
tion, and tempering to provide functionality. The aeration function of plastic short-
enings was correlated with the polymorphic form of the triglyceride while shelf life
and frying stability were attained by the level of saturation to resist oxidation. Plas-
tic shortenings that exist in the small beta-prime crystal form aerate batters much
more thoroughly than those in the large beta form, and saturation of the unsaturated
fatty acids eliminates a reaction site for oxygen to extend flavor and frying stability.
These rules are still applicable but emulsifiers, antioxidants, and antifoamers have
significantly reduced the dependence on the plastic consistency for functionality.
This reduced dependence has allowed the development of liquid shortenings that
combine the functional characteristics of plastic shortenings with the bulk handling
characteristics of a liquid oil.
A liquid shortening is a stable dispersion of solids with the proper polymorphic
form in a continuous oil phase that is both flowable and pumpable over a tempera-
ture range of 6090 F, or 15.632.2 C. The solids are derived from either hardfats
or emulsifiers and sometimes both. For some applications, the primary function of
the liquid shortening is as a delivery system for emulsifiers that alter the character-
istics of the food product. Like the plastic shortenings, liquid shortenings are
designed for the specific end-use application. Fluid shortenings require a wide
range of functional and physical attributes. For optimum fluidity the base oil should
be liquid at room temperature and have no suspended solids. Oxidative stability
is dependent on the base oil, which may require the use of a high-stability oil. Uti-
lization of a hydrogenated basestock reduces fluidity and requires more exacting
processing to prevent product separation. Antioxidant and antifoamer additions
improve oxidative stability and frying life, but the improvement does not approach
the stability of a plastic shortening with these same additives.
Saturated fatty acid composition of the hardfat or fat-based emulsifiers added to
the suspension are important to the functionality and physical form of the fluid
shortening system. Palmitic fatty acids are more soluble in oil, but tend to supercool
more and take longer to crystallize. Stearic fatty acids do not supercool as much
but are not very soluble. For the greatest emulsifier loading in the system, emulsi-
fiers high in palmitic fatty acid content are the best. However, for the most stable
viscosity over time, stearic fatty acids are preferred as they readily crystallize out of
the liquid oil and form a stable crystalline matrix. Therefore, in contrast to plastic
shortenings, it is desirable to formulate nonemulsified liquid shortenings with beta-
stable blends whose large crystals tend to form a stable dispersion. Normally, a low-
iodine-value, beta-forming hardfat is used to seed crystallization for liquid shorten-
ings. The hardstock levels can vary from as low as 1.0% to higher levels, as required
to produce the required viscosity, usually no higher than 10%. The ease in which
beta-forming hardfats convert to the stable beta crystal form make them ideal for
SHORTENINGS 265
liquid shortening crystallization. The beta crystals do not intertwine to form a

matrix that can enmesh the liquid phase and form a thick product as found with
beta-prime hardfats. The liquid opaque shortening SFI slope plotted on Figure 6
is typical of a liquid bread shortening formulated with 10% soybean hardfat crystal-
lized in 90% unhardened liquid oil. In contrast, emulsified liquid shortenings should
be formulated with surfactants made from cottonseed oil, which has a high palmitic
fatty acid content, to take advantage of the greater emulsifier loading capability (154).
Shortening Flakes and Chips. Another shortening form is represented by pro-
ducts crystallized with a chill roll. Edible oil products crystallized on a chill roll
solidify into a thin flake form that affords ease of handling of these higher melting
products. For quite some time this crystallization process was limited to hardfats
commonly referred to as stearines. Now, chill-roll use has been expanded to include
several different products for various applications. These shortening products use
both the steep and flat SFI slopes produced with selective and nonselective hydro-
genation techniques depending on application.
Stearines, Titers, or Low-Iodine-Value FlakesMelting before use is the

standard procedure for these high-melting, almost completely saturated oil
products. Cottonseed hardfat with a 5 or less IV has a melting point of
approximately 140 F or 60 C. Some of the applications for these flakes are:
melting point adjustment, lubricant, encapsulation base, bread stabilizer,
chewing gum ingredient, and so forth.
Hard EmulsifiersDistribution in frozen desserts, whipped toppings, pasta,
peanut butter, yeast-raised products, shortenings, and other applications
require that these high-melting surfactants be added as a liquid. The thin
flake form allows the melting process to be accomplished at a quicker rate
with less damage to the product.
Shortening ChipsThese thicker and larger chips are incorporated into baked
products to provide a flaky product similar to danish pastry without the labor
intensive roll-in process. Normally, the thickness of these chips is controlled
at a mean of 0.05 inches while the standard flake thickness is no more than
half this thickness. Shortening chips are formulated to resist incorporation into
doughs and batters during mixing and still have a palatable mouthfeel. Steep
SFI basestocks with high solids contents at 5080 F, or 1026.7 C, and
declining rapidly thereafter provide this functionality.
StabilizersThese slightly high-melting, flat SFI profile flakes were designed
to stabilize roll icings, syrups, donut glazes, and butter creme icings. An icing
stabilizers performance requires a flat SFI slope to maintain a soft consis-
tency with solids contents high enough to facilitate a rapid set. The rapid set
allows the icing to resist fingerprinting when handled while the flat SFI slope
allows the icing to retain an elasticity to prevent product brittleness and
flaking. A range of melting point products, usually from 110 F, or 43.3 C, to
125 F, or 51.7 C, are made for this application, which has been extended to
confections, cake mixes, and other food products.
266 COTTONSEED OIL
Confectionery FatsConfectionery fats require a very steep SFI curve, which

makes them brittle with a short melting range that ensures a quick meltdown
and a pleasant mouthfeel. These fats, formulated to resemble the character-
istics of cocoa butter, are usually referred to as hard butter. Quality hard
butters have a relatively high solid fat index at room temperature of above
50%, as lower levels can provide a greasy or tacky feel. The SFI then falls off
quickly for a complete melt for most products at 95102 F. (3539.2 C).
Processes used to prepare hard butters include hydrogenation, interesterifica-
tion, solvent and dry fractionation, and blending. The hard butters can use
both lauric and nonlauric source oils.
9.3. Cottonseed Oil Utilization

Shortenings are a unique food ingredient in that a high degree of interchangeability
among the raw materials is possible for many products and uses. However, in order
for a particular oil to substitute for another in a given product, it may be necessary
for it to undergo additional processing steps, which may increase its cost to become
an adequate replacement. After this additional processing, if the replacement fat or
oil can be substituted in terms of physical and analytical properties in the end pro-
duct, then price becomes a major consideration for the employment of the raw
material replacement. Experience has shown that small cost differences in compet-
ing source oils can markedly change the proportion of the oils used in a shortening.
Shifts in the use of various fats and oils in the composition of an individual
shortening were more common in the past that at present. Source oil labeling
requirements have made alternate formulations for the same product difficult. How-
ever, shortening customers may still substitute alternative source oil produced
shortenings that have comparable performance characteristics. Table 29 tracks
the changes in the source oil use for U.S. shortenings from 1940 through 2000
(13, 2628).
TABLE 29. U.S. Shortening Source Oils Usage (Metric Ton).
Year 1940 1950 1960 1970 1980 1990 2000
Coconut oil 8.2 9.1 4.5 20.4 46.7 15.4a

Corn oil 0.5 0.5 1.8 5.4 W 122.5 12.2
Cottonseed oil 373.3 249.0 165.6 125.2 85.7 114.3 75.3
Palm oil NR NR NR 40.8 85.3 94.8a 84.8a
Peanut oil 10.4 5.4 0.9 7.3 NR NR NR
Soybean oil 96.2 381.5 530.3 989.7 1202.5 1816.2 2591.9
Lard 7.7 80.3 217.7 195.0 171.5 119.7 108.4
Tallow 26.3 14.1 121.6 236.8 305.3 288.9 128.8
Unidentified 20.0 12.2 1.4 3.2 8.2 6.4
Total 542.5 783.4 1043.7 1623.9 1905.1 2578.2 3001.5
Per capita, kg 5.0 5.7 7.8 8.3 10.1 10.5
a
Estimated.
W withheld.
NR none recorded.
MARGARINE AND SPREADS 267
Shortenings were developed as lard substitutes to provide an outlet for the then
plentiful and inexpensive cottonseed oil. The first shortenings, prepared by blending
liquid cottonseed oil with stearines, were no better than the product imitated. Intro-
duction of the hydrogenation process enabled cottonseed oil processors to develop
shortenings that outperformed the natural lard products. Shortenings, initially mar-
keted as economic lard replacements, became the desired product that commanded
premium pricing. Specialty shortening development was enhanced with the intro-
duction of mono-and diglyceride emulsifiers in 1933. Throughout these develop-
ments, cottonseed oil retained its position as the dominate source oil for
shortenings until soybean oil rose from a minor, little known, problem-related oil
before 1940 due to a shortage of cottonseed oil created by World War II. A plentiful
supply of lower cost soybean oil and the advanced technologies developed during
the war years made it impossible for cottonseed oil to regain its dominant source oil
position for shortenings. However, it could not be totally replaced, cottonseed oil
has some definite advantages over soybean and most other vegetable oils, one being
its beta-prime (b0 ) crystal habit. A smooth consistency, fine texture, wide plastic
range, good creaming properties, and a tolerance to high temperatures have become
the standard for shortening. Most of these desirable characteristics are contributed
by the tiny, needle-like beta-prime (b0 ) crystals that pack together into dense, fine
grained, ridged structures to form a three-dimensional network capable of immobi-
lizing a large amount of liquid oil. Soybean oil, as well as most of the other available
vegetable oils, has a beta (b) crystal habit, which are large, course, high-melting,
self-occluding crystals that clump, allowing separation of the liquid oils, and
responsible for a visible grainy appearance. A beta-prime (b0 ) crystal form can
be induced with the addition of hydrogenated cottonseed oil, or another oil with
a beta-prime (b0 ) crystal habit (see Table 12 Crystal Habit of Hydrogenated Oils
and Fats), usually at levels of 10% or higher.
10. MARGARINE AND SPREADS
Margarine is a flavored food product containing 80% fat, made by blending selected
fats and oils with other ingredients, fortified with Vitamin A, to produce a table,
cooking, or baking fat product that serves the purpose of dairy butter but is different
in composition and can be varied for different applications (145). Margarine was
developed to fill both an economic and a nutritional need when it was first made
as a butter substitute. The ability to be physically altered to perform in many varied
applications was a major factor in the growth of margarine. There are over ten dif-
ferent types of margarine produced today, including regular, whipped, soft tub,
liquid, diet, spreads, no fat, restaurant, bakers, and specialty types, which are pack-
aged in as many different packages. These margarines are made from a variety of
fats and oils, including cottonseed, soybean, palm, corn, canola, safflower, sun-
flower, lard, and tallow. Margarine products cater to the requirements of all the
different consumers; retail, foodservice, and food processor.
268 COTTONSEED OIL
TABLE 30. U.S. Margarine and Spread Source Oil Usage.
Metric Ton (MT)
Year 1950 1960 1970 1980 1990 2000
Corn oil 0.5 30.8 83.9 101.2 94.3 25.4

Cottonseed oil 232.7 76.2 30.8 11.3 D D
Safflower oil 5.9 10.0
Soybean oil 173.3 621.4 639.6 749.8 793.3 758.9
Animal fats 7.3 34.5 44.9 47.2 15.9 5.9
Unidentified 11.3 4.5 15.4 49.9 5.9
Total 425.0 768.8 813.8 924.9 953.5 796.1
Per Capita, lbs 2.8 4.3 5.0 5.2 4.9 3.8
D data withheld by census.
10.1. Cottonseed Oil Utilization

Cottonseed oil lost its dominant source oil position for margarine oils in 1951. Soy-
bean oil became the leading margarine source oil due to the technologies developed
for use of this lesser cost raw material. Cottonseed oil use in margarines steadily
decreased thereafter to also be surpassed by corn oil in 1963. Margarine manufac-
tures found that cottonseed oil was no longer required in the formulation to main-
tain a beta-prime crystal habit, due to the use of multiple high-trans basestocks and
the improvements in shipping, storage, and grocery store display refrigeration cap-
abilities. This eliminated the 10% or more hydrogenated cottonseed oil in most con-
sumer margarine oil formulations. Most of the industrial margarine oil formulations
still maintain a hydrogenated cottonseed oil hardstock in the formulation to provide
the plasticisticy and consistency desired. The changes in source oil use in margarine
and spread base oils are outlined in Table 30.
10.2. Margarine Oil Formulation

The physical and functional aspects of a margarine product are primarily dependent
on the characteristics of the major ingredientthe margarine oil or marbase. Mar-
garine consistency, flavor, and emulsion stability depend on crystallized fat. In the
United States, hydrogenation is the preferred process used to change the solid-
liquid relationship of margarine basestocks. A direct relationship exists between
the fat solids content and the structure, consistency, and plasticity of the finished
margarine. SFI values at 50 F, 70 F, and 92 F, or 10.0 C, 21.1 C, and 33.3 C,
are used by most margarine manufactures in the United States for margarine
consistency control. The SFI values are indicative of the crystallization tendencies
and the finished product quality as shown in Table 31 (106).
Consumer Margarine Oil Formulations. Consumer margarines are formulated
by blending two or more basestocks with different degrees of hardness. This per-
mits the margarines to be spreadable directly out of the refrigerator and to maintain
a solid consistency at room temperature. Hydrogenated cottonseed oil was a
TABLE 31. Consumer Margarine Solid Fat Index (SFI) Affect on Product Characteristics.
Solid Fat Index (SFI)

Characteristic
Temperature % Solids Affected Solid Fat Index (SFI) Influence

10 C or 50 F 10 to 28 Spreadability Optional range for spreadibility at refrigerator
temperatures.
10 C or 50 F 10 to 28 Printability Low SFI Colder Chilling Temp. Required
High SFI Channeling Possible with Cooling
21.1 C or 70 F 5 to 18 Consistency Body and Resistance to Oil Separation
Too High Brittle, Firm, Poor Spreadability
Too Low Soft, Soupy, Oil Separation
33.3 C or 92 F 3.5 max Mouth Feel Quality Consumer Tablegrade Margarine
Melts Quickly With a Cooling Sensation
Too High Lingering Pasty, Greasy, Waxy,
Sensation Due to Palate Coating
component of most vegetable oil margarines to induce a beta-prime crystal habit to

prevent graininess for quite some time. However, the development of more spread-
able margarines, the use of multiple basestocks, and uniformly low cold storage
temperatures has reduced the transition of margarines formulated with soybean
and corn oils to the beta crystal form that causes sandiness.
The hydrogenated vegetable oil basestocks best suited for table-grade products
have steep SFI slopes to provide the desired eating, melting, and nonoiling physical
characteristics along with machinability. The cottonseed oil basestocks previously
presented in Table 20 may be used to produce margarine oils as well as shortenings.
Margarines have been able to conform to the increased health consciousness of the
U.S. consumer by responding with increased polyunsaturated fatty acids and liquid-
oils levels. Although polyunsaturate level is less emphasized now, apparently it is
still perceived as healthy by the consumer. Therefore, it is still advantageous to have
a high polyunsaturate-to-saturates ratio contributed by a high liquid-oil level.
Table 32 identifies the typical SFI values for six table-grade marbases; three are
TABLE 32. Typical SFI Values for Margarine Products.
10 C 21.1 C 26.7 C 33 C 40 C

Solids Fat Index: 50 F 70 F 80 F 92 F 104 F
Consumer prints:
Soft stick 22.0 13.5 2.0
All hydrogenated 28.5 17.5 3.0
High liquid 30.0 17.0 1.5
Consumer Tub:
All hydrogenated 12.0 7.5 3.0
High liquid oil 11.0 5.5 0.7
Consumer liquid 7.0 6.0 6.0 5.5 5.0
Bakers margarine 24.0 18.0 17.0 13.0 7.0
Roll-in margarine 29.0 24.0 22.0 16.0 10.0
Puff pastry 34.0 30.0 28.0 24.0 17.0
270 COTTONSEED OIL
for stick products, two are soft tub types and a liquid or squeezable margarine. The
soft stick marbase probably represents the softest stick type margarine that can be
packaged successfully. The present day packaging equipment will deposit, wrap,
and carton product that is softer, but the margarine could not withstand the normal
abuse after packaging in storage and distribution. This product was introduced after
the soft tub products in an effort to take advantage of the soft concept in the old
familiar stick package (106).
The all-hydrogenated stick table-grade product represents the type that was the
main consumer margarine product for quite some time. It is still the most preferred
table fat product for baking and some cooking due to better oxidative stability. This
product also prints well due to the high 50 F or 10 C SFI, especially in equipment
that doesnt deposit the margarine into preformed quarters parchment. The high
liquid-oil stick marbase represents the majority of the stick margarine production.
The high liquid oil in the formula provides a marketing claim for low saturates but
reduces the finished margarines oxidative stability and therefore shelf life. The sur-
face of the margarine develops a darker color from oxidation that is quite noticeable
when the surface layer is scraped away during use.
Soft tub marbase compositions are somewhat like a compound shortening for-
mulation: a blend of a soft and a hard basestock. The hard basestock cant be as
saturated as the low-iodine-value hardfats used for shortenings and should have a
steep SFI slope for good eating characteristics. Two different soft tub compositions
are shown in Table 32. The all-hydrogenated product provides the oxidative stabi-
lity and the firmest margarine consistency. Slight consistency differences can be the
difference between a soft tub margarine with a picture perfect surface appearance or
one with excessive lid slosh. The all-hydrogenated margarine oil base has a
better chance of retaining a smooth surface because it should set quicker than
the product formulated with liquid oil.
Liquid margarine has been marketed for quite some time but it has never
achieved significant consumer acceptance. Food processors have accepted and
used this product for specialty applications more so than consumers. Liquid
margarine has been prepared using both beta and beta-prime type hardfats. Beta-
prime (b0 ) hardfats, like cottonseed oil, have been found more suitable for the pro-
duct packaged without a tempering for crystallization process stage. Product
prepared with beta crystal forming hardfats requires tempering of the supercooled
mixture at an elevated temperature for a period of time under agitation to develop
and stabilize the beta crystal. This liquid margarine process resembles the liquid
shortening process closely. The beta crystal formulation and procedure produces
a more fluid, less viscous product with better suspension stability than the beta-
prime product but costs more to produce (151). The beta-prime hardfat direct
process provides a better mouthfeel and flavor but requires constant refrigeration
to avoid separation.
Industrial Margarines. Foodservice and food processor margarines and spreads
are considered industrial products. These products are formulated or packaged
for more specific applications than the consumer products. The most popular
foodservice margarine is the consumer stick margarine formulation packaged in

one-pound solids for use in food preparation. Individual serving or portion control
soft spread products are popular foodservice dinning room products. The major
difference that affects the desired functionality is the margarine oil composition.
The consumer table-grade margarine formulation is also packaged in 50 pound
and larger containers for food processor applications. It is plasticized like a short-
ening instead of the margarine print procedure for roll-in and other applications.
Bakers margarine is designed to have a wide plastic range with good creaming
properties like the standard all-purpose shortening. In fact, the marbase for bakers
margarine can be the same composition as an all-purpose shortening using an 80
85 IV hydrogenated cottonseed or soybean oil basestock with a fully hydrogenated
beta-prime crystal forming cottonseed oil hardfat. Special roll-in margarines for
danish and other pastries are prepared with fat blends similar to anhydrous or short-
ening products. Margarine formulations provide a buttery flavor and color to the
finished product as well as moisture to produce steam during baking to help expand
the dough layers and improve flakiness. Several other specialty products also rely
on the marbase formulation for functionality and the margarine process to provide
flavor, color, and moisture for functionality (106).
10.3. Spread Formulations

All products resembling margarine that contain less than 80% but more than 40%
fat are required to be labeled as spreads. However, these products must conform
with the margarine standard in all respects except the fat content and that safe
and suitable ingredients not provided for in the standard may be added to improve
functional characteristics so that the spreads are not inferior to margarine. Soft tub
and stick spreads containing 40% to 75% fat are usually formulated from the same
fat and oil blends as those used for regular consumer margarines. The other ingre-
dients used are also basically the same with the following exceptions (155, 156):
Milk protein acts as an oil-in-water emulsifier. Consequently, the use of milk,

casein, or caseinates can result in a phase reversion. Therefore, most spread
formulations are milk and milk-protein free.
Emulsifier levels are increased slightly to improve the physical characteristics
of the emulsion and its stability; typically 0.4% to 0.6% alpha monoglyceride
levels with mono-and diglyceride use. Soft mono-and diglycerides are
essential in protein-free spreads. Mono-and diglyceride and polyglycerol
emulsifier systems have been found effective in spreads containing significant
quantities of protein.
Lecithin use in low fat spreads may decrease the emulsion stability and
increase the tendency to oil off; however, it also functions to slow the
emulsion breakdown in the mouth. Therefore, the use of lecithin and the
level of use must be evaluated for each formulation.
272 COTTONSEED OIL
Gelatin, pectin, carrageenans, agar, xanthan gum, starch alginates, or methy-

lecellulose derivatives are gelling or thickening agents used in some spreads to
improve the body.
The higher emulsifier levels used for spread can produce tighter emulsions and
the gelling or thickening agents can affect the rate and order that flavors are
perceived. The flavor content and types must be defined to produce oral
responses similar to the high-fat products.
Preservatives are more important in spreads than in regular margarines as a
result of the higher moisture content.
The light reflection of a spread is different from that of regular margarine as a
result of the increased number of water droplets present. Therefore, it is
necessary to add about twice the amount of color used in normal margarine to
obtain the same color intensity.
11. OTHER COTTONSEED OIL USES
Major changes are continually occurring in the markets for fats and oils as dis-
cussed previously. Food applications have been a major use for cottonseed oil
but it has also been used in soap, lubricants, sulfonated oil, pharmaceuticals, pro-
tective coastings, rubber, as a carrier for nickel catalysts, and, to a lesser degree, in
the manufacture of leather, textiles, printing ink, polishes, synthetic plastics, and
resins (23). More recently, cottonseed oil was used in the synthesis of sucrose
polyesters as a zero-calorie fat substitute that has a trade name of Oleans, or a com-
mon name of Olestra. This product, developed by Proctor & Gamble, was approved
by the U.S. Food and Drug Administration in 1996 (157160). Cottonseed oil was
chosen for this application because it imparted a pleasing nutty flavor to the fat
substitute.
REFERENCES
1. Anonymous, INFORM, 11(8), 820839 (2000).

2. S. R. Gregory, E. Hernandez, and B. R. Savoy, in C. W. Smith J. T. and Cothren, eds.,
Cottonseed Processing. Cotton Origin, History, Technology, and Production, John Wiley
& Sons, New York, 1999, pp. 793823.
3. R. D. OBrien and P. J. Wan, in R. F. Wilson, ed., Proceedings of the World Conference
on Oilseed Processing and Utilization, AOCS Press, Champaign, Illinois, 2001, pp. 90
140.
4. National Cottonseed Products Association, Inc., Cottonseed And Its Products, 9th ed.,
1989, pp. 47.
5. L. B. Wrenn, Cinderella of the New South A History of the Cottonseed Industry, 1855
1955, The University of Tennessee Press, Knoxville, Tennessee, 1995, pp. 9, 15, 44,
7374, 76, 80, and 81.
REFERENCES 273
6. L. A. Jones and C. C. King, eds., in Cottonseed Oil, National Cottonseed Products

Association, Inc. and the Cotton Foundation, Memphis, Tennessee, 1990, pp. 12, 57,
910, and 14.7. L. B. Wrenn, INFORM, 4(1), 619 (1993).
7. L. B. Wrenn, INFORM, 4(1), 619 (1993).
8. L. B. Wrenn, INFORM, 6(1), 98 (1995).
9. L. B. Wrenn, INFORM, 9(1), 100106 (1998).
10. L. H. Meyer, in Food Chemistry, The AVI Publishing Co., Westport, Connecticut, 1975,
p. 55.
11. R. D. OBrien, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, CRC Press,
12. W. H. Goss, in Proceedings of A Six-Day Short Course in Vegetable Oils, The American
Oil Chemists Society and The University of Illinois, Champaign, Illinois, 1948, pp. 53
55.
13. K. F. Mattil, in D. Swern, ed., Baileys Industrial Oil and Fat Products, 3rd ed., Wiley
Interscience New York, 1964, pp. 249263, and 296.
14. E. M. James, U. S. Patent 2,050,844, 1936.
15. B. Clayton, U.S. Patents 2,100,276 and 2,100,277, 1937.
16. B. H. Thurman, U.S. Patent 2,150,733, 1939.
17. H. C. Black, A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
Publisher, Inc., New York, 1948, pp. 732762.
18. C. E. Morris, J. Amer. Oil Chem. Soc., 26(10), 607 (1949).
19. R. C. Hastert, INFORM, 9(12), 1140 (1998).
20. R. D. OBrien, in P. J. Wan et al., eds., Introduction to Fats and Oils Technology, 2nd ed.,
21. L. C. Brown, J. Amer. Oil Chem. Soc., 26(10), 634 (1949).
22. J. E. Slaughter, Jr., in Proceedings of A Six-Day Short Course in Vegetable Oils, The
American Oil Chemists Society and The University of Illinois, Champaign, Illinois,
1948, pp. 119131.
23. A. E. Bailey, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
24. H. J. Dutton, J. Amer. Oil Chem. Soc., 58(3), 234236 (1981).
25. USDA, Fats and Oils Situation, April 1971, pp. 2425, 27.
26. USDA, Fats and Oils Situation, July 1977, pp. 1415.
27. USDA, Oil Crops Situation and Outlook Yearbook, August 1985, p. 8. USDA, Fats and
Oils Situation Report, January 1981, Table IX.
28. USDA, Oil Crops Situation and Outlook Yearbook, October 2002, pp. 40, 44, 53, 6266,
and 68.
29. L. A. Jones and C. C. King, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products,
Vol. 2, 5th ed., Wiley-Interscience, New York, 1996, pp. 188189.
30. National Cottonseed Products Association, The Cottonseed Industry Statistical Hand-
book, Memphis, Tennessee, 2002.
31. C. E. Lund, in C. E., in A. E. Bailey, ed., Cottonseed and Cottonseed Products,
Interscience Publisher, Inc., New York, 1948, pp. 58, 62, and 64.
32. USDA, Oil Crops Situation and Outlook Yearbook, October, 1997.
274 COTTONSEED OIL
33. USDA, Oil Crops Situation and Outlook Yearbook, October, 1998, p. 72.
34. P. J. Wakelyn, in Kirk-Othmer Encyclopedia of Chemical Technology, 5th ed., John
Wiley & Sons, Inc. New York, 2002.
35. R. J. Kohel and C. F. Lewis, eds., Cotton, American Society of Agronomy, Crop Society
of America, Soil Society of America, Madison, Wisconsin, 1984.
36. F. D. Gunstone, in F. D. Gunstone, ed., Vegetable Oils in Food Technology, CRC Press,
Baca Raton, Florida, 2002, p. 11.
37. K. Esau, Anatomy of Seed Plants, John Wiley & Sons, Inc., New York, 1977.
38. A. E. Percival, J. E. Wendel, and J. M. Stewart, in C. W. Smith and J. T. Cothren, eds.,
Cotton Origin, History, Technology, and Production, John Wiley & Sons, New York,
1999, pp. 3363.
39. D. S. Calhoun and D. T. Bowman, in C. W. Smith and J. T. Cothren, eds., Cotton
Origin, History, Technology, and Production, John Wiley & Sons, New York, 1999, pp.
361414.
40. M. F. Stansbury, A. F. Cucullu, and G. T. Den Hartog, J. Agric. Food Chem., 2, 693
(1954).
41. H. H. Ramsey, in Proceedings 33rd Oilseed Processing Clinic, New Orleans, Louisiana,
January 1617, 1984.
42. W. H. Tharp, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
Publishers, Inc., New York, 1948, pp. 117, and 142149.
43. N. O. V. Sonntag, in D. Swern, ed., Baileys Industrial Oils and Fat Products, Vol. 1, 4th
ed., Wiley Interscience Publishing Co., New York, 1979, pp. 289, 352362.
44. M. L. Meara and E. H. Steiner, Studies on the Fatty Acid Composition and the
Crystallizing Behavior of Some of the Major Components, to Obtain Fundamental
Information That Will Contribute to the Development of Imroved Edible Products and
Hence to Expanded Utilization of Cottonseed Oil, British Food Manufacturing Industries
Research Association, Leatherhead, Surrey, U.K., 1966.
45. H. R. Leffler, in Proceedings 33rd Oilseed Processing Clinic, New Orleans, Louisiana,
January 1617, 1984.
46. Z. M. Sawan, A. A. El-Farra, and S. A. El-Latif, J. Agron. Crop Sci., 161, 50 (1988).
M. K. Schwitzer, Margarine and Other Food Fats, Interscience Publisher, New York,
1956, p. 64.
47. J. F. OKelly, W. W. Hull, and M. Geiger, Technical Bulletin 20, Mis. Agr. Exp. Sta.,
State College, Mississippi, 1933.
48. C. H. Boatner, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
Publisher, Inc., New York, 1948, p. 213.
49. F. A. Norris, in D. Swern, ed., Baileys Industrial Oil and Fat Products, Vol. 2, Wiley
Interscience New York, 982, pp. 175, 253, 295, and 479.
50. L. C. Dasilva-Ramos and R. J. Kohel, J. Amer. Oil Chem. Soc., 64, 1337 (1987).
51. C. M. Lawhon, Cater, and K. F. Mattil, J. Amer. Oil Chem. Soc., 54, 75 (1977).
52. P. J. Wan, Oil Mill Gaz., 103(11), 3845 (1998).
53. Agricultural Research Service and National Cottonseed Products Association, in
Proceedings of a Conference, Southern Regional Research Center, ARS, USDA,
New Orleans, Louisiana, 1978.
54. E. W. Lusas and G. M. Jividen, J. Amer. Oil Chem. Soc., 64, 839 (1987).
REFERENCES 275
55. P. J. Wakelyn and P. J. Wan, in C. Tzia, ed., Extraction Optimization, Marcel Dekker,
New York, 2003, pp. 391427.
56. T. A. Wilkins and R. Mishra, in Proc. 11th International Plant and Animal Genome
Conference, Scherago Int., New York, 2003, p. 61.
57. International Cotton Advisory Committee, ICAC Recorder, XX(4), 812 (2002).
58. International Cotton Advisory Committee, Report of the Second Expert Panel on
Biotechnology of Cotton, 2004.
59. D. L. Nida, K. H. Kolacz, R. E. Buehler, W. R. Deaton, W. R. Schuler, T. A. Armstrong,
M. L. Taylor, C. C. Ebert, G. J. Rogan, S. R. Padgett, and R. L. Fuchs, J. Agric. Food
Chem., 44, 19601966 (1996); D. L. Nida, S. Patzer, P. Harvey, R. Stipanovic, R. Wood,
and R. L. Fuchs, J. Agric. Food Chem., 44, 19671974 (1996a).
60. O. A. Momtaz, A. A. Diab, and M. A. Madkour, in Proceedings. Beltwide Cotton
Conference, National Cotton Council, Memphis, Tennessee, 2000, pp. 513517.
61. R. G. Cantrell, in Proceedings World Cotton Research Conference 3, Capetown,
South Africa, International Cotton Advisory Committee. Washington, D.C., 2003.
62. Y. L. Cousins, B. R. Lyon, and D. J. Llewellyn, Aust J. Plant Physiol., 18, 481494
(1991).
63. X. Y. Wu and X. H. Xiang, J. Hunan Agric. Univ., 24, 494498 (1998).
64. Q. Liu, S. Singh, and A. Green, J. Amer. Coll. Nutr., 21(3), 205S211S (2002).
65. K. S. Rathore, in Proceedings Beltwide Cotton Conference, National Cotton Council,
Memphis, Tennessee, 1997.
66. J. H. Benedict and D. W. Altman, in J. J. Jenkins and S. Saha, eds., Emerging
Technologies in Cotton Breeding, Science Publishers, Enfield, New Hampshire,
2000, pp. 137201.
67. C. R. Benedict, J-L. Lu, G. S. Martin, J. Lu, and R. D. Stipanovic, in C. R. Benedict and
G. M. Jividen, eds., Genetic Control of Cotton Fiber and Seed Quality, Cotton
Incorporated, Cary, North Carolina, 2000, pp. 244249.
68. G. S. Martin, C. R. Benedict, J. Liu, and L. Puelchaber, in C. R. Benedict and G. M.
Jividen, eds., Genetic Control of Cotton Fiber and Seed Quality, Cotton Incorporated,
Cary, North Carolina, 2000, pp. 262270.
69. P. J. Wakelyn and P. J. Wan, Oil Mill Gaz., 106(8), 1422 (2001).
70. K. Blaine, S. Kamalden, and D. Powell, J. Food Sci., 67(9), 32003208 (2002).
71. European Commission, GM-Derived Cottonseed Oils Notified for Marketing in the EU,
Midday Express Notification, December 19, 2002.
72. P. J. Wan, in R. D. OBrien, W. E. Farr, and P. J. Wan, ed., Introduction to Fats and Oils
Technology, 2nd ed., American Oil Chemists Society, Champaign, Illinois, 2000,
pp. 2048.
73. G. Jurriens and A. C. J. Kroesen, J. Amer. Oil Chem. Soc., 42, 9 (1965).
74. J. P. Cherry, J. Amer. Oil Chem. Soc., 60, 360 (1980).
75. P. J. Van Niekerk and A. E. C. Burger, J. Amer. Oil Chem. Soc., 62, 531 (1985).
76. J. B. Reeves and J. C. Weirauch, Composition of Foods, Agriculture Handbook No. 8-4,
United States Department of Agriculture, Washington, D.C., 1979.
77. D. Firestone, Physical and Chemical Characteristics of Oils, Fats, and Waxes, AOCS
Press, Champaign, Illinois, 1999, pp. 3233.
276 COTTONSEED OIL
78. J. P. Cherry, J. G. Simmons, and R. J. Kohel, in Proceedings of Belwide Cotton

Production Research Conference, 1978.
79. J. P. Cherry, R. J. Kohel, L. A. Jones, and W. H. Powell, in J. R. Mauney and J. M.
Stewart, eds., Cotton Physiology, The Cotton Foundation, Memphis, Tennessee, 1986,
p. 557.
80. J. M. Bland, E. J. Conkerton, and G. Abraham, G., J. Amer. Oil Chem. Soc., 68, 840
(1991).
81. R. A. Phelps, F. S. Shenstone, A. R. Kemmerer, and R. J. Evans, Poult. Sci., 44, 358
(1965).
82. E. Allen, A. R. Johnson, A. C. Fogerty, J. A. Pearson, and F. S. Shenstone, Lipids, 2, 419
(1967).
83. F. L. Carter and V. L. Frampton, Chem. Rev., 64, 497 (1964).
84. J. S. Harris, F. C. Magne, and E. L. J. Amer. Oil Chem. Soc., 41, 309 (1964). R. C. Hastert,
The Partial Hydrogenation of Edible Oils, AOCS Short Course, Phoenix, Arizona, 1988.
85. A. V. Bailey, J. A. Harris, and E. L. Skau, E. L., in J. Amer. Oil Chem. Soc., 43, 107
(1966).
86. F. S. Shenstone and J. R. Vickery, Nature, 190, 168 (1961).
87. P. H. Eaves, H. P. Depuy, L. L. Holenthol, E. T. Rayner, and L. E. Brown, J. Am. Oil
Chem. Soc., 45, 293 (1968).
88. V. L. Frampton, Poult. Sci., 45, 527 (1966).
89. R. J. Evans, D. H. Bauer, and C. J. Flegal, Poutl. Sci., 54, 824 (1975).
90. L. Marchlewski, J. fuer Prakusche Chemis, 60, 81 (1899).
91. W. A. Pons, Jr., C. L. Hoffpauir, and OConnor, J. Amer. Oil Chem. Soc., 28, 8
(1951).
92. Y. Yabe, S. Tan, T. Ninomiya, and T. Okada, J. Food Hygien. Soc. Jpn., 25, 264 (1984).
93. B. J. F. Hudson and M. Ghavami, Lebensm. -Wiss. -u Technol., 17, 191 (1984).
94. N. A. M. Eskin, B. E. McDonald, R. Przybylski, L. J. Malcolmson, S. Scarth, T. Mag,
K. Ward, and D. Adolph, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products,
Vol. 2, 5th ed., Wiley-Interscience, New York, 1996, pp. 5253 and 67.
95. T. Harper, in R. F. Wilson, ed., Proceedings of the World Conference on Oilseed
Processing and Utilization, AOCS Press, Champaign, Illinois, 2001, p. 25.
96. L. N. Norcia and B. E. Rosenthal, JAOCS, 43, 168170 (1996).
97. J. A. Pennington, Bowes and Churchs Food Values of Portions Commonly Used,
J. B. Lippincott Co., New York, 1994.
98. E. H. Pryde, in D. R. Erickson et al., eds., Handbook of Soy Oil Processing and
Utilization, Soybean Assoc. and Am. Oil Chem. Soc., St Louis, Missouri and Cham-
paign, Illinois, 1980, p 13.
99. K. J. Smith, P. B. Polen, D. M. DeVries, and F. B. Coon, J. Amer. Oil Chem. Soc., 45(9),
866869 (1968).
100. T. L. Mounts, C. D. Evans, H. J. Dutton, and J. C. Cowan, J. Amer. Oil Chem. Soc., 46(9),
482484 (1969).
101. M. M. Chaudry, A. I. Nelson, and E. G. Perkins, J. Amer. Oil Chem. Soc., 53(11), 695
697 (1976).
102. F. J. Flider and F. T. Orthoefer, J. Amer. Oil Chem. Soc., 58, 270272 (1981).
REFERENCES 277
103. Official Methods and Recommended Practices of the AOCS, 5th ed., 2nd printing,
AOCS, Champaign, Illinois, 1999.
104. D. Swern, in D. Swern, ed., Baileys Industrial Oils and Fat Products, 3rd ed.,
Interscience Publishers, New York, 1964, pp. 128131.
105. C. E. Lund, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
Publishers, New York, 1948, pp. 58, 62, and 64.
106. R. D. OBrien, Fats and Oils Formulating and Processing for Applications, Technomic
Publishing Co., Lancaster, Pennsylvania, 1998, pp. 14, 4754, 53121, 129131, 168
172, 182183, 188190, 199200, 251253, 258260, 264266, 266290, 327341,
437458, 483, and 525.
107. D. R. Erickson and G. R. List, in T. H. Applewhite, ed., Baileys Industrial Oil and Fat
Products, Vol. 3, 4th ed., John Wiley & Sons, Inc., New York, 1985, pp. 275277.
108. R. D. OBrien, Fats and Oils: Formulating and Processing for Applications, Technomic
Publishing, Lancaster, Pennsylvania, 1998.
109. L. A. Johnson, Report to Cotton Inc., Raleigh, North Carolina, 1981.
110. A. M. Altschul, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
111. J. R. Mauney and J. M. Stewart, Cotton Physiology, The Cotton Foundation, Memphis,
Tennessee, 1986.
112. D. L. Park, L. S. Lee, R. L. Price, and A. E. Pohland, J. Assoc. Off. Anal. Chem., 71, 685
(1985).
113. F. A. Norris, L. J. Hilfman, and C. E. Land, J. Amer. Oil Chem. Soc., 35, 684
(1958).
114. K. C. Rhee, E. W. Lusas, L. R. Watkins, and D. H. Kinard, in Proceedings 37th Oilseed
Processing Clinic, New Orleans, Louisiana, March 1415, 1988.
115. R. A. Rusca and F. L. Gerdes, U.S. Department of Agriculture Circular 651, 1942.
116. O. H. Alderks, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
117. J. P. Galloway, J. Amer. Oil Chem. Soc., 53, 271 (1976).
118. J. A. Ward, J. Amer. Oil Chem. Soc., 53, 261 (1976).
119. P. J. Wan and P. J. Wakelyn, INFORM, 9(12), 11551160 (1998).
120. K. F. Carlson and J. D. Scott, Inform 2, 10341060, 1991.
121. L. R. Watkins, Proceedings 37th Oilseed Processing Clinic, New Orleans, Louisiana,
March 1415, 1988.
122. D. P. French, in Proceedings 37th Oilseed Processing Clinic, New Orleans, Louisiana,
March 1415, 1988.
123. E. M. James, in A. E. Bailey, ed., Cottonseed and Cottonseed Products, Interscience
124. L. S. Crauer, J. Amer. Oil Chem. Soc., 41, 656661 (1964).
125. G. C. Cavanagh, J. Amer. Oil Chem. Soc., 53, 361363 (1976).
126. W. B. Hendrix, J. Amer. Oil Chem. Soc., 61, 13691372, 1984.
127. T. J. Weiss, in Food Oils and Their Uses, AVI Publishing Co., Inc., Westport, Connecticut,
1983, p. 39.
128. T. R. Bessler and F. T. Ortheofer, J. Amer. Oil Chem. Soc., 60(10), 17651766 (1983).
278 COTTONSEED OIL
129. G. C. Cavanagh, Food Processing, 22(4), 3845 (1961).

130. R. C. Hastert, AOCS short course, Phoenix, Arizona, 1988.
131. R. D. OBrien, in Fats and Oils Formulating and Processing for Applications, CRC
Press, Boca Raton, Florida, 2003.
132. R. D. OBrien, in P. J. Wan et al., eds., Introduction to Fats and Oils Technology, 2nd ed.,
133. D. Lampert, in P. J. Wan, ed., Introduction to Fats and Oils Technology, 2nd ed., AOCS
134. F. A. Norris, J. Amer. Oil Chem. Soc., 25(8), 274275 (1947).
135. P. J. Wakelyn and P. K. Adair, in S. S. Koseoglu, K. C. Rhee, and R. F. Wilson, eds.,
Emerging Technologies, Current Practices, Quality Control, Technology Transfer, and
Environmental Issues, Vol. 1, Proc. of the World Conference on Oilseed and Edible Oil
Processing, AOCS Press, Champaign, Illinois, 1998, pp. 305312.
136. P. J. Wakelyn, P. K. Adair, and S. R. Gregory, Oil Mill Gaz., 103(6), 2330 (1997).
137. P. J. Wakelyn, Cotton Gin Oil Mills Press, 92(17), 1215, 18 (1991).
138. P. J. Wakelyn and L. A. Forster, Jr., Oil Mill Gaz., 99(12), 2130 (1994).
139. U. S. EPA, The Toxic Release Inventory (TRI) and Factors to Consider when Using TRI
Data. http://www.epa.gov/tri/2002_tri_brochure.pdf.
140. P. J. Wakelyn, in P. J. Wan, D. OBrien, and W. Farr, eds., Introduction to Fats and Oils
Technology, 2nd ed., AOCS Press, Champaign, Illinois, 2000, pp. 302325.
141. European Union (EU), J. Eur. Commun. L, 157, 00280033 (1988).
142. A. H. Allen, Food Product Design (1996).
143. B. L. Oser and R. A. Ford, Food Technol., 27(1), 64 (1973).
144. R. D. OBrien, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, 5th ed., Wiley,
New York, 1996, pp. 190191.
145. R. D. OBrien, Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, Vol. 3, 5th ed.,
Wiley-Interscience, New York, 1996, pp. 190191.
146. S. F. Riepma, The Story of Margarine, Public Affairs Press, Washington, D.C., 1970, p. 5,
109, 148, and 153.
147. Anonymous, INFORM, 2(4), 373 (1991).
148. C. C. Price, Food Rev., 19(3), 11 (1996).
149. R. J. Bell, R. J., Shortening and Margarine Products, in P. J. Wan, ed., Introduction to
Fats and Oils Technology, American Oil Chemists Society, Champaign, Illinois, 1991,
pp. 183210.
150. W. W. Nawar, in O. R. Fennema, ed., Food Chemistry, 2nd ed., Marcel Dekker,
New York, 1985, p. 139.
151. L. H. Weiderman, J. Amer. Oil Chem. Soc., 55(11), 823 (1978).
152. E. G. Latondress, in D. R. Erickson et al., eds., Handbook of Soy Oil Processing and
Utilization, Am. Soybean Assoc. and Am. Oil Chem. Soc., 1980, pp. 146154.
153. H. B. Patterson, Oleagineux, 28(12), 585 (1973).
154. A. G. Herzing, INFORM, 7(2), 165167 (1996).
155. M. M. Chrysam, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, Vol. 3, 5th
ed., Wiley-Interscience, New York, 1996, pp. 6768 and 9296.
REFERENCES 279
156. A. Moustafa, in D. R. Erickson, ed., Practical Handbook of Soybean Processing and

Utilization, AOCS Press and United Soybean Board, Champaign, Illinois and St. Louis,
Missouri, 1995, pp. 352353.
157. D. M. Prince and M. A. Welshenbach, J. Amer. Diet. Assoc., 98(5), 565569 (1998).
158. J. B. Klis, Food-Technol., 50(2), 124 (1996).
159. J. Giese, Food Technol., 50(3), 130131 (1996).
160. D. Buss, New-Nutrition-Business, 6(10), 710 (2001).
6
Flax Oil and
High Linolenic Oils
Roman Przybylski
University of Manitoba
1. INTRODUCTION
Many species in the Europhorbiaceae and Labiatae families produce seeds with a
high content of oil and contribution of linolenic acid of up to 76% (1). Flaxseed has
been used for years in the production of paints, varnishes, inks, and linoleum. In
food applications, flaxseed is more often used than oil because of its better stability
and because of the presence of fiber, lignans, and a-linolenic acid (ALA), which
have health benefits. Cold pressed flaxseed oil is not considered suitable for
deep-frying, although Chinese use it in stir-frying (2). In this chapter, oilseeds of
flax, perilla, camelina, and chia are discussed as sources of oils with elevated con-
tent of ALA. These oilseeds are produced in industrial quantities and can be con-
sidered as potential sources of new oils with specific nutritional and functional
properties.
281
282 FLAX OIL AND HIGH LINOLENIC OILS
2. FLAX
2.1 Origin
Flax, widely adapted to warm and cool climates, has been cultivated for centuries in
various parts of the world for its stem fiber, linen cloth, and seed. Linseed is an
alternative name used for flax. Crops grown for seed are termed linseed in India
and in the United Kingdom and flaxseed in Canada and the United States, and
flax oil or flax seed is used in many European countries.
Flaxseed/linseed is the annual cultivar of Linum usitatissimum L. Flax is a mem-
ber of the Linaceae family that includes ten genera and more than 150 species (3).
Approximately 200 species of Linum are known (3).
The crops grown for both seed and fiber are generally called dual-purpose flax.
Initially, the same variety was used for both oil and fiber production. Today, oil and
fiber varieties are different and specifically designed to serve the actual end use.
Fiber varieties usually have longer stem, 80120 cm tall, with fewer branches,
fewer seed capsules, and smaller seeds. Although oil type has shorter and heavily
branched stems, 6080 cm tall, with a higher number of seed capsules and larger
seeds.
All registered flax varieties in Canada have a dark brown seed coat. There are
available yellow seed-coated varieties grown in other countries such as the Omega
variety in the United States. Transition to different color is mainly esthetic, lighter
colored flaxseed flour is produced from these seeds, and appearance of the product
is less affected when it is applied as an ingredient.
2.2 Production
More than 60 years ago, the average world production of flaxseed was about
3.4 million metric tons (MMT), which was more than sunflower, 2.5 MMT, and
slightly lower than rapeseed, 3.8 MMT. In the same period, soybean was produced at
a level of 12.6 MMT (4). In those years, flaxseed was the third-most produced oil-
seed in the world by volume. Since then, world production of flaxseed has remained
between 2 and 3 MMT, and the production of other oilseeds has increased consid-
erably (4). In 20002001, world production of flaxseed was 2.34 MMT, with Canada
being the largest producer and exporter of this oilseed (See graph in Canola
chapter).
The total average yearly world production of flaxseed for the past ten years was
2.52 MMT (5). The principal growing areas for flaxseed are Canada, China, India,
Argentina, the United States, the United Kingdom, former USSR, and some
European countries (5). The average contribution of mentioned countries in the
world production of flaxseed is presented in Figure 1. Among mentioned producers,
Canada, China, and India contributed 34.9%, 18.7%, and 11.9%, respectively, to the
world production. The eight main flaxseed producers listed contributed up to 82%
of the total yearly flaxseed production.
FLAX 283
Figure 1. Major world producers of flaxseed (Ten-year average from 1990 to 2000). Production
averaged 2.52 million metric tons/year. aFormer U.S.S.R. Source: Canadian Grains Council,
Statistical Handbook 2001 (5).
Canada is one of the major flaxseed producers and exporters, where a minimal
amount of seeds is crushed to produce flax oil. Flax oil is mainly considered as a
health food product but not a commodity oil. Figure 2 shows yearly production of
flaxseeds in Canada for the past ten years. On average, Canada is producing above
800,000 MT (metric tons) of flaxseed per year (5). Part of this production is low
linolenic acid varieties, which contribute from 10% to 15% to the total production.
Recently, the food industry in North America and Europe has shown an
increased interest in utilization of flaxseed in food product formulations. This is
1200
1100
1000
Metric Tons (x1000)
900
800
700
600
500
400
300
200
19911992 19931994 1995 199619971998 1999 2000 2001
Figure 2. Flaxseed production in Canada. Data include low linolenic flaxseed. Source: Canada
Grains Council Statistical Handbook 2001 (5).
mainly because of the presence of ALA, dietetic fiber, and plant lignans, which
according to scientific evidence provide important health benefits.
2.3. Physicochemical Properties of Flax Oil

Some physicochemical properties of conventional flaxseed oil and low linolenic
varieties are presented in Table 1. The higher specific gravity of 0.935 observed
for flaxseed oil than other vegetable oils can be directly attributed to the high con-
tribution of linolenic acid. It is in line with the specific density of fatty acids that
increases from 0.895 to 0.9038 and to 0.914 for oleic, linoleic, and linolenic acids,
respectively (7).
The amount of polyunsaturated fatty acids (PUFA) affects both melting and
flashpoints of vegetable oils. Melting temperature of oil is directly related to the
melting point of fatty acids, which decreases with unsaturation (7). The flash point
of flaxseed oil is relatively low compared with other vegetable oils; this can be
attributed to a high contribution of PUFA.
Unsaponifiable matter content, saponification value, and iodine value are char-
acteristic for a high contribution of PUFA in the flax oil. The content of unsaponifi-
able matter in flax oil is similar to other vegetable oils.
2.4. Chemical Composition of Flax Oil

Main components of vegetable oils, including flax oil, are triglycerols and usually
contribute more than 90% of all components (Table 1). Minor components in flax
oils were found to be at the similar level as in canola and soybean oils (10). The
presence of chlorophyll in flax oil usually indicates immaturity of flaxseed.
TABLE 1. Properties of Flaxseed Oils (69).
LinolaTM

Parameter Flaxseed Oil Crude RBD
Relative density (20 C/water at 4 C) 0.9250.935 0.921 0.920

Refractive Index (nD 20 C) 1.4751.475 1.4657 1.4665
Melting Point ( C) 20 to 24
Flash Point, min. ( C, open cup) 120135
Viscosity (cp) 46.8 46.4
Iodine Value 182203 142 144
Unsaponifiable Matter (%) 0.11.7 1.2 0.6
Saponification Value (mgKOH/g) 187195 185 185
Phosphorus (ppm) 1.030 300 1.0
Chlorophyll (ppm) 0.01.5 0.4 0.1
Free Fatty Acids (% of oleic) 0.12.0 0.3 <0.02
Triglyceride (%) 9498 9398 9698
RBDRefined, bleached, and deodorized.

FLAX 285
TABLE 2. Composition of Flaxseed and Major Oils (6, 10, Przybylski Unpublished Data).
Component Flax LinolaTM Canola Soybean Sunflower
Fatty Acids (%)

C16:0 5.3 6.1 3.8 11.2 6.0
C18:0 3.3 3.8 1.7 4.1 4.0
C18:1 17.9 15.5 58.2 24.3 16.5
C18:2 14.7 71.3 20.1 54.6 72.4
C18:3 58.7 2.0 9.6 8.3 0.5
SFA 9.0 10.0 6.2 15.6 11.2
MUFA 18.1 17.1 64.2 23.4 16.7
PUFA 72.9 72.9 29.6 61.0 72.1
Tocopherols (ppm)
Alpha 20 15 272 116 613
Gamma 200 200 423 737 19
Delta 7 5 275
Plastochromanol-8 120 110 75
Total 347 330 770 1128 632
Phytosterols (%)
Brassicasterol 1 1 14
Campesterol 27 23 28 18 7
Stigmasterol 8 4 1 15 7
b-Sitosterol 50 54 52 54 58
5-Avenasterol 10 18 5 2 4
Total sterols (g/kg) 2.3 2.2 6.9 2.6 3.1
Abbreviations: Fatty Acids: SFAsaturated; MUFAmonounsaturated; PUFApolyunsaturated; Plasto-

chromanol-8derivative of gamma tocotrienol with longer side chain.
Fatty acid composition of regular flax oil is different from other commercial oils
because of the very high contribution of ALA, usually above 50% (Table 2).
Because of the high content of this unique fatty acid, flaxseed and flax oil are often
used as food supplements, where enrichment with omega-3 fatty acids is needed.
This fatty acid is susceptible to oxidation; it oxidizes 2040 times faster than oleic
acid and 24 times faster than linoleic acid (8). This property makes the oil a good
material for paint and plastic production where fast oxidation is required. Flax oil
contains low amounts of saturated fatty acids (SFA) compared with low linolenic
flax oil (Linola), soybean, and sunflower oils; however, it is higher than canola
oil (Table 2). Canola oil contains the lowest amount of SFA among all commercial
oils.
The contribution of linolenic acid in flaxseed oil showed a wide range and was
affected by the growing conditions. Flax varieties grown in Western Canada, aver-
age from 495 samples analyzed, contained 5% palmitic acid (16:0), 3% stearic acid
(18:0), 17% oleic acid (18:1), 15% linoleic acid (18:2), and 59% linolenic acid
(18:3) (11). Although similar varieties were grown in North Dakota, the 11 cultivars
assessed showed the following fatty acid composition: 56% of 16:0, 36% of 18:0,
1929% of 18:1, 1418% of 18:2, and 4552% of 18:3 (12).
Cool temperatures during the 1025 days after flowering are the main cause for
higher amounts of linolenic acid in flax oils (14). For the same reason, flaxseed
grown in the Canadian prairies, northern latitude, produce oils with higher levels
of polyunsaturated fatty acids and lower contributions of oleic acid and saturated
fatty acids. This phenomenon was also observed for other oilseeds such as sun-
flower, canola, and soybean (7, 13, 14). Similarly, a wide variation in fatty acid
composition in Australian flaxseed samples was observed: 1325% of 18:1 and
4664% of 18:3 (6).
Analysis performed on varieties of flaxseeds collected from different flax grow-
ing regions of the world and later grown in Morden, Manitoba, Canada, showed
even wider distributions of oleic acid 1460%, linoleic acid 321%, and ALA
3172% (13). All of these data indicate that within flax, there is a wide distribution
of fatty acids, and this variability can be used for developing specialty oils based on
traditional breeding and to avoid GMO oils.
Flaxseed oils contain much lower amounts of tocopherols, half of the amount
present in sunflower and canola oils and one-third of that present in soybean oil
(Table 2). A lower content of these antioxidants makes these oils even more suscep-
tible to oxidation. Gamma-tocopherol was found as the main tocopherol in flax oils,
with a contribution of about 80% to the total amount. This makes flax oil compar-
able with soybean oil. Among unique antioxidants detected in flax oils was plasto-
chromanol-8. This compound is a derivative of gamma tocotrienol with twice as
long unsaturated side chain. Plastochromanol-8 was found to be a more efficient
antioxidant than any tocopherols isomer (15). A low content of tocopherols in flax-
seed did not make them more susceptible to oxidation; experiments showed that
milled flaxseed could be stored for 28 months at ambient temperatures without
measurable changes in oxidation products. This can be attributed to the presence
of antioxidants other than tocopherols in the seeds (16).
Sterols or phytosterols are present in flax oils at a level lower than those in many
vegetable oils, 2.3 mg/g in flaxseed oil versus 4.1 to 6.9 mg/g in other oils (Table 2).
The composition of sterols was similar to other oils, where b-sitosterol was the
main component followed by campesterol and 5-avenasterol. Brassicasterol was
found in trace amounts in flax oil. This phytosterol is characteristic to plants from
the Brassica family and often is used as a marker for oil adulteration (Table 2).
As discussed above, triacyglycerols are the main components of vegetable oils
and the composition of flax acylglycerols is presented in Table 3.
As expected from fatty acid composition, the main triacylglycerols contain lino-
lenic acid in their molecules and 84% of all triacylglycerols have this acid in their
structure (Table 3). Among them, 21% of total acylglycerols contained three ALA
in molecule, second by contribution were acylglycerols with two ALA, and linoleic
acid had the second-most abundant fatty acid present in the flax oil (17).
Flaxseed is the richest source of plant lignans containing 75800 times more
than that in other oilseeds, cereals, legumes, fruits, and vegetables (18). These plant
origin components act in mammalians as hormone-like phytoestrogens. Lignans are
compounds with a dibenzylbutane skeleton, which have been found in many higher
plants (1820). Plant lignans, namely, secoisolariciresinol diglycoside (SDG) and
FLAX 287
TABLE 3. Composition of Triacyglycerols

in Flaxseed Oil (17).
Triacyglycerols1 Contribution (%)
PLnLn 7.6
PLLn 6.7
PLL 1.5
POL 1.6
LnLnLn 20.9
LLnLn 13.8
LLLn 3.7
OLnLn 8.4
LLL 0.9
OLLn 5.3
OLL 0.9
SLLn 1.1
OOL 3.4
OOLn 7.3
POLn 4.0
SLnLn 3.2
POL 1.6
PLL 1.5
OOO 3.3
1
Abbreviations of fatty acid: Ppalmitic; Lnlinolenic;
Llinoleic; Ooleic; Sstearic.
matairesinol (MAT), are the main compounds among flaxseed lignans. Both are
structurally different from animal and human lignans, enterodiol (ED) and entero-
lactone (EL). Mammalian lignans are formed by intestinal microorganisms from
plant precursors (Figure 3). The concentration of mammalian lignan precursors is
measured by adding a particular food ingredient to the model of the intestinal
microorganism and assessing the amounts of released ED and EL (18). Similarly,
excretion of animal lignans in urine may be measured (18, 19). Figure 4 shows urin-
ary excretion of ED and EL when different plant components were included in the
diet. Flax oil is the second dominant source of mammalian lignans excreted after
flaxseed, in far higher amounts than other oils, oilseeds, and cereals.
Lignans from flaxseed have been shown to reduce mammary tumor size by more
than 50% and tumor number by 37% in carcinogen-treated rats (19, 20). Further-
more, it has been suggested that lignans have antimiotic, antiestrogenic, antiviral,
antibacterial, antifungal, and antioxidant properties (2033).
The presence of plant lignans in flax oil makes it nutritionally more valuable
than any other oil. When high levels of ALA and linoleic acid are considered in
the whole equation, flaxseed oil serves as the best oil in terms of its nutritional
and health value.
The Food and Drug Administration (FDA) regulations allow inclusion of flax-
seed in food products, but the amount allowed is limited to 12% (34).
Plant Lignans
O
H3CO H3CO
OR
O
OR
HO HO
OCH3 OCH3
OH OH
Secoisolariciresinol Matairesinol
Diglycoside (SDG)
Bacterial Bacterial
Fermentation Fermentation
O
HO HO
OH
O
OH
Bacterial
Fermentation
OH OH
Enterodiol (ED) Enterolactone (EL)
Mammalian Lignans
Figure 3. Mammalian lignan formation in digestive tract and their plant precursors (19).
Figure 4. Total excretion of human lignans in the urine of rats after diet was supplemented with
various foods (18).
FLAX 289
2.5. Low Linolenic Flaxseed Oil

Low linolenic acid varieties with yellow-seed coat flax trademarked Linola were
developed by the Commonwealth Scientific and Industrial Research Organization
(CSIRO) in Australia and distributed elsewhere under this name by United Grain
Growers, Canada (4). The Linola seed color has been changed to yellow to make
it distinguishable from the traditional flaxseed dark brown color. The generic com-
mon name Solin has been assigned by Flax Council of Canada for all low linolenic
flax varieties produced in Canada. Developmental work on Solin (Linola is a brand
name within Solin family) is continuing mainly to reduce saturated fatty acid and to
increase linoleic acid content above the 70% level and to increase the content of
antioxidants as well as to enhance nutritional properties of the meal.
The new oilseed crop is grown wherever flax and linseed varieties are currently
cultivated (35, 36). The climate in northern Europe is highly suitable for production
of Linola, where sunflower and corn/maize cannot be produced. Linola seed can be
processed by existing crushing plants using similar processing parameters. Linola
meal is used for ruminant feed in the same way as linseed meal.
The fatty acid composition of the new crop has been modified, and the level of
linolenic acid has been reduced from over 50% to 2% (6). This greatly improves
oxidative stability of the oil, which by fatty acid composition is very close to sun-
flower and soybean oils (Table 2). Linola has been found to be more resistant to
oxidation than regular flax oil, and its stability is comparable with soybean, canola,
and sunflower oils (Przybylski, unpublished data).
Refining of crude Linola oil by conventional steps, namely, degumming, alkali
refining, bleaching, and deodorization, produces colorless and odorless oil, which
has good oxidative stability (9). In addition, properties of crude and refined,
bleached, and deodorized (RBD) Linola oil are comparable with other commodity
oils (Table 1).
The FDA granted Generally Recognized as Safe (GRAS) status for Solin/Linola
oil in 1998 (38). This oil can be used as an ingredient in food product formulations
such as salad oil, cooking, and frying oil, and in fat phase to formulate margarine,
spreads, and shortenings (19, 37).
Because of several beneficial nutritional properties, mainly related to the high
level of linoleic acid and lignans, there is a growing interest to use Linola seeds
and oil in bakery and confectionery applications. The golden-yellow-colored Linola
seeds can serve as an attractive and appealing topping for baking goods. It seems
evident that Linola/Solin seed and oil can have promising future applications in
food products (35).
2.6. Processing of Flaxseed and Oil

Flaxseed is covered with fibrous hull accounting for 25 to 45% of the seed weight
and contains 27% by weight of water-soluble carbohydrates. These components
called mucilage can interfere during processing (38). Flaxseed contains approxi-
mately 25% protein, 10% moisture, and 3545% of oil (6, 38, 11). In immature
seeds, cyanogenic glucosides such as linamarin, linustatin, and neolinustatin can be

present at the level of 200650 mg/100 g of seeds (9). Enzyme linase is always pre-
sent in flaxseed, and it decomposes glucosides to many products, including
hydrocyanic acid, one of the most toxic substances. Newly developed varieties of
flax have lower amounts of glucosides in the seed. During processing, small
amounts of glucoside can be transferred into oil, whereas these compounds are
water-soluble.
Flaxseed contains a high amount of oil, but expressing oil from it is difficult and
often double pressing is required to efficiently remove oil from the seeds. Proces-
sing steps for flax oil production are presented in Figure 5. Before crushing, cleaned
seeds are tempered to achieve a moisture level of 9.5% to 10%, this will minimize
the formation of fine particles when seeds are cracked or flaked and will maximize
removal of oil from them. Moisturized seeds are passed through sets of corrugated
and smooth rolls to be cracked and flaked, respectively. From the next processing
step, production of flax oil is differentiated from that for Solin/Linola oil (7). The
flax oil for human consumption is cold-pressed, and further purification of oil is
not applied. According to industry standards, cold pressing is achieved when the
temperature of oil coming from the extruder does not exceed 35 C and pressing
is performed under protection from oxygen, usually under a blanket of nitrogen.
Good practice requires utilization of expellers, which have the ability to cool parts
of the press, which are in contact with seeds and oil to control the temperature
during processing (38).
Figure 5. Processing of flaxseed to produce cold-pressed flax oil.

FLAX 291
Oil from expeller is filtered, packaged under nitrogen or other neutral gas into bot-
tles protecting from light exposure, and ready for distribution. Flax oil is very suscep-
tible to oxidative deterioration, and treatment to eliminate oxygen needs to be
applied. On the North American continent, flax oil is considered as a health food oil.
When flax oil is processed for industrial use, standard processing steps are
applied as described in Figure 6. Flaxseeds are tampered and then flaked, passing
through a set of smooth rolls. Flaked seeds are sent to a cooker where they are
heated to a temperature of 80100 C to inactivate enzymes and help release the
oil during pressing. At this stage, formation of toxic substances is prevented. The
cooked seeds are transferred to the expeller, and expelled oil through filtration is
placed in a storage tank, where it is combined with oil from solvent extraction.
Cake/meal after pressing is fed to the solvent extractor, where hexane is used as
Figure 6. Processing of flaxseed to produce refined, bleached, and deodorized flax oil.
a solvent. From the extractor, cake is moved to the desolvatizer where the solvent
is removed at 100 C. Meal is then cooled and used as an animal feed ingredient.
Combined oils are purified by the standard refining process, typical to all vegetable
oils (7). Degumming is applied to remove phospholipids, refining to lower the
content of free fatty acids, bleaching to eliminate chlorophylls and other pigments
as well as to decompose hydroperoxides, and deodorizing to make the oil
odorless through elimination of oxidation products (Figure 6). Processing of
low linolenic flaxseed oil is similar to that described for flax oil and other com-
modity oils.
3. PERILLA OIL
3.1. Origin and Application

Perilla, Perilla frutescens, L. Britton, is a member of the mint family, Lamiaceae
(Labiatae). This plant is a common annual weed in the eastern United States (1). In
Asia, perilla is considered a commercial crop where seeds are used to produce oil
and plant parts are used as garnish, flavoring, and sources of nutritional components
in combination with cereals or vegetables. In the United States, perilla food pro-
ducts are available in the Korean ethnic markets, and red-leafed plants are used
in landscaping. After the Second World War, the United States imported perilla
oil, which was used as a drying oil (1). Perilla plant and seed is used in Asia as
seeds for birds and human consumption; seed oil is used as a fuel, a drying oil,
or a cooking oil; leaves are used as a pot-herb, for medicine, food coloring, flavor-
ing dishes, and source of functional nutrients; foliage is distilled to produce an
essential oil for flavoring.
Wilson et al. (39) isolated the toxin, perilla ketone, which causes pulmonary
edema (fluid in the lung cavity) in many animal species, although not in pigs
and dogs (40). In Japan, 2050% of long-term workers in the perilla industry devel-
oped dermatitis on their hands because of contact with perillaldehyde (41). Small
amounts of these components have been detected in perilla oil where it works as an
efficient antioxidant.
Perilla was never grown commercially as an oilseed in the United States; how-
ever, several agronomists have investigated the crop (42, 43). Rabak and Lowman
(43) determined that perilla is well adapted to the climate of the southeastern
United States; it would be unprofitable to cultivate it, unless seed shattering can
be controlled. Seed yields ranged from 220 to 1400 kg/ha in Illinois (44), 1020
to 1440 kg/ha in Korea (45), and 1110 to 1670 kg/ha in Japan agricultural produc-
tion (41). Perilla was also experimentally grown as a crop in many parts of the
British Empire (46, 47). Production of perilla seeds and oil has been continued
in Korea for a long time (48, 45). Annual production of perilla seed is approxi-
mately 40,000 MT, and perilla oil is the third largest among edible oils used in
the Korean market (49). Perilla plant and seed is widely used in Asian countries
as food ingredients, including Japan, China, and India.
PERILLA OIL 293
3.2. Perilla Seed and Oil

The seed of perilla contains 3151% of oil, which is similar in composition to
flaxseed oil, with a higher contribution of PUFA of over 70% (Table 4). The
oil is highly unsaturated, with an iodine value of 192208-g iodine /100-g oil
(Table 4). Perilla oil contains over 60% linolenic acid with equal amounts of
both linoleic and oleic acids (Table 4). Specific gravity of this oil is higher than
flax oil because of a higher contribution of PUFA. Other physical parameters of
this oil reflect the composition of its fatty acids.
TABLE 4. Composition and Properties of Perilla, Camelina, and Chia Oils.
Parameter Perillaa Camelinab Chiac
Fatty Acids (%)

C16:0 7 6 6
C18:0 2 2 3
C18:1 14 13 7
C18:2 17 16 20
C18:3 61 39 63
Saturated 8 12 9
Monounsaturated 14 34 8
Polyunsaturated 78 54 83
Tocopherols (ppm)
a-Tocopherol 31 46
g-Tocopherol 461 420
d-Tocopherol 7 10
Total 499 500
Lipid Classes (%)
Sterol Esters 2
Glycerides 91 97
Glycolipids 4 2
Phospholipids 2 0.9
Sterols (%)
Cholesterol 5
Brassicasterol 4
Campesterol 25
Stigmasterol 3
b-Sitosterol 52
5-Avenasterol 11
Total Sterols (mg/kg) 3604
Physicochemical Properties
Refractive Index (nD 20 C) 1.4761 1.4698 1.4753
Specific Gravity (at 15.5 C/15.5 C) 0.937 0.925 0.936
Iodine Value 192208 127155 190199
Saponification Value (mgKOH/g) 188197 180190 180192
Unsaponifiable Matter (%) 1.31.5 1.21.5 1.11.3
Oil Content (%) 3550 3542 3240
Protein Content (%) 1728 2530 2030
Camelina contains 15% of eicosenoic acid (C20:1) and 35% of erucic acid (C22:1).
Source: a(49); b(50, 51); c(5054).
The amount of tocopherols in perilla oil is higher compared with flax oil, and a
similar contribution of gamma-tocopherol, above 90%, was observed (Table 4).
Shin and Kim (49) analyzed perilla oil for lipid composition and established that
it contained more than 90% triacylglycerols, 4% glycolipids, and 2% of each phos-
pholipids and sterol esters.
Perilla oil has been used as a drying oil in paints, varnishes, linoleum, printing
ink, lacquers, and for protective waterproof coatings on cloth. It has also been used
for cooking and as fuel (56). The meal produced after oil extraction is often used as
an animal feed ingredient.
3.3. Perilla Oil Processing

Perilla oil in Korea is processed like other cold-pressed oils, where pressing and
filtration are the main processing steps. To improve the flavor of perilla oil used
in food applications, roasting of seeds is practiced. This will provide oil with a
distinctive roasted, nutty flavor and improved stability. Roasting of perilla seeds
is often applied in Korea and China (57). Kim et al. (57) analyzed different
parameters of roasting and established that temperature above 170 C provided
the best flavor and stability for the oil. Nonenzymatic browning components are
mainly responsible for flavor and antioxidant activity (52). When perilla oil is pro-
duced for the industrial applications, additional processing such as refining, bleach-
ing, and deodorization is carried out (58).
4. CAMELINA
Standard oilseed crops are not often suitable to marginal lands where factors such
as low moisture, low fertility, and saline soils play an important role in the possible
crop to be grown. In recent years, there has been increasing interest in developing
agronomic systems with low requirements for fertilizer, pesticides, and energy,
which provide better soil erosion control than conventional systems. Camelina
can grow in these extreme conditions and provide oilseed with enhanced nutritional
value (59, 60).
4.1. Origin
Camelina sativa (L.) Crantz., plant from the Brassicaceae family, known as false
flax, linseed dodder, and Gold-of-Pleasure, originated in the Mediterranean area
and Central Asia (61). Seeds are small (0.7 mm 1.5 mm), pale yellow-brown,
oblong, rough, and with a ridged surface. Camelina is listed as being adapted to
the flax-growing region on the Prairies, in Europe, and other countries (59, 62).
It is primarily a minor weed in flax, which does not have seed dormancy (63).
Camelina is short-seasoned, 85100 days, so it could be incorporated into double
cropping systems during cool periods in warmer environments (55).
Cultivation of camelina probably began in Neolithic times, and by the Iron Age
in Europe, when the number of crop plants approximately doubled, this crop was
CAMELINA 295
commonly used as an oil-supplying plant. Cultivation, as evidenced from carbo-

nized seed, has been shown to occur in regions surrounding the North Sea during
the Bronze Age (64). Camelina monoculture occurred in the Rhine River Valley as
early as 600 B.C. Camelina probably spread in mixtures with flax and as monocul-
tures, similar to small grains, which also often spread as crop mixtures. It was cul-
tivated in antiquity from Rome to Southeastern Europe and the Southwestern Asian
(64).
Camelina production declined during medieval times because of unknown fac-
tors, but it continued to coevolve as a weed with flax, and this is the possible intro-
duction of it to the Americas. Like rapeseed oil, camelina oil has been used as an
industrial oil after the industrial revolution (64). The seeds have been fed to caged
birds, and the straw has been used for fiber. There has been scattered production of
camelina in Europe in modern times, mostly in Germany, Poland, and the USSR. In
the 1980s, breeding and germplasma screening were applied to modify fatty acid
composition and the content of glucosinolates in camelina seeds (6569).
Camelina has been evaluated in Canada, North Dakota, and Minnesota for its
agronomical performance (63, 70, 50). Recent interest in the species is mainly
because of the demand for alternative low-input oilseed crops with the potential
for food and nonfood utilization of the seed oil (60, 71). Unique agronomic features
such as compatibility with reduced tillage and cover crop and competitiveness with
weeds or winter surface seeding showed suitability of camelina for sustainable agri-
culture systems. Furthermore, the species has a potential as a low-cost crop for
green manuring (60).
Long-term yield of camelina cultivars in North America has been averaging
from 1100 to 1200 kg/ha with a maximum of about 2000 kg/ha. It should be noted
that the yield of many commodity oilseeds, especially B. napus, has been improved
through plant breeding, whereas camelina has not been modified yet (63).
4.2. Seed Composition

The oil content of camelina seed ranges from 29% to 45% in North American
crops, and in Germany, it is between 37% and 44%. The seed protein content varies
from 23% to 30% (60, 50, 71, 72). Camelina protein content and composition is
similar to flax, although higher sulfur content has been observed for camelina oil
(63). Camelina meal is comparable with soybean meal, containing 4547% crude
protein and 1011% fiber (73). Like other cruciferous plants, camelina meal con-
tains glucosinolates at levels of 1520 mmol/g (74). This is a low content of gluco-
sinolates compared with other brassicaceous species, hence making the utilization
of meals easier (73, 75).
4.3. Fatty Acid Composition and Use of the Oil

Camelina oil has a unique fatty acid pattern and is characterized by a linolenic
acid (C18:3) content ranging from 30% to 40%, eicosenic acid (C20:1) content
of around 15%, and less than 4% erucic acid (21). The fatty acids in camelina oil are
primarily unsaturated, with only about 12% being saturated (Table 4). About 54%
of the fatty acids are polyunsaturated, primarily linoleic (18:2) and linolenic (18:3),
and 34% are monounsaturated, primarily oleic (18:1) and eicosenoic (20:1).
The fatty acid composition of camelina oil can be influenced by both environ-
ment and variety, although the effects detected were small. Nine varieties were
tested, and the maximum differences between oleic, linoleic and linolenic acid
levels were 3%, 2.4%, and 2.2%, respectively (76). Also, a 2% less linolenic
acid was observed in camelina grown during a dry warm year compared with the
normal year. Although these differences are statistically significant, they are rela-
tively small in absolute terms and have no significant effect on the properties of the
extracted oil (68, 50, 76).
With its high contribution of polyunsaturated fatty acids, mainly linoleic and
linolenic, and relatively low saturated fatty acid content, camelina oil could be con-
sidered a high-quality edible oil. Camelina oil is less unsaturated than flax oil but
more than sunflower or canola oils (Tables 2 and 4). This oil seems to be unique
among vegetable oils in having a high content of 11-eicosenoic acid. Most of the
camelina lines assessed contain 24% erucic acid (Table 4), which is higher than
the maximum limits for canola-quality rapeseed oil. However, screened germplasm
of camelina showed that lines with zero erucic acid content are available and,
through plant breeding, zero erucic varieties can be obtained.
Plant sterols identified in this oil consist mainly of b-sitosterol and campesterol
(Table 4). About 4% brassicasterol was detected in the oil, which is typical for
Brassica family plants (51). The total content of sterols in oil is comparable with
other commercial oils (Tables 2 and 4). The presence of cholesterol in camelina oil
makes it unique among vegetable oils, where only a trace has been detected in some
tropical oils (51).
Composition and content of tocopherols in camelina oil was similar to perilla
oil, where more than 80% of all tocopherols were gamma isomer (Table 4). Alpha
and delta tocopherols were detected as minor antioxidants (77). The total content of
tocopherols was comparable with perilla oil, and higher than that in flax oil (Tables
4 and 2). The total content of tocopherols in camelina oil is higher than canola, flax,
soybean, and sunflower.
4.4. Processing of Camelina Seed, Oil Stability, and Utilization

Cold-pressed camelina oil had an attractive yellow color, a mustard-like taste, and a
characteristic pleasant odor. This type of flavor is acceptable in India and other
Asian countries, but in Europe and North America, it is difficult to find acceptability
among consumers, mainly because of a different expectation from vegetable oils.
However, commercial camelina oil needs to be refined and deodorized to produce
an odorless and colorless product as expected by consumers (76). Crude camelina
oil, refined following typical steps as described for flax oil (Figure 6), afforded a
product similar to typical commercial oils (76).
CAMELINA 297
To establish storage stability of camelina oil, an accelerated Schaal Oven storage

test was carried out at 65 C with crude and refined canola and linseed and camelina
oils (76). Conjugated dienes, peroxide, and p-anisidine values were determined.
The results indicated that the storage stability of camelina oil was similar to flax
oil, but it was less stable than canola oil. Crude camelina oil showed a higher
oxidative stability than the refined product (76). During storage, refined camelina
oil had a 30% higher peroxide level when compared with crude camelina oil (76).
Comparison with fish oil, which is rich in omega-3 fatty acids, proved that camelina
oil is much more resistant to oxidative deterioration than fish oil (76). At room
temperature, crude camelina oil was far more stable than could be expected
from its high linolenic acid content. This unusual oxidative stability can be attrib-
uted to the presence of natural antioxidants. However, the content of tocopherols
discussed above was in the middle range compared with other commercial
oils but slightly higher than that of flax oil (Tables 2 and 4). Oxidative stability
is not only related to the content and composition of tocopherols, but also to pre-
sence of other components, such as phenolic acids and polyphenols. The content of
antioxidants in oils is also affected by the processing, and the amounts of antiox-
idants can be lowered even by 50% when particular processing conditions are
applied (15).
The frying performance of camelina oil was compared with soybean oil and
assessed under deep frying conditions. Oil deterioration was monitored by asses-
sing changes in viscosities, free fatty acids, p-anisidine values, and the formation
of oxidized triacylglycerols (76). During the first 5 days of frying, camelina oil
deterioration was similar to that of soybean oil. After that time of frying, camelina
oil deteriorated much faster than soybean oil, probably because its antioxidants
were depleted. In fact, after 7 days of frying, the levels of oxidized triacylglycerols
in camelina oil reached the level permitted in Europe, 25%, and in soybean the
amount of these components was at 14% (76). Similarly, viscosity of camelina
oil increased 100% by the end of the heating period, whereas in soybean oil, it
increased only by 30%. Total carbonyl level, measured by p-anisidine values,
was three times higher in camelina oil than in soybean oil. In addition, deterioration
of camelina oil during 5 days of potato frying caused formation of the strong and
objectionable paint-like flavor (76).
Refined camelina oil was blended into fat phase to produce margarines and
spreads enriched in omega-3 fatty acids. The resulting spreads had physical proper-
ties similar to a product based on typical commercial oils. The stability of the new
product was satisfactory, and off-flavors were not detected after 6 months of storage
(76).
Camelina oil was also included in formulation of salad dressings. Produced dres-
sings showed a similar stability to conventional products during several months of
storage at ambient temperature without off-flavor formation (76).
Taking into consideration that camelina oil production will be less expensive and
the oil is more stable than fish oil, this oil can be an excellent ingredient to enrich
spreads, margarines, and other fat-containing food products, in omega-3 fatty acids,
and by this way change the ratio of omega-3 to omega-6 fatty acids.
5. CHIA
5.1. Origin
Chia (Salvia hispanica L.) is an annual herbaceous plant from the mint family,
Labiatae, and it is native to southern Mexico, northern Guatemala, and can be
grown in South America and the Southwestern United States (52). This plant
was used by the Aztec and other tribes of Central America as an important crop
not only for food, but also for medicine and paint. Chia oil is a century-old ingre-
dient that has been rediscovered today as a potential ingredient for cosmetic and
food industries (52). Although chia has been cultivated for several centuries, pre-
sently it is cultivated only in some states in Mexico. The total area cultivated is less
than 450 hectares per year. Trials to adopt this cultivar to other regions of America
have been done with positive results (52). Chia seeds and oil are available on the
American continent in health food stores.
5.2. Oil and Seed Composition

Chia seed contains 2540% oil and 1830% protein. The chia meal is high in pro-
tein and fiber similar to flaxseed and soybean (52, 53). Chia seed, oil, and meal can
be used as ingredients with high nutritional value for human food and animal feed.
Chia seed contains mucilage and water-soluble fiber, may possibly contain lignans,
and is similar to flax (53). Trials conducted in 1995 and 1996 showed yield and
oil contents to be affected by growing conditions and harvested yields were up to
1500 kg/ha (52).
Chia oil is high in polyunsaturated fatty acids, particularly a-linolenic acid; the
content of this fatty acid is higher than flax oil (Table 4). Linoleic acid is the sec-
ond-most abundant acid in chia with a contribution of 1726%, which gives PUFA
content of 83%, the highest amount among edible oils. Additionally, chia oil has the
lowest content of saturated fatty acids (Tables 2 and 4).
The physical properties of chia oil are similar to perilla and camelina with
the same effect of PUFA discussed above. Lipid class composition in chia oil is
also typical for vegetable oils where triacylglycerols are the main components
(Table 4)(52).
REFERENCES
1. J. M. Hagemann, F. R. Earle, and I. A. Wolff, Lipids, 2, 371 (1967).

2. Q. Pan, Flax Production, Utilization and Research in China, 53, 5963 (1990).
3. T. P. Freeman, in S. C. Cunnane and L. U. Thompson, eds., Flaxseed in Human Nutrition,
4. T. L. Krawczyk, Inform, 10, 1029 (1999).
5. Canada Grains Council, Statistical Handbook, 2001.
6. A. G. Green and D. R. Marshall, Aust. J. Agric. Res., 32, 599 (1981).
REFERENCES 299
7. N. A. M. Eskin, B. E. McDonald, R. Przybylski, L. J. Malcolmson, R. Scarth, T. Mag,

K. Ward, and D. Adolph, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, Wiley,
New York, 1996, p. 1.
8. E. N. Frankel, Trends Food Sci. Technol., 4, 220 (1993).
9. A. G. Green, and P. J. C. Dribnenki, Lipid Technol., 6, 29 (1994).
10. F. D. Gunstone, J. L. Harwood, and F. B. Padley, eds., The Lipid Handbook, Chapman and
Hall, London, 1994.
11. D. R. Declercq, J. K. Daun, and K. H. Tipples, in Crop Bulletin, Canadian Grain
Commission, Winnipeg, Manitoba, Canada, No. 202, 1992, p. 1.
12. N. S. Hettiarachy, G. A. Hareland, A. Ostenson, and G. Baldner-Shank, in Proceedings of
the 53rd Annual Flax Institute of the US Meeting, Fargo, North Dakota, 1990, p. 36.
13. D. G. Dorrell, Fette Seifen Anstrichm., 77, 258 (9175).
14. A. G. Green, Crop Sci., 26, 961 (1986).
15. D. Olejnik, M. Gogolewski, and M. Nogala-Kalucka, Nahrung, 41, 101104 (1997).
16. N. A. M. Eskin and R. Przybylski, in N. A. M. Eskin and D. S. Robinson, eds., Food Shelf
Life Stability, CRC Press, Boca Raton, Florida, 2001, p. 175.
17. R. B. Tarandjiiska, I. N. Mrekov, B. M. N. Damyanova, and B. S. Amidzhin, J. Sci Food
Agric., 72, 403 (1996).
18. L. U. Thompson, P. Robb, M. Serraino, and F. Cheung, Nutr. Cancer, 16, 43 (1991).
19. M. Axelson, J. Sjovall, B. E. Gustafsson, and K. D. R. Setchell, Nature, 298, 659 (1982).
20. K. D. R. Setchell, in S. C. Cunnane and L. U. Thompson, eds., Flaxseed in Human
Nutrition, AOCS Press, Champaign, Illinois, 1995, p. 82.
21. L. U. Thompson, S. E. Rickard, L. J. Orcheson, and M. M. Seidl, Carcinogenesis, 17, 1373
(1996).
22. L. U. Thompson, M. M. Seidl, S. Rickard, L. Orcheson, and H. H. S. Fong, Nutr. Cancer,
26, 159 (1996).
23. W. R. Phipps, M. C. Martini, J. W. Lampe, J. L. Slavin, and M. S. Kurzer, J. Clin.
Endocrinol., 77, 1215 (1993).
24. Y. Mousavi and H. Adlercreutz, J. Steroid. Biochem. Mol. Biol., 41, 615 (1992).
25. M. E. Martin, M. Haourigui, C. Pelissero, C. Benassayag, and E. A. Nunez, Life Sci., 58,
429 (1996).
26. T. Fotsis, M. Pepper, and H. Adlercreutz, Proc. Natl. Acad. Sci. USA, 90, 2690 (1993).
27. T. Hirano, K. Fukuoka, K. Oka, and Y. Matsumoto, Cancer Invest., 9, 145 (1991).
28. K. Prasad, Mol. Cell Biochem., 168, 17 (1997).
29. G. Block, B. Patterson, and A. Subar, Nutr. Cancer, 50, 207 (1992).
30. D. M. Parkin, C. S. Muir, S. L. Whelan, Y. Gao, J. Ferlay, and J. Powell, in Cancer
Incidence in Five Continents, Vol VI, International Agency for Research on Cancer, Lyon,
France, 1992, p. 865.
31. H. Adlercreutz, T. Fotsis, and J. Lampe, Scand. J. Clin. Lab. Invest., 53, 5 (1993).
32. H. Adlercreutz, H. R. Heikkinen, and M. Woods, Lancet, 2, 1295 (1982).
33. K. D. R. Setchell and H. Adlercreutz, in I. R. Rowland, ed., Role of the Gut Flora in
Toxicity and Cancer, Academic Press, London, 1988, pp. 315345.
34. J. E. Vanderveen, in S. C. Cunnane and L. U. Thompson, eds., Flaxseed in Human
Nutrition, AOCS Press, Champaign, Illinois, 1995, p. 363.
35. B. F. Haumann, Inform, 1, 934 (1990).

36. E. A.Weiss, Oils & Fat Int., 9, 23 (1993).
37. Inform, 9, 628 (1998).
38. P. P. Kolodziejczyk and P. Fedec, in S. C. Cunnane and L. U. Thompson, eds., Flaxseed in
Human Nutrition, AOCS Press, Champaign, Illinois, 1995, p. 261.
39. B. J. Wilson, J. E. Garst, R. D. Linnabary, and R. B. Channell, Britton. Sci., 197, 573
(1977).
40. J. E. Garst, W. C. Wilson, N. C. Kristensen, P. C. Harrison, J. E. Corbin, J. Simon, R. M.
Philpot, and R. R. Szabo, Anim. Sci., 60, 248 (1985).
41. N. Okazaki, M. Matsunaka, M. Kondo, and K. Okamoto, Skin Res., 24, 250 (1982).
42. H. A. Gardener, U.S. Cir., 52, 1 (1917).
43. Rabak, F. and M. S. Lowman, in Perilla, USDA, Agriculture Research Administration,
Bureau of Plant Industry, Soils, and Agricultural Engineering, Beltsville, Maryland, 1945.
44. R. O. Weibel and W. L. Burlison, Soybean Dig., 8, 14 (1948).
45. I. S. Choi, S. Y. Son, and O. H. Kwon, Korean Soc. Hort. Sci., 25, 68 (1980).
46. Imperial Institute, Bul. Imp. Inst., 18, 479 (1920).
47. Imperial Institute, Bul. Imp. Inst., 24, 205 (1926).
48. Yu, I. S. and S. K. Oh, Res. Rpt. Off. Rural Dev., 18, 187 (1976).
49. H. S. Shin and S. W. Kim, J. Amer. Oil Chem. Soc., 71, 619 (1994).
50. E. C. Leonard, Inform, 9, 831 (1998).
51. V. K. S. Shukla, P. C. Dutta, and W. E. Artz, J. Amer. Oil Chem. Soc., 79, 965 (2002).
52. W. Coates, and R. Ayerza, J. Amer. Oil. Chem. Soc., 75, 1417 (1998).
53. A. A. Bushway, A. M. Wilson, L. Houston, and R. J. Bushway, J. Food Sci., 49, 555 (1984).
54. R. Ayerza, J. Amer. Oil Chem. Soc, 72, 1079 (1995).
55. H. S. Park, J. G. Kim, and M. J. Cho, J. Korean Agric. Chem. Soc., 24, 224 (1981).
56. Publications and Information Directorate, in The Wealth of India, Vol.7, New Delhi, India,
1966.
57. S. J.Kim, H. N. Yoon, and J. S. Rhee, J. Amer. Oil Chem. Soc., 77, 451 (2000).
58. S. J. Kim, S. H. Song, H. N. Yoon, and U. Y. Kong, Inform, 6, 514 (1995).
59. National Research Council (NRC), in Alternative Agriculture/Committee on the Role of
Alternative Farming Methods in Modern Production Agriculture, National Academy
Press, Washington, D.C., 1989.
60. D. H. Putnam, J. T. Budin, L. A. Field, and W. M. Breene, in J. Janick and J. E. Simon, eds.,
New Crops, Wiley, New York, 1993, pp. 314322.
61. J. Schultze-Motel, Archaeo-Physika, 8, 267 (1979).
62. North Central Regional Technical Committee NC-121, Weeds of the North Central States,
North Central Regional Res. Pub. 281, Bul. 772, Agr. Expt. Sta., Univ. Illinois, Urbana-
63. R. G. Robinson, Minnesota Agr. Expt. Sta. Bull., 579 (1987).
64. K. H. Knorzer, Bererichte der Deutschen Botanischen Gesellschaft, 91, 187 (1978).
65. R. Seehuber and M. Dambroth, Landbauforschung Voelkenrode, 37, 219 (1987).
66. G. Enge and G. Olsson, Sveriges Utsadesforenings Tidskrift, 96, 220 (1986).
67. B. G. Kartamyshev, Selektsiya i Semenovodstvo, 6, 9 (1985).
REFERENCES 301
68. R. Seehuber and M. Dambroth, Landbauforschung Volkenrode, 33, 183 (1983).

69. R. Seehuber and M. Dambroth, Landbauforschung Volkenrode, 34, 174 (1984).
70. R. K. Downey, J. Amer. Oil Chem. Soc., 48, 718 (1971).
71. R. Seehuber, Fette-Seifen-Anstrichmittel, 86, 177 (1984).
72. R. Marquard and H. Kuhlmann, Fette Seifen Anstrichmittel, 88, 245 (1986).
73. G. O. Korsrud, M. O. Keith, and J. M. Bell, Can. J. Anim. Sci., 58, 493 (1978).
74. J. Zubr, Indust. Crops Products, 6, 113 (1997).
75. R. Lange, W. Schumann, M. Petrzika, H. Busch, and R. Marquard, Fat Sci. Technol., 97,
146 (1995).
76. J. G. Crowley and A. Frohlich. (2002, November 29). Factors Affecting the Composition
and Use of Camelina. Technical Report, Crops Research Centre, Oak Park, Carlow,
Ireland. Available: http://www.teagasc.ie/research/reports/crops/4319/eopr-4319.htm.
77. J. T. Budin, W. M. Breene, and D. H. Putnam, J. Amer. Oil Chem. Soc., 72, 309 (1995).
7
Olive Oil
David Firestone
United States Food and Drug Administration
Washington, DC
1. INTRODUCTION AND HISTORY
Olive oil, an important component in the diet of Mediterranean people, is obtained

by mechanical extraction from the fruit of the Olea europaea L. tree, which belongs
to the Olive family, comprises some 400 species, and thrives in temperate and
tropical climates (1). Of the 35 species in the genus Olea, mainly of African, Indian,
and Australian origin, O. europaea is the only Mediterranean species. Although its
origin is not known, one theory is that it originated in ancient Iran and Turkestan,
spreading westward to Anatolia, Syria, and Israel along commercial and migratory
routes (2).
Olives appeared in Israel about 45,000 years ago (1). Charred pieces of olive
wood have been found in excavations at Lower Boker-Har Hanegev in layers dating
to 42,980 B.C. Both charred wood and carbonized stones have been found in many
archeological sites in Israel dating from 8000 B.C. onward, and indirect evidence
suggests the use of wild olives (O. oleaster) by humans as early as the seventh
millennium B.C. (3). It is not known whether the carbonized stones and charred
wood obtained from Chalcolithic (fourth millennium B.C.) and Early Bronze Age
(29002700 B.C.) sites represented cultivated or wild olives.
Olive farming and an olive oil industry appear to have been well established
throughout the region bordering the Mediterranean from Palestine and Syria to
Greece in the middle and late Bronze Age (4). Olive farming in Palestine and Syria
303
304 OLIVE OIL
increased dramatically at the turn of the first millennium B.C. (2). An olive oil indus-
try became well established in Palestine, and the export of olive oil from Palestine
to Egypt is documented in Old Kingdom Egypt. Olive cultivation provided materi-
als useful as a lamp fuel, lubricants, and body ointments; the fruit was easily cured
by salting, and the wood was used for carpentry and fuel. Later, the olive fruit
became a source of edible oil.
The manufacturer of olive oil became a mass production industry during the
Israelite period when processing methods improved (3). In Judea, oil presses gen-
erally consisted of large stone beds with a collecting vat in the center of the pressing
surface. A beam, which acted as a lever and was weighted down by several stones,
was used for pressing. The end of the beam was anchored to a wall behind the press
(niche wall) with a special niche stone. Olives were crushed in a rectangular basin
by a roller, which an operator set into forward and backward motion by means of an
attached shaft.
A typical Iron Age industrial site is that of the seventh century B.C. biblical town
of Timnah (Tel Batach), which was a center of olive oil production along with other
towns in the Tel Aviv area (5). The oil presses of the town were constructed simi-
larly to those of other Iron Age sites in the area. Each press consisted of a crushing
basin with two pressing vats on either side. Olives were crushed in the basin with
stone rollers, each of which had wooden handles fitted into sunken depressions at
the sides. The crushing basin was a shallow trough made of one large chalk stone.
Each pressing vat contained a large stone with a flat top and an inner hollowed
space for collecting the oil. Because there was no means of draining the oil from
the vet, pottery jugs were used to withdraw the oil. Baskets with crushed olives
were pressed by wooden beams anchored at one end to niches in the wall; the other
end of each beam was pressed down with three heavy stone weights (Figure 1).
Olive growing reached Cyprus and the Aegean area around the sixteenth century
B.C. As Renfrew (6) pointed out, the olive was one of three important constituents,
along with the vine and domesticated wheat, that contributed to the emergence of
civilization in the Aegean region. Oil production and trade played important roles in
the MinoanMycenaen economy of Crete and main-land Greece in the second mil-
lennium B.C. (7). Olive oil was used in the manufacture of scented perfumes and
unguents in the palace industries of Crete and Mycenae. Wild rather than cultivated
olives were apparently preferred for Aegean perfume and unguents because of the
low fat content of the wild olive.
Initially, olives were harvested by beating the trees with flails (6). After harvest-
ing, the olives were drenched in hot water and pressed to extract the oil. The oil was
separated from the water in a vat from which the water was drawn off, and then
stored in jars similar to those used to hold wine. Oil was used locally for lighting,
hygienic purposes (to clean the body), and as food, especially for cooking.
Mycenaen documents suggest that scented olive oil was used for religious purposes
and as a body ointment for the rich (8).
The earliest evidence of olive oil extraction in Cyprus dates to about 1300 B.C.
(9). (Wild olives grew on the island at least as early as 4000 B.C.) An olive press
(probably a lever and weight press) found at a Maroni excavation site consisted
INTRODUCTION AND HISTORY 305
Figure 1. Oil press at Tel Batach (biblical Timnah), 7 km west of Beth-Shemesh, Israel.
Isometric plan and sections (5). [Reprinted with kind permission of the editors of Olive oil in
Antiquity (5).]
306 OLIVE OIL
of a large rectangular trough (pressing bed) set on a mudbrick platform and sloping
downward. The trough is a flat stone with channels cut to meet at a small projection,
permitting the liquid to pour off into a jar standing on the floor below the press.
Other presses of the late Roman period found on the island were of the lever and
screw type in which the horizontal beam is immobile while the screw presses
down on the pressing bed. The screw was already used for pressing in Italy in
the first century B.C., initially as a lever and screw press and then in a direct frame
press (10).
Olive orchards continued to be extensively cultivated in Palestine throughout the
Byzantine and Arab periods (11). The chain of mountains from the upper Galilee to
Hebron were covered with olive trees. Olive oil was used regularly for food and
cooking as well as for lighting and manufacture of soap by boiling the oil with
ashes. During the early period under the Umayyides (661750 A.D.) and Abbasides
(7501258 A.D.), oil surpluses were exported from Palestine by land to neighboring
countries. With revival of maritime commerce under the Fatimids (9091171 A.D.),
oil was transported to Egypt and other countries by boat.
Phoenician settlements in the Mediterranean basin introduced olive farming into
Sicily, Sardinia, southern France, and Spain (2). The Greeks later spread farming
independently of the Phoenicians, reintroducing the olive into Sicily. The Romans
spread olive farming throughout their territories and used the olive tree in their land
reclamation projects, particularly in North Africa where they instituted olive farm-
ing and other projects to reclaim desert areas. Although of variable quality, olive oil
was a staple food and an important industrial product in Roman times.
Olive growing continued to prosper in the Mediterranean region until the fifth
century A.D., when the Roman Empire was invaded from the north and maritime
routes were closed (2). Olive farming was later revived with commercial develop-
ment of Venice and other maritime republics during the Renaissance. In 1709, olive
growing entered a new modern age when all of Europe was hit with a deep cold
spell and new orchards were planted to replace those destroyed by the cold weather.
As modern farming techniques evolved, large-scale state enterprises were begun
and olive farming reached a peak in the first half of the nineteenth century.
2. STATISTICS AND DEFINITIONS
Currently, more than 95% of the worlds olive trees grow in the Mediterranean
Basin. About 81% of total olive production comes from the European Community
(EC) (Spain, Italy, Greece, Portugal, and France), with the Near East contributing,
ca 7% and North Africa supplying about 11%. The remaining 1% is of American
origin, chiefly from Argentina, Mexico, Peru, and the United States (Table 1). Olive
oil consumption is growing in the developed countries that produce little or no olive
oil (Table 2).
The fruit of the olive tree is an egg-shaped drupe, consisting of a pericarp and an
endocarp. The pericarp includes an epicarp (skin) of variable thickness according to
the variety, and a mesocarp (pulp) surrounding the endocarp (woody pit) in which
STATISTICS AND DEFINITIONS 307
TABLE 1. World Production of Olive Oil (Thousand Metric Tons).a
2002/03 2003/04
Country 1997/98 1998/99 1999/00 2000/01 2001/02 (prov.) (est.)
Algeria 15.0 54.5 33.5 26.5 25.5 16.5 40.0

Argentina 8.0 6.5 11.0 4.0 10.0 11.0 22.0
Cyprus 1.5 2.5 3.5 5.5 6.5 7.0 7.0
EC 2,116.5 1,707.0 1,878.5 1,940.5 2,463.5 1,942.5 2,307.0
Croatia 5.0 9.0 5.5 5.0 7.0 3.0
Israel 3.0 4.5 2.5 7.0 3.5 9.0 2.5
Jordan 14.0 21.5 6.5 27.0 14.0 28.0 11.5
Lebanon 3.5 7.0 5.0 6.0 5.0 6.0 4.0
Morocco 70.0 65.0 40.0 35.0 60.0 45.0 80.0
Palestine 9.0 5.5 2.0 20.0 18.0 21.5 5.0
Syria 70.0 115.0 81.0 165.0 92.0 165.0 110.0
Tunisia 93.0 215.0 210.0 130.0 35.0 70.0 180.0
Turkey 40.0 170.0 70.0 175.0 65.0 160.0 60.0
Australia 0.5 0.5 1.0 1.0 2.0 3.0
Egypt 1.0 0.5 2.5 0.5 1.5 5.0 2.0
USA 1.0 1.0 1.0 0.5 0.5 1.0 1.0
Iran 3.0 2.5 2.5 3.0 2.5 1.5 4.0
Libya 6.0 8.0 7.0 4.0 7.0 6.5 6.5
Mexico 2.0 2.5 1.0 1.5 2.0 2.5 2.5
Yugoslavia 0.5 1.0 1.0 0.5 0.5 0.5 0.5
Serbia and
Montenegro 8.5 7.5 6.5 7.5 7.5 7.5 7.5
World Total 2,465.5 2,402.5 2,374.5 2,565.5 2,825.5 2,515.0 2,859.0
a
Source: International Olive Oil Council (IOOC).
the seed is enclosed. The yield per hectare is about 2.45 tons. Oil yield per 100 kg
of fruit is 19.6 kg (based on yields in Italy during the past 10 years).
In addition to oil, the pulp and epicarp contain a variety of natural components
soluble in the oil. As will be seen later, the oil is obtained from the olive by a
variety of techniques, always physical, leaving a residue (pomace) that contains
up to 8% oil, which is then extracted by solvent (usually hexane) and named
pomace oil.
Because of the behavior of the solvent, solvent-extracted oil contains more
minor components at higher levels than those found in physically extracted oil.
This provides the basis for designating pomace oil as a commercial product distinct
from virgin oil (obtained only by mechanical means) or refined (lower grade) virgin
oil mixed with virgin oil (olive oil, Riviera type).
The following internationally recognized definitions of oils derived from olives
and available on the market were promulgated by the International Olive Oil
Council (IOOC) (12):
1. Olive oil is that oil produced by extraction of the fruit of the olive tree (Olea
Europaea Sativa Hoffman et Link) to the exclusion of oils obtained using
solvents or reesterification processes and of any mixture with oils of other
308 OLIVE OIL
TABLE 2. Olive Oil Consumption (Thousand Metric Tons).a
2002/03 2003/04
Country 1997/98 1998/99 1999/00 2000/01 2001/02 (prov.) (est.)
Algeria 31.5 44.0 42.0 26.0 25.0 16.0 39.0

Argentina 8.0 8.0 7.0 6.0 5.5 5.5 6.0
Cyprus 2.0 2.5 4.0 5.0 5.5 6.0 6.0
EC 1,705.5 1,709.0 1,728.0 1,835.0 1,894.0 1,904.5 1,932.0
Croatia 4.0 8.5 6.5 5.0 6.0 3.0
Israel 6.5 9.5 12.5 13.5 14.5 14.5 13.5
Jordan 19.0 19.0 9.0 17.0 20.0 25.0 15.5
Lebanon 8.0 9.0 8.0 8.0 7.0 7.0 7.0
Morocco 55.0 55.0 55.0 45.0 60.0 55.0 70.0
Palestine 5.5 4.0 4.0 8.0 10.0 12.0 12.0
Syria 95.0 88.0 90.0 110.0 86.0 100.5 115.0
Tunisia 52.0 49.0 60.0 58.0 28.0 30.0 60.0
Turkey 85.5 85.0 60.0 72.5 55.0 55.0 40.0
2,073.5 2,086.0 2,088.0 2,210.5 2,215.5 2,237.0 2,319.0
Australia 17.5 24.0 25.5 31.0 27.5 31.0 31.0
Brazil 29.0 23.5 25.0 25.0 22.5 20.0 21.0
Chile
Egypt 1.0 1.0 1.5 1.0 1.5 3.5 2.5
USA 142.5 151.0 169.5 194.5 188.5 190.0 195.0
Iran 4.0 2.5 2.5 3.0 2.0 2.0 3.5
Libya 7.0 16.0 11.0 7.0 8.0 8.5 8.5
Mexico 4.5 5.0 5.0 6.5 8.0 10.0 10.0
Yugoslavia/Serbia and
Montenegro 0.5 1.0 1.0 0.5 0.5 0.5 0.5
Other producing
countries 13.5 12.5 13.0 13.0 14.0 14.5 14.5
219.5 236.5 254.0 281.5 272.5 280.0 286.5
Saudi Arabia 5.0 5.5 4.5 4.0 5.0 7.0 7.5
Canada 17.5 18.5 23.0 24.5 24.0 24.0 24.5
Japan 34.0 28.5 27.0 30.0 31.5 32.5 33.0
USSR/Russia 1.5 2.0 3.0 4.0 4.0 6.0 7.0
Switzerland 5.5 6.0 8.0 8.0 9.0 10.0 10.0
Taiwan 4.5 7.0 6.0 8.0 6.5 5.5 6.0
Other nonproducing
countries 20.5 23.0 29.0 20.0 38.0 38.5 38.5
88.5 90.5 100.5 98.5 118.0 123.5 126.5
Total World 2,381.5 2,413.0 2,442.5 2,590.5 2,606.0 2,640.5 2,732.0
a
Source International Olive Oil Council (IOOC).
kinds. In no case shall the designation olive oil be used to refer to olive
pomace oils.
A. Virgin olive oil is the oil obtained from the fruit of the olive tree solely
by mechanical or other physical means under conditions, particularly
thermal conditions, that do not lead to alterations in the oil, and which
has not undergone any treatment other than washing, decantation,
centrifugation, and filtration.
STATISTICS AND DEFINITIONS 309
Virgin olive oil fit for consumption as is (and can be designated as

natural) includes:
a. Extra virgin olive oil: virgin olive oil that has an organoleptic
rating of 6.5 or more as determined by the IOOC method (13) and
a free acidity, expressed as oleic acid, of not more than 1 g per
100 g.
b. Fine virgin olive oil: virgin olive oil that has an organoleptic
rating of 5.5 or more and a free acidity, expressed as oleic acid, of
not more than 1.5 g per 100 g.
c. Semifine virgin olive oil (or ordinary virgin olive oil): virgin olive
oil that has an organoleptic rating of 3.5 or more and a free
acidity, expressed as oleic acid, of not more than 3.3 g per 100 g.
(This class of olive oil is normally traded in bulk for blending
purposes.)
B. Virgin olive oil not fit for human consumption, also designated as
lampante virgin olive oil: virgin olive oil that has an organoleptic rating
of less than 3.5 and/or a free acidity, expressed as oleic acid, of more
than 3.3 g per 100 g. This class of olive oil is used to produce refined
olive oil or is intended for technical (nonfood purposes).
C. Refined olive oil: olive oil obtained from virgin olive oils by refining
methods that do not lead to alterations in the original triglyceride
structure.
D. Olive oil: the oil consisting of a blend of refined olive oil and virgin
olive oil in various proportions.
2. Olivepomace oil: the oil obtained by solvent extraction of olivepomace and

not including any oil obtained by a reesterification procedure or any mixture
with other kinds of oils. (The various categories of olivepomace oil are
described below.)
A. Crude olivepomace oil: olivepomace oil intended for refining to

produce a product (as B, below) suitable for human consumption, or
intended for technical purposes.
B. Refined olivepomace oil: the oil obtained from crude olivepomace
oil by refining methods that do not lead to alterations in the original
triglyceride structure.
C. Olivepomace oil: a blend of refined olivepomace oil and virgin olive
oil (any A, B, or C). In no case may this be called olive oil.
Because the yearly production of olive oil is variable, low-production years can fol-
low years of high production. Therefore, it is customary to record average values
(Table 1).
310 OLIVE OIL
3. EXTRACTION TECHNOLOGY
Ripe olives contain a variety of components, including water, oil, sugars, proteins,
organic acids, and cellulose. Olive cultivars with medium-size fruits generally pro-
vide the best oil yields. The pulp-to-kernel ratio of olives for oil production ranges
from 4:1 to 8:1.
The epicarp contains a number of components of relatively high polarity that are
not removed by mechanical extraction and remain in the pomace. Removal of these
components along with the oil by solvent extraction of the pomace accounts for
the higher unsaponifiable content of olivepomace oil.
Most of the oil (9698%) is in the pulp along with most of the water vegetation
water (VW), which accounts for 4060% of the weight of the fruit.
The woody pit inside the mesocarp holds a seed whose oil is more unsaturted
than the mesocarp (pulp) oil because of a higher content of linoleic acid. The ratio
of fruit oil to seed oil is 50:1.
The approximate chemical composition of olive fruit is as follows: water 52.4%;
oil 19.6%; proteins 1.6%; sugars 19.1%; cellulose 6.8%; and ash 1.5%. Oil yield
and quality depend on the cultivar of olive tree, ratio of the various anatomical
parts, and levels of minor components as well as growing conditions and health
of the trees. Soil moisture is very important during fruit development.
Harvesting of fruit for oil production begins in the middle of autumn and lasts
until the end of February. In some regions, it begins earlier, and in other locales, it
lasts until March. Accordingly, differences in oil quality and composition can be
expected along with variations caused by climatic and soil conditions. Variations
in quality are chiefly related to the levels of minor components and flavor com-
pounds, acidity, and the presence of mono- and diglycerides (1416).
Analytical and organoleptic data show that oil content is lower at the
beginning than at the end of the harvesting period, but it is of higher quality
(15). Harvesting technology is very important for production of high-quality
oil. Olives should be collected as soon as they reach optimal maturity; however,
it is difficult to have mechanical collection devices available where and when
needed. In addition, because of the conformation of the tree branches, strong adher-
ence of the fruit to the tree, and limited accessibility, most olives are picked by
hand.
Another harvesting procedure is to wait until the olives drop naturally and then
collect the fruit with a system of nets. When the ripening period is delayed, both
this procedure and handpicking are used. Although attempts have been made in the
past to use chemicals to influence dropping time, chemicals are seldom used.
Mechanical devices must be used with caution so that neither the tree nor the
branches are damaged. When mechanical devices are used, the olives are caught
in nets to avoid contact with the ground and damage to the fruit.
Under optimum conditions, the olives are transferred from the nets to cages
(usually plastic), forming layers not higher than 30 cm each, and the olives are
sent promptly to the extraction plant. In most regions of Italy and Greece, cages
are stored no more than 3 to 5 days before extraction. This procedure ensures
EXTRACTION TECHNOLOGY 311
high-quality oil if climatic conditions were good, the trees received proper care, and
the fruit was not damaged by pests.
If proper precautions are not taken and the olives are collected in large
batches and held in piles several meters high, the fruit may be damaged. The
enzymes released will cause hydrolytic and oxidative transformations resulting in
off-flavors that affect the quality of the oil. Even with low acidity, such oils will
have an unpleasant taste not acceptable for virgin oil and will have to be refined.
Because of the difference in price between virgin and refined oils, economic losses
to the farmer can be high.
Three systems are used for mechanical extraction of oil from the olive fruit:
pressure processing (Figure 2); centrifugation (Figure 3); and adhesion filtering
(Figure 4) (17). Pressing is the oldest and most often used method for olive oil
extraction. High-speed rotating machines are used for centrifugation extraction.
With adhesion filtration, a series of steel plates or blades are dipped into olive paste;
when withdrawn, the oil drips off the blades.
Several processing steps are required before extraction. The fruit must first be
cleaned to eliminate branches and leaves and any extraneous materials that might
damage plant equipment. The fruit is then washed to remove dirt and agricultural
contaminants, and finally crushed and milled to a coarse paste (Figure 5). During
the last step, enzymatic action breaks up the bitter components and reduces the level
of peppery constituents while increasing the amount of minor polar components
Figure 2. Pressure extraction of oil. 1, movement of the rack; 2, movement of the oil; A, mobile
head; B, fixed head.
312 OLIVE OIL
Figure 3. Centrifuge for oil extraction from olivepomace.
and tocopherols in the oil. If enzymatic action is prolonged, the minor polar com-
ponents break down into water-soluble compounds that are removed from the oil,
causing the loss of much of the antioxidant strength of the oil. Milling releases the
oil from the oil-bearing cells and helps smaller droplets of oil to merge into larger
drops, thus preparing the fruit for the following extraction step. A solid residue and
vegetation water are produced during extraction in addition to oil (Figure 6). The
vegetation water must be purified before discharge into a municipal sewer. Waste
water has been used to grow yeast, to produce butanol using microorganisms, to
isolate anthocyanin compounds for use in the food industry, and to produce steam.
Figure 4. Diagram of adhesion extraction of oil.

EXTRACTION TECHNOLOGY 313
Figure 5. Flow diagram of steps to prepare olives for extraction of oil.
Figure 6. Flow diagram of olive oil extraction and processing to yield olive oil products and
byproducts.
314 OLIVE OIL
Figure 7. Flow diagram of solvent extraction of pomace.
Efforts are being made to reduce waste water by recycling in the milling process
and to decrease its environmental pollution by treatment with biological or physical
processes prior to its discharge (22). A number of alternative technologies are avail-
able for waste water purification (1821); however, they are costly and difficult to
apply.
If suitable for consumption, the oil is centrifuged after extraction to eliminate
solid impurities and residual water. If the free fatty acid content is too high or orga-
noleptic properties are unsatisfactory, the oil is refined.
At the solvent extraction plant, the cake (pomace) containing up to 8% residual
oil is dried in a rotary kiln before proceeding to the solvent extraction unit, usually a
semicontinuous system (Figure 7). The extracted pomace oil is always refined.
Spent cake is used as fuel or is separated into two fractions, the pulp (including
skin) and the pit. In addition to use as fuel, the pit is occasionally used to produce
fiberboard (23).
4. REFINING OF OLIVE OILS
Olive oil refining is carried out in either of two ways: by alkali refining, generally
used for animal and vegetable oils and fats; or by physical refining, a technology
not usually used for seed oils. Flow diagrams of the two procedures are shown in
Figures 8 and 9.
REFINING OF OLIVE OILS 315
Figure 8. Flow diagram of alkali refining.
Figure 9. Flow diagram of physical refining.

316 OLIVE OIL
In the first procedure, the oil is treated with dilute acid to precipitate the gums
(phosphatides and proteinaceous material), which are separated by settling or
centrifugation. Phosphoric acid and citric acid are the two most common degum-
ming agents. After degumming, the oil is neutralized (alkali refined) either in a
batched or continuous system. Batch neutralization is currently preferred because
centrifuging of only the settled soap fraction lowers the neutralization coefficient
values, thereby shortening the washing time of the oil. The separated soap solution
is acidified with sulfuric acid to recover the free fatty acids (containing 3040%
triglycerides) for industrial applications.
The alkali-refined oil is then bleached under vacuum with mixtures of various
adsorbents (bleaching earth or clay and sometimes small amounts of activated
carbon) and filtered by any of a number of available filter presses occasionally
equipped with a solvent system for recovering oil entrained in the bleaching
earth.
The bleached oil is deodorized in a semicontinuous or continuous deodorizer
operating at a vacuum of less than 2 mm Hg. The final step involves mixing refined
oil with virgin oil to improve the organoleptic and keeping properties of the oil. A
good olive oil will contain at least 20% virgin oil, but the product must, of course,
meet consumer preference, which sometimes requires a very light flavor and
taste.
With physical refining, the oil is first degummed and bleached and then fed to a
continuous distillation (deodorization) unit, which removes the free fatty acids (92
95%) and volatiles. The refined oil is blended as above. Frequently, distillation is
stopped before removal of all of the free fatty acids, and the oil is alkali refined to
remove the remainder of the free fatty acids. This procedure has the advantage of
eliminating oxidation byproducts and pro-oxidant metals, thus improving product
stability.
5. REFINING OF POMACE OIL
The technologies adopted for refining pomace oil are based primarily on physical
refining because the acidity of pomace oils is about 10% (expressed as oleic acid).
Because degumming of pomace oil requires more drastic conditions than those for
pulp oil, larger amounts of acidulant are used (phosphoric acid is preferred), and
occasionally, the precipitate (gum) that entrains a high proportion of oil is centri-
fuged to recover the oil. Larger amounts of bleaching earths are required to remove
the intense green color of the oil. Additional processing of the bleached oil usually
follows the same procedures described for physical refining of olive oil, including
incomplete distillation (deodorization) followed by alkali refining of the partially
deodorized oil.
Dewaxing (winterization) of pomace oil is mandatory because of its high content
of waxes (olive oil may also be winterized, especially if it is used to produce
margarine or mayonnaise). Higher melting point triglycerides are also removed.
Winterization can be carried out after bleaching or following partial deodorization
OLIVE OIL COMPONENTS 317
and alkali neutralization (alkali refining). If alkali neutralization is performed at a

low temperature, winterization can be carried out simultaneously. A continuous
apparatus is generally used for winterization (Figure 9) coupled with continuous
filtering units. The winterization oil is then blended with virgin oil to restore the
oils antioxidant properties.
6. OLIVE OIL COMPONENTS
Glycerides account for at least 97% of a virgin oil if the acidity is disregarded. The
free fatty acid content is used to distinguish the various classes of virgin oil, from
extra virgin to lampant. It must be emphasized that virgin olive oil is a natural
product and therefore subject to variations in composition, both qualitative and
quantitative. The origin, cultivar, extraction technology, state of ripening of the
fruit, climatic conditions, and rainfall all influence biosynthesis within the fruit
and, therefore, the composition and quality of the oil. The fatty acid composition
of olive oil is shown in Table 3, which lists typical compositions of European,
Turkish, and African (Tunisian) oil as well as IOOC limits (12). Differences in
composition are due chiefly to linoleic, linolenic, and palmitic acid content. Olive
oils from Argentina resemble those from Tunisia. The triglyceride composition of
European, Turkish, and Tunisian olive oils is shown in Table 4 (main glycerides
are shown). Fatty acid distribution in the triglycerides follows the 1,3-random,
2-random rule (2426).
Several classes of minor components are present in virgin olive oil. The struc-
ture, concentration, and number of these substances are characteristic of virgin oils.
Some are minor glyceridic components (MGCs); others fall into other categories as
listed below.
TABLE 3. Fatty Acid Composition of Olive Oil (%).
Acid CANa European Turkish Tunisianb Limits (12)
Palmitic 16 : 0 8.4 12.1 15.3 7.520.0

Palmitoleic 16 : 1 0.7 0.7 1.6 0.33.5
Heptadecanoic 17 : 0 0.1 0.2 0.1 0.00.3
Heptadecenoic 17 : 1 0.1 0.2 0.1 0.00.3
Stearic 18 : 0 2.5 3.1 2.1 0.55.0
Oleic 18 : 1 78.0 71.3 62.5 55.083.0
Linoleic 18 : 2 8.3 10.6 16.5 3.521.0
Linolenic 18 : 3 0.8 0.7 0.8 0.30.9
Arachidic 20 : 0 0.5 0.4 0.5 0.20.6
Eicosenoic 20 : 1 0.3 0.3 0.3 0.10.4
Behenic 22 : 0 0.1 0.2 0.1 0.00.2
Lignoceric 24 : 0 0.2 0.2 0.1 0.00.2
a
CAN Carbon atom number.
b
Typical values for Tunisian olive oil analyzed during 1994.
318 OLIVE OIL
TABLE 4. Main Triglycerides of Olive Oil (%).
ECNa Triglycerideb European Turkish Tunisianc
42 LLL, TLOd, TLPd 0.5 0.8 1.6

44 LLOd 2.4 3.2 10.6
TOOd , LLP 2.6 2.9 1.7
46 LOOd 13.3 13.8 16.0
LOPd, PLP 8.0 9.7 16.2
48 OOO 39.9 34.0 23.2
POO 26.0 24.4 22.0
POP 5.1
50 SOO 5.1 5.1 4.3
SOP 1.0 1.4 1.2
52 OSS, PSS 0.8 0.5
a
ECN equivalent carbon number.
b
L C18 : 2; T 18 : 3; O C18 : 1; P C16 : 0; S C18 : 0.
c
Typical values for Tunisian olive oil analyzed in 1994.
d
Mixture of isomers.
Hydrocarbons
Tocopherols
Linear short chain alcohols and their esters
Linear long chain alcohols and their esters
Sterols and their esters
a-Methyl sterols
Monohydroxytriterpenes
Dihydroxytriterpenes
Triterpenic acids
Phytol
Geranylgeraniol
Phenols and related compounds
Flavor components
Methyl and ethyl esters
Other components
6.1. Minor Glyceridic Components

Monoglycerides (MGs) and diglycerides (DGs) in the olive fruit are caused by
enzymatic hydrolysis of the triglycerides and incomplete triglyceride biosynthesis
(16). In general, DGs are more abundant than MGs. Determination of DG concen-
tration is useful for evaluating oil freshness and time of fruit harvesting because the
DG level is strongly related to climatic influences. DG concentration can even be
used to determine the source of an oil, even a refined oil, because the DG content of
edible virgin olive oils differs from that of high acidity oils or solvent-extracted oils.
Phospholipids are essentially absent from olive oil.
OLIVE OIL COMPONENTS 319
6.2. Nonglyceridic Minor Components

Hydrocarbons. Both even- and odd-chain n-paraffins, including branched-chain
(iso and anteiso) compounds, which are minor components of the hydrocarbon
fraction, are present in virgin olive oil. The polyunsaturated triterpenic hydrocarbon
squalene, and biochemical precursor of sterols, is the main component of the hydro
carbon fraction. The squalene content of olive oil ranges from 150 to 700 mg per
100 g (2730). b-Carotene is also present in olive oil as are aromatic hydrocarbons,
including benzenoid, napthalenic, and more complex aromatic hydrocarbons
(3037).
Linear Short Chain Alcohols and Their Esters. Methanol and ethanol esters of
the fatty acids present in olive and in the same proportions as in the olive are
present among the volatile compounds in virgin olive oil (3137).
Straight Long Chain Alcohols. Linear long-chain alcohols with carbon numbers
between C22 and C32 are present in olive oil both free and esterified (waxes). The
components are abundant in the epicarp of the fruit and concentrate in solvent
extracted oil. Phytol, probably derived from biodegradation of chlorophyll, is
also present along with geranyl (38).
Cyclic Monohydroxy Compounds. Triterpenic tetra- and pentacyclic mono-
hydroxy compounds are characteristic of olive oils (3449). The following com-
pounds have been shown to be present, accompanied by small amounts of
lanosterol and obtusifoliol:
Tetracyclic: cycloartenol
24-methylene cycloartanol
Pentacyclic: a-amyrin
b-amyrin
Methylsterols (4-desmethyl triterpenes) and sterols (4,4-di-desmethyl triterpenes)

present in olive oils are derived from the tetracyclic alcohols. The following methyl
sterols (4a-methyl-7-cholesten-3b-ol compounds) are present: 24-methylene,
24-methyl-, 24-ethylidene, and 24-ethyl.
The main sterols of olive oil are (40, 43, 4566) campesterol, stigmasterol,
clerosterol, b-sitosterol, sitostanol, and d-5-avenasterol.
These are accompanied by small amounts of cholesterol (max. 0.5%), brassica-
sterol (max. 0.1%), 24-methylenecholesterol (max. 0.5%), campestanol (max.
0.5%), d-5,24,-stigmastadienol (max. 1%), d-7-stigmastenol (max. 0.5%), and
d-7-avenasterol (max. 1.1%).
Analysis of the sterol fraction isolated from the unsaponifiable fraction is very
important, as will be seen later, for determining the authenticity of the oil. The
triterpenes and sterols are present both as free alcohols and as fatty acid esters
(46, 47).
Cyclic Dihydroxy Compounds. Pentacyclic triterpenes in olive oil include
3b,17b-dihydroxy-12-oleanene (erythrodiol) and its parent compound uvaol,
obtained largely from the epicarp and therefore characteristic of solvent extracted
oils (42, 65).
320 OLIVE OIL
Triterpenic Acids. The following pentacyclic mono- and dihydroxy triterpenic

acids are present in virgin olive oil (35, 43, 44): 3b-hydroxy-17-carboxy-d-12-olea-
nene (oleanolic acid); 3b,2a-dihydroxy-17-carboxy-d-12-oleanene (maslinic acid);
3b-hydroxy-17-carboxy-d-12-ursene (ursolic acid); 2a,3b-dihydroxy-17-carboxy-
d-12-ursene (2a-hydroxyursolic acid); and deoxyursolic acid (structure not fully
elucidated).
Chlorophylls. Both chlorophyll a and chlorophyll b are present in olives and are
partially extracted into the oils.
Flavor Components. Olive oil volatiles contain at least 100 compounds (3337)
in several categories: hydrocarbons (5 compounds), aliphatic alcohols (13 com-
pounds), terpenic alcohols (4 compounds), aldehydes (27 compounds), ketones
(8 compounds), ethers (2 compounds), furans (3 compounds), thiophenes (6 com-
pounds), and esters (29 compounds).
6.3. Minor Polar Components

The olive mesocarp contains a number of phenolic and polyphenolic compounds
and their esters, small amounts of which are present in olive oil (35, 43, 44). These
include monohydroxy- and dihydroxy-phenylethanol, including tyrosol and other
phenols and a series of carboxyphenols, including caffeic, o-coumaric, p-coumaric,
cinnamic, ferulic, gallic, p-hydroxybenzoic, protocatechuic, sinapic, syringic,
and vanillic acids. Benzoic and cinnamic acids are produced by hydrolysis of
flavonoids. The hydroxyphenylethanols arise from hydrolysis of oleoeuropein.
Their esters are responsible for the bitterness and pepperlike sensation occasionally
dominant in the taste of olive oils.
Olive oil contains a-tocopherol in the range of 12190 mg/kg. According to one
report (43), olive oil tocopherols were found to consist of 88.5% a-tocopherol,
9.9% b- g-tocopherol, and 1.6% d-tocopherol. Tocopherol content can be used
to detect adulteration of olive oil with seed oils.
7. ANALYSIS OF OLIVE OILS
Olive oil is initially examined to determine purity, then to place it in the proper
category, and finally to establish its quality.
7.1. Determination of Purity

Sterol Composition. Sterol analysis involves preparation of the unsaponifiable frac-
tion, fractionation by thin-layer chromatography (TLC), and gas chromatographic
analysis of the TMS derivatives (66). The following limits apply to all types of olive
oil (12):
ANALYSIS OF OLIVE OILS 321
Sterol Sterol Fraction (%)
Cholesterol Max. 0.5

Brassicasterol Max. 0.1
Campesterol Max. 4.0
Stigmasterol Less than 4.0
d-7-Stigmastenol Max. 0.5
The sum of the following sterols must be more than 93.0% of
the sterol function:
b-Sitosterol
d-5-Avenasterol
d-5,23-Stigmastadienol
Clerosterol
Sitostanol
d-5,24-Stigmastadienol
Total Sterol Content. The gas liquid chromatographic method for sterol determi-
nation using an internal standard (cholestanol) is used to calculate the absolute
(total) sterol content of an oil (68, 69). Gravimetric, enzymatic, colorimetric, and
liquid chromatographic methods have also been reported (69). Limits (mg/100 g)
are as follows (12): virgin olive oil, refined olive oil, and olive oil (mixture of
refined and virgin) >100; crude olivepomace oil >250; and refined olivepomace
oil, olive oil and olivepomace oil (mixture) >180.
Fatty Acid Composition. Olive oil triglycerides are converted into methyl esters,
and the methyl esters are analyzed by gasliquid chromatography (GLC) (70, 71).
The limits of genuine olive oil are as follows (% m/m) (12):
Acid CANa Minimum Maximum
Myristic 14:0 0.05

Palmitic 16:0 7.50 20.00
Palmitoleic 16:1 0.30 3.50
Heptadecanoic 17:0 0.30
Heptadecenoic 17:1 0.30
Stearic 18:0 0.50 5.00
Oleic 18:1 55.00 83.00
Linoleic 18:2 3.50 21.00
Linolenic 18:3 0.90
Arachidic 20:0 0.60
Eicosenoic 20:1 0.40
Behenic 22:0 0.20
Lignoceric 24:0 0.20
a
CAN carbon atom number.
Saturated Fatty Acids in Position 2 of the Triglycerides. Hydrolysis with pan-

creatic lipase is followed by thin-layer chromatographic isolation of the monogly-
ceride fraction, which is converted to methyl esters. The methyl esters are analyzed
322 OLIVE OIL
by GLC (72, 73). Maximum acceptable level is the sum of palmitic and stearic acid
(% m/m) (12):
Virgin olive oil 1.5

Refined olive oil 1.8
Olive oil (mixture of refined and virgin) 1.8
Crude olivepomace oil 2.2
Refined olivepomace oil 2.2
Absolute Difference Between Found and Theoretical Equivalent Carbon Number

(ECN) 42 (Trilinolein) Values. The triglyceride composition of the oil is deter-
mined by high-performance liquid chromatography (HPLC) (74). (A chromatogram
of an olive oil sample (ECN 42, 0.8%) is shown in Figure 10.) The theoretical tri-
glyceride composition is calculated with a Lotus 123 program provided by the
IOOC. The maximum difference of theoretical ECN 42 vs. ECN 42 found is calcu-
lated. (ECN CN-2n, where CN is the carbon number and n is the number of dou-
ble bonds.) The maximum difference between the real and theoretical ECN content
Figure 10. HPLC chromatogram of olive oil triglycerides. Column: LC-18, 200 4:6 mm i.d.;
mobile phase : acetone : acetonitrile (60 : 40, v/v); flow rate : 0.75 mL/min; refractive index
detector; oven and detector temperature : 40 C. IUPAC Method 2.324 (72) with injection of 10-mL
test sample diluted 1 : 20 with acetone. ECN 42, 0.8% of total glycerides.
of olive oils and olivepomace oils should be 0.3 and 0.5, respectively. This proce-
dure avoids errors because of miscalculation of trilinolein alone (75).
Trans-Fatty Acid Content. Trans-fatty acids arise during refining of vegetable
oils as well as during hydrogenation, or from attempts to eliminate the sterol frac-
tion of seed oils with a fatty acid composition similar to that of olive oil. Methyl
esters are analyzed by capillary column GLC (76, 77). The following limits
(% m/m) are mandatory (12):
18:1 18:2 trans

Oil trans 18:3 trans
Virgin olive oil <0.03 <0.03

Lampant olive oil 0.10 0.10
Refined olive oil 0.20 0.30
Olive oil (mixture of refined and virgin) 0.20 0.30
Crude olivepomace oil 0.20 0.10
Refined olivepomace oil 0.40 0.35
Olivepomace oil and olive oil mixture 0.40 0.35
7.2. Differentiation Between Olive Oil and Olive Pomace Oil

Wax Content. Olive oil fatty acid esters of straight chain alcohols (wax esters pre-
sent in solvent extracted olivepomace oil are isolated by column chromatography
on silica gel (LC) and quantitated by GLC to determine if olivepomace oil is
present in olive oil (78). LC separation of the wax esters can be replaced with
HPLC to automate the separation step and improve reliability and repeatability
(79). Limits for content of C40 C42 C44 C46 wax esters (mg/kg) are as
follows (12):
Virgin olive oil 250

Lampant olive oil 350
Refined olive oil 350
Olive oil (mixture of refined and virgin) 350
Dihydroxyterpene Alcohol Content. Olivepomace oil contains relatively high

levels of erythrodiol, uvaol, and wax esters. Erythrodiol and uvaol (total diol) con-
tent is determined by the same procedure as that used for sterol analysis (80, 81).
Limits for total diol content (as % of total sterols) are as follows:

Lampant olive oil 4.5
Olive oil (mixture of refined and virgin) 4.5
324 OLIVE OIL
7.3. Differentiation Between Virgin and Refined Olive Oil and Detection
of Refined Olive Oil and Seed Oils in Virgin Olive Oil
Concentration of Stigmasta-3,5-Diene. When olive oil and seed oils are refined,
stigmasta-3,5-diene is produced by dehydration of b-sitosterol, the parent
sterol (82). Refined olive oils contain significant amounts of stigmasta-3,5-diene
(3 100 mg/kg) not present in any significant amount in virgin olive oils. Refined
seed oils also contain significant amounts of steroidal hydrocarbons, including
campesta-3,5-diene and stigmasta-3,5,22-triene in addition to stigmasta-3,
5-diene. The relative amounts of these steroidal hydrocarbons can be used to
detect refined seed oils or seed oils desterolized for the purpose of adulterating
olive oil. Isolation of the hydrocarbon fraction from the unsaponifiables by col-
umn chromatography on silica gel followed by GLC is used to determine the con-
centration of stigmasta-3,5-diene and accompanying hydrocarbons (83, 84). A
chromatogram of the hydrocarbon fraction from an olive oil is shown in Figure 11.
Figure 11. Capillary GLC of the hydrocarbon fraction of olive oil (blend of refined and virgin olive
oil). Column: DB-5, 25 m 0:25 mm i.d., 0.2-mm film thickness; split ratio; 1 : 15; temperature
program: 235 C, 6 min; 20 C/min; 285 C final temperature; injector: 300 C; detector; 320 C,
1, cholesta-3,5-diene (internal standard); 2, stigmasta-3,5-diene.
A chromatogram of the hydrocarbon fraction from an olive oil admixed with des-
terolized, refined seed oil is shown in Figure 12. Ratios of stigmasta-3,5-diene to
campesta-3,5-diene (R1) and stigmasta-3,5-diene to stigmasta-3,5,22-triene (R2)
are determined when the level of stigmasta-3,5-diene exceeds 4 ppm (12).
Oil Stigmasta-3,5-diene (ppm) R1a R2b;c

Lampant olive oil 0.5
Refined olive oil 50.0 15 15
Olive oil 50.0 15 15
Crude olivepomace oil 0.5 15 15
Refined olivepomace oil 120.0 15 15
Pomace and olive oil mixture 120.0 15 15
a
R1 ratio of stigmasta-3,5-diene to campesta-3,5-diene.
b
R2 ratio of stigmasta-3,5-diene to campesta-3,5,22-triene.
c
Provisional limits.
However, a July 1994 IOOC report (84) noted that the R1 and R2 values of many
Italian and Greek olive oils were considerably lower than those proposed by the
IOOC (12) and that the composition of steroidal hydrocarbons should be identical
to that of the sterols from which they are derived when the R1 and R2 ratios are
used to identify extraneous oils in refined olive oil.
UV Absorption at 268 nm. K (1%, 1 cm) and related value, d-K, are useful
for readily classifying olive oil quality according to the following values
(12, 85):
Oil K 270 nm d-K a
Extra virgin olive oil 0.25 0.01

Virgin olive oil (fine) 0.25 0.01
Virgin olive oil (semifine) 0.30 0.01
Lampant olive oil No limits No limits
Refined olive oil 1.10 0.16
Olive oil 0.90 0.15
Crude olivepomace oil No limits No limits
Refined olivepomace oil 2.00 0.20
Pomace and olive oil mixture 1.70 0.18
a
d-K K 268 K 262 K 274 =2.
Both K and d-K are altered when oxidation products are present. In this case, the oil
is dissolved in hexane and passed through an alumina column before measurement
of K and d-K.
326 OLIVE OIL
Figure 12. Capillary GLC of the hydrocarbon fraction of olive oil admixed with a desterolized
seed oil (GLC column and operating conditions as described for Figure 11). 1, Cholesta-3,
5-diene (internal standard); 2, campesta-3,5-diene; 3, stigmasta-3,5,22-triene; 4, stigmasta-3,
5-diene.
7.4. Quality Parameters

Organoleptic Characteristics. Organoleptic properties of virgin oil can be deter-
mined by a panel test (13, 86), which gives results that are often controversial.
Organoleptic testing is currently undergoing revision. Currently the IOOC is pre-
paring a draft method for the organoleptic assessment of virgin olive oil using a
designation of origin (DO) code. It is intended for use by DO authorities to ensure
that the oil meets requirements (87). The panel test method is based on examination
of virgin oil by a panel of 8 to 12 trained personnel who grade various character-
istics and defects that are then converted into a number score. The following scores
apply to various grades of virgin olive oil:
Extra virgin olive oil >6.5
Fine virgin olive oil >5.5
Semifine virgin olive oil >3.5
Lampant virgin olive oil <3.5
Free Fatty Acid Content. Free fatty acid content (expressed as % oleic acid) (88)
is used to define the various grades of virgin olive oil (12):
Extra virgin olive oil <1.0

Fine virgin olive oil <1.5
Semifine virgin olive oil <3.3
Lampant virgin olive oil >3.3
Refined olive oil and mixtures have the following limits (12):

Olive oil 1.5
Refined olivepomace oil 0.3
Olivepomace and olive oil 1.5
Olive oil and mixtures of olivepomace and olive oil have higher free fatty acid
contents because they are generally mixed with virgin olive oils of high acidity.
Peroxide Value (PV). PV (expressed in meq per kg oil) (89) allowed for various
grades of olive oil is as follows (12):
Extra virgin, fine, and semifine virgin olive oil 20

Refined olive oil 10
Olive oil 15
Refined olivepomace oil 10
Pomace oil and olive oil mixture 15
Virgin olive oil contains components that interfere with conventional PV determi-
nation. Even freshly expressed olive oil has PV values of about 10, and under
certain climatic conditions (dry weather), the PV value can be higher than 10.
Tocopherol Content. Tocopherols can be determined by colorimetry or GLC
(90), or by HPLC (91, 92). Added tocopherols are not permitted in virgin olive
oils and crude olivepomace oils (12). Added a-tocopherol is allowed in refined
olive oil, olive oil, refined olivepomace oil, and olivepomace oil to restore natural
tocopherol lost during refining with a maximum level of 200 mg/kg of total a-toco-
pherol in the final product (12).
Impurities. Water content (93) of virgin olive oil should not exceed 0.2% (m/m);
for refined oil and mixtures (olive oil, olivepomace and olive oil), the maximum
value is 0.1%; for lampant olive oil, 0.3%; for crude olivepomace oil, 1.5% (12).
Allowable hydrocarbon (hexane, petroleum ether) residues (94) are as follows
(% m/m):
Extra virgin, fine, and semifine virgin olive oil 0.10

Refined olive oil, olive oil 0.05
Refined olivepomace oil, olivepomace and olive oil 0.05
328 OLIVE OIL
The occurrence of mg/kg to mg/kg amounts of tetrachloroethylene in some olive

oils (95) led to an EEC regulation limiting the tetrachloroethylene content of olive
oil and products containing olive oil to not more than 0.1 mg/kg, as determined by a
head space/electron capture GLC method (96).
Maximum allowable contents of iron and copper (97) are 3 ppm and 0.1 ppm,
respectively.
Smoke point (98) is a function of acidity level in the oil. The smoke point for
olive oil generally ranges from 150 C to 163 C.
7.5. Combined Gas ChromatographyMass Spectrometry (GC/MS)

GC/MS is a powerful tool for identification and confirmation of the various com-
ponents of olive oil. With GC/MS in the selective ion mode, unresolved GC peaks
can be identified and accurately quantitated. For example, an apparent b-sitosterol
peak in the sterol fraction was resolved into clerosterol (m/z 218) and -5-avenas-
terol (m/z 314), and both sterols measured regardless of inadequate GC resolution.
Italian and Spanish olive oil from the 19911992 crop year contained a very high
level of 9,19-cyclolanosterol (>400 mg/kg), which was not found with the standard
method for sterol analysis. Two isomers of this sterol were identified by GC/MS of
the unsaponifiable fraction, and their levels were found to be inversely proportional
to the levels of b-sitosterol in the oils. GC/MS of the unsaponifiable fraction with
high-resolution GC capillary columns provides a relatively rapid means of checking
product purity and the identity of individual components. Thus, triterpene diols
were identifiable at m/z 203, a-tocopherol at m/z 165, squalene at m/z 69, choles-
terol at m/z 386, and brassicasterol, characteristic of canola oil and other Brassica
oils, at m/z 398.
REFERENCES
1. N. Liphschitz in M. Heltzer and D. Eitam, eds., Olive Oil in Antiquity, University of Haifa,
Haifa, Israel, 1987, pp. 139145 (Sorgon Publications, Padova, Italy, 1994).
2. P. Fiorino and F. Nizzi Griffi, Olivae, (44), 912 Dec. 1992).
3. D. Eitam, in Ref. 1, pp. 1636.
4. D. Zohary and P. Spiegel-Roy, Science 187, 319 (1975).
5. G. I. Kelm and A. Mazar, in Ref. 1, pp. 121125.
6. C. Renfrew, The Emergence of Civilization, The Cyclades and the Aegean in the Third
Millenium BC, Methuen & Co., Ltd., London, 1972.
7. S. Bunimovitz, in Ref. 1, pp. 1115.
8. J. Chadwick in W. A. McDonald and G. R. Rapp, Jr., eds., The Minnesota Messenia
Expedition. Reconstructing a Bronze Age Regional Environment, The University of
Minnesota Press, Minneapolis, 1972, p. 115.
9. S. Hadjisavvas, in Ref. 1, pp. 98105.
10. R. Frankel, in Ref. 1, pp. 6380.
REFERENCES 329
11. J. Frenkel, in Ref. 1, pp. 5762.

12. International Olive Oil Council (IOOC), Doc. IOOC/T.15/NC no. 1/Rev. 6, June 10, 1993;
see also Codex Alimentarius Commission Doc. CL 1993/15-FO, May 1993 (proposed draft
revised standard for olive oils).
13. IOOC, Organoleptic Assessment of Virgin Olive Oil, IOOC/T.20/Doc. 3/Rev. 2, May 28,
1992; see also Official Journal of the European Communities, No. L 248/49, Annex, XII,
1991.
14. E. Fedeli in Proceedings of the Meeting of the Association of Italian Oils and Fats
Producers, June 67, 1991.
15. E. Fedeli and N. Cortesi, Riv. Ital. Sostanze Grasse 70, 419 (1993).
16. N. Cortesi, P. Rovellini, and E. Fedeli, Riv. Ital. Sostanze Grasse 69, 305 (1992).
17. L. Di Giovacchino, Olivae, (36), 1441 (Apr. 1991).
18. P. Bondioli, A. Lanzani, E. Fedeli, M. Sala, and G. Gerali, Riv. Ital. Sostanze Grasse 69,
487 (1992).
19. C. Carola, A. Arpino, and A. Lanzani, Riv. Ital. Sostanze Grasse 52, 335 (1975).
20. A. Lanzani, P. Bondioli, E. Fedeli, A. Ponzetti, and G. Pieralisi, Riv. Ital. Sostanze Grasse,
117 (1988).
21. P. Bondioli, A. Lanzani, and E. Fedeli, Riv. Ital. Sostanze Grasse 68, 1 (1991).
22. I. M. Ordovas, in F. J. Francis, Encyclopedia of Food Science and Technology, Vol. 3,
John Wiley & Sons, Inc., New York, 2000, p. 1775.
23. P. Bondioli, A. Lanzani, E. Fedeli, and L. Casarotto, Riv. Ital. Sostanze Grasse 66, 623,
661 (1989).
24. N. Cortesi, P. Rovellini, and E. Fedeli, Riv. Ital. Sostanze Grasse 69, 1 (1992).
25. C. Mariani and E. Fedeli, Riv. Ital. Sostanze Grasse 62, 3 (1985).
26. R. V. Flor, L. T. Hecking, and B. D. Martin, J. Amer. Oil Chem. Soc. 70, 199 (1993).
27. P. Capella, E. Fedeli, M. Cirimele, and G. Jacini, Riv. Ital. Sostanze Grasse 41, 635 (1964).
28. E. Fedeli, Rev. Franc. Corps Gras 15, 281 (1968).
29. C. Mariani, S. Venturini, and E. Fedeli, Riv. Ital. Sostanze Grasse 70, 321 (1993).
30. C. Mariani and E. Fedeli, Olivae (45), 3439 (Feb. 1993).
31. R. Vazquez, A. J. DelValle, and J. L. M. DelValle, Grasas y Ace 27, 185 (1976).
32. E. Fedeli, D. Baroni, and G. Jacini, Riv. Ital. Sostanze Grasse 50, 38 (1973).
33. E. Fedeli, G. Favini, D. Baroni and G. Jacini, La Chimica e lIndustria 55, 681 (1973).
34. R. A. Flath, R. R. Forrey, and D. G. Guadagni, J. Agric. Food Chem. 21, 948 (1973).
35. E. Tiscornia, M. Forina, and F. Evangelisti, Riv. Ital. Sostanze Grasse 59, 519 (1982).
36. G. Montedoro, M. Bertuccioli, and F. Anichini in G. Charalampous and G. Inglett, eds.,
Flavor of Foods and Beverages. Chemistry and Technology, Acacemic Press, Inc.,
New York, 1978.
37. H. Guth and W. Grosch, Fat Sci. Technol. 93, 335 (1991).
38. E. Fedeli, A. Lanzani, P. Capella, and G. Jacini, J. Lipid Res. 7, 437 (1966).
39. J. Eisner, J. L. Iverson, A. K. Mozingo, and D. Firestone, J. Assoc. Off. Agric. Chem. 48,
417 (1965).
40. P. Capella, E. Fedeli, M. Cirimele, A. Lanzani, and G. Jacini, Riv. Ital. Sostanze Grasse 40,
645 (1963).
330 OLIVE OIL
41. A. Fiecchi, P. Capella, E. Fedeli, A. Lanzani, and G. Jacini, La Ricerca Scientifica 36, 1316
(1966).
42. E. Fedeli and G. Jacini, Advan. Lipid Res. 9, 335 (1972).
43. E. Fedeli, Prog. Chem. Fats Other Lipids 15, 57 (1977).
44. E. Fedeli, Rev. Franc. des Corps Gras 30, 51 (1983).
45. P. Capella, E. Fedeli, M. Cirimele, and G. Jacini, Riv. Ital. Sostanze Grasse 40, 296 (1963).
46. P. Capella, E. Fedeli, M. Cirimele, A. Lanzani, and G. Jacini, Riv. Ital. Sostanze Grasse 40,
660 (1963).
47. C. Mariani, E. Fedeli, K. Grob, and A. Artho, Riv. Ital. Sostanze Grasse 68, 179 (1991).
48. C. Mariani, S. Venturini, P. Bondioli, E. Fedeli, and K. Grob, Riv. Ital. Sostanze Grasse 69,
393 (1992).
49. G. Bianchi and G. Vlahov, Fat Sci. Technol 96, 72 (1994).
50. P. Capella, E. Fedeli, M. Cirimele, A. Lanzani, and G. Jacini, Fette Seifen Anstrichm. 66,
997 (1964).
51. E. Fedeli, A. Lanzani, P. Capella, and G. Jacini, J. Amer. Oil Chem. Soc. 43, 254 (1966).
52. G. Jacini, E. Fedeli, and A. Lanzani, J. Assoc. Off. Anal. Chem. 50, 84 (1967).
53. E. Fedeli in Proceedings of the Second Meeting on Mass Spectrometry, Ispra, Italy 1971, 391.
54. E. Fedeli and C. Mariani, Riv. Ital. Sostanze Grasse 51, 129 (1974).
55. G. Jacini and E. Fedeli, Fette Seifen Anstrichm. 77, 1 (1975).
56. E. Fedeli, N. Cortesi, C. Mariani, A. Govoni, and D. Baroni, Riv. Ital. Sostanze Grasse 49,
105 (1972).
57. E. Fedeli, A. Daghetta, D. Baroni, and N. Cortesi, Riv. Ital. Sostanze Grasse 49, 159
(1972).
58. N. Cortesi, C. Mariani, and E. Fedeli, Riv. Ital. Sostanze Grasse 50, 411 (1973).
59. E. Fedeli, N. Cortesi, C. Mariani, D. Baroni, and G. Jacini, Sci. Tecnol. degli Alimenti 3,
143 (1974).
60. E. Fedeli, Annali di Chimica 64, 213 (1974).
61. E. Fedeli, N. Cortesi, and C. Mariani, Riv. Ital. Sostanze Grasse 53, 130 (1976).
62. N. Cortesi, E. Fedeli, A. Gasparoli, and E. Tiscornia, Riv. Ital. Sostanze Grasse 54, 16
(1977).
63. N. Cortesi, E. Fedeli, and E. Tiscornia, Riv. Ital. Sostanze Grasse 55, 168 (1978).
64. A. M. Leone, E. La Notte, and F. Lamparelli, Riv. Ital. Sostanze Grasse 53, 259 (1976).
65. T. Itoh, K. Yoshida, T. Tamura, and T. Matsumoto, and G. F. Spencer, J. Amer. Oil. Chem.
Soc. 58, 545 (1981).
66. C. Mariani, E. Fedeli, and K. Grob, Riv. Ital. Sostanze Grasse 68, 233 (1991).
67. E. Fedeli and C. Mariani, Riv. Ital. Sostanze Grasse 50, 164 (1973).
68. Official Journal of the European Communities, Determination of the Composition and
Content of Sterols by Capillary-Column Gas Chromatography, no. L 248/15, Annex V,
May 9, 1991.
69. G. Morchio, R. De Andreis, and E. Fedeli, Riv. Ital. Sostanze Grasse 64, 185 (1987).
70. International Standards Organization, Animal and Vegetable Fats and OilsAnalysis by
Gas Chromatography of Methyl Esters of Fatty Acids, International Standard ISO 5508,
Geneva, 1990.
REFERENCES 331
71. D. Firestone, ed., Official Methods and Recommended Practices of the American Oil
Chemists Society, 4th ed., American Oil Chemists Society, Champaign, Illinois, Method
Ch 2-91.
72. C. Paquot and A. Hautfenne, eds, Standard Methods for the Analysis of Oils, Fats and
Derivatives, (IUPAC), 7th ed., Blackwell Scientific Publications, Boston, 1987, Sec-
tion 2.210.
73. D. Firestone, ed., in Ref. 71, Method Ch 3-91.
74. A. Dieffenbacher and W. D. Pocklington, eds., Standard Methods for the Analysis of Oils,
Fats and Derivatives, (IUPAC), 1st Supple. to the 7th ed. Blackwell Scientific Publica-
tions, Boston, 1992, Section 2.324.
75. IOOC, IOOC/T.20/Doc. No. 9, 1991; IOOC/T.20/Doc. No. 18.3, Annex 2; IOOC/T.20/
Doc. No. 19-2, Annex 1, Madrid, June 1990.
76. Riv. Ital. Sostanze Grasse 68, 646 (1991).
77. M. Amelio and co-workers, Riv. Ital. Sostanze Grasse 70, 561 (1993).
78. Riv. Ital. Sostanze Grasse 66, 368 (1989).
79. M. Amelio, R. Rizzo, and F. Varazini, J. Amer. Oil Chem. Soc. 70, 793 (1993).
80. C. Paquot and A. Hautfeene, eds., Standard Methods for the Analysis of Oils, Fats and
Derivatives, (IUPAC) 7th ed., Blackwell Scientific Publications, Boston, 1987, Section
2.431.
81. J. Gracian Tous, W. D. Pocklington, and A. Hautfenne, Rev. Franc. Corps Gras 34, 465
(1987).
82. K. Grob, A. Artho, and C. Mariani, Fat Sci. Technol. 94, 394 (1992).
83. International Union of Pure and Applied Chemistry, Commission on Oils, Fats and
Derivatives, Minutes of the Meeting in Budapest, Sept. 911, 1992, pp. A9A11; meeting
at Lisbon, working report, July 1993, pp. A1A4; meeting at Bruges, Belgium, working
report, July 1994, pp. A1A2 and M1M11.
84. IOOC, Spain, IOOC/T.20/Doc.9 (1991), Doc.11/Rev. 1 (1994), Doc. 12 (1993), Doc. 22-2,
Annex 3 (1992), Doc. 28-4, Annex 2 (1993) Madrid; report on the collaborative trial run
by the Italian Technical Board on the determination of steroidal hydrocarbons in olive oils,
July 1994.
85. Italian Method NGD C40 (1976); see also D. Firestone, ed. in Ref. 71, Method Ch 5-91.
86. E. Fedeli, Riv. Ital. Sostanze Grasse 70, 81 (1993).
87. Technical methods, IOOC, www.internationaloliveoil.org, accessed july 2004.
88. C. Paquot and A. Hautfenne, eds., in Ref. 80, Section 2.201.
91. A. Dieffenbacher and W. D. Pocklington, eds., in Ref. 74, Standard Methods Section 2.432.
92. D. Firestone, ed., in Ref. 71, Method Ce8-89.
93. C. Paquot and A. Hautfenne, eds., in Ref. 80, Section 2.601 or 2.603.
94. A. Dieffenbacher and W. D. Pocklington, eds., in Ref. 74, Section 2.607.
95. M. Matta, G. Gaetano, and M. A. Giannandrea, Riv. Ital. Sostanze Grasse 66, 581 (1989).
96. Official Journal of the European Communities, No. L 166/12, (July 1, 1988).
97. A. Dieffenbacher and W. D. Picklington, eds., in Ref. 74, Section 2.631.
98. D. Firestone, ed., in Ref. 71, Method Cc 9a-48.
8
Palm Oil
Yusof Basiron
1. INTRODUCTION
1.1 Scope
The rapid expansion in world production of palm oil over the last three decades has
attracted the attention of the oils and fats industry. Many are interested to know how
palm oil has been able to compete successfully to gain an increasing share of the
international oils and fats markets. The increasing importance of palm oil has natu-
rally led to a steady buildup of scientific, technical, and trade data and information.
Many palm oil producing countries have established dedicated organizations and
research institutes that generate data and information to add to the body of knowl-
edge on oil palm cultivation, palm oil processing, and applications. It would be
impossible to include the voluminous body of existing and new information as
space constraint necessitates the coverage to be selectively confined to describing
only the essential aspects of the palm oil industry. Specialists in the field could refer
to numerous books and journals or databases that are now available on the subject.
This chapter will provide the reader with an understanding of the role of the
palm oil industry in the international oils and fats market, the technology involved
in extracting and processing of the oil, and the various quality parameters useful in
understanding the applications of palm oil products. Some of the common applica-
tions of palm oil in food and nonfood end products are described.
333
334 PALM OIL
1.2. General Considerations

Palm oil is an edible oil referred to by the FAO/WHO Codex Alimentarius (1) as
being derived from the fleshy mesocarp of the oil palm fruit. In the unprocessed
form palm oil is reddish brown in color, and it has a semisolid consistency at ambi-
ent temperature. Readers should not confuse palm oil with palm kernel oil, which is
another product obtained from the kernel of the oil palm fruit while palm oil is
derived from the mesocarp or fruit flesh. The two oils have different chemical com-
position and physical characteristics, and they are used and marketed separately
according to their own supply and demand situations.
1.3. Production
World production of palm oil had increased tremendously during the last 30 years
as a result of rapid expansion of oil palm planting in South East Asian countries
spearheaded by Malaysia and Indonesia. Papua New Guinea is also a significant
producer. Significant amounts of palm oil continue to be produced by the traditional
producer countries in West Africa but the growth was much slower. Nevertheless,
toward the end of the 1980s, Cote dIvoire (Ivory Coast) has emerged as a leading
palm oil producer and exporter in Africa with projection of further expansion in its
production in the future. Countries of South America are also striving to expand
their cultivation of oil palm, and increasing output of palm oil is projected for
Columbia, the leading producer from the region.
The continuing investment in oil palm cultivation in South East Asia, Africa, and
Central and Latin America especially in the late 1980s and early 1990s contributed
further to the growth in the future share of palm oil in the world supply of oils and fats.
Many countries plant oil palm to produce the oil to fulfill their local consump-
tion. In contrast, Malaysia and to a certain extent Indonesia are unique in that the
production of palm oil is meant for export. For these countries, palm oil production
for export purposes is found to be highly viable, and oil palm has become a favorite
cash crop to replace other traditional crops such as rubber. The viability of palm oil
for export is determined by the ability of the oil palm to be grown successfully in
the country concerned. High yield of the palm throughout the year is essential to
achieve viability for the export market.
Oil palm grows well in the tropical climate within 5 north and south of the
equator. Ideal growing conditions include adequate rainfall of over 2000 mm per
year spread evenly through the year, adequate sunshine of over 2000 h per annum,
and moderately high temperature of 2533 C. Many countries keen to grow oil
palm unfortunately experience a few months of drought during each year, and
this will severely affect the yield of the palm. Monsoon rains that can cause flood-
ing and problems of fruit evacuation may also affect crop yield. Countries not
having the ideal conditions for growing oil palms are reported (2) to have high
cost of production to the extent that exporting of the product would not be viable.
World oilseed production, vegetable oils production, and protein meal production
are listed in Tables 12.
INTRODUCTION 335
TABLE 1. World Oilseed Production, 106 t 1995/96 to Date (2).
Years

Item 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03a
Production
Soybeans 124.90 132.22 158.07 159.82 159.90 175.10 183.78 184.49
Cottonseed 35.15 33.61 34.35 32.62 32.93 33.53 36.61 33.37
Peanuts 27.47 28.96 27.29 29.77 28.99 31.12 33.11 31.84
Sunflowerseed 25.72 23.80 23.21 26.63 27.22 23.29 21.25 23.33
Rapeseed 34.44 31.53 33.23 35.89 42.47 37.52 35.87 32.17
Copra 5.13 6.05 5.33 4.38 5.46 5.90 5.26 5.30
Palm kernel 4.87 5.21 5.05 5.62 6.41 6.91 7.24 7.40
Total 257.67 261.38 286.53 294.72 303.37 313.36 323.10 317.89
a
Forecast.
TABLE 2. World Vegetable Oils and Protein Meal Production, 106 t 1995/96 to Date (2).
Years

Item 1995/96 1996/97 1997/98 1998/99 1999/2000 2000/01 2001/02 2002/03a
Production, Vegetable oil

Soybeans 20.17 20.53 22.57 24.65 24.74 26.80 28.72 29.85
Palm 16.26 17.64 16.97 19.25 21.80 23.93 24.88 25.37
Sunflowerseed 9.01 8.61 8.29 9.18 9.63 8.41 7.57 8.32
Rapeseed 11.24 10.52 11.43 11.81 13.64 12.96 12.20 11.41
Cottonseed 4.15 3.70 3.70 3.57 3.57 3.52 3.82 3.56
Peanut 4.15 4.38 4.18 4.44 4.15 4.30 4.75 4.51
Coconut 3.16 3.69 3.29 2.71 3.34 3.63 3.26 3.23
Olive 1.45 2.46 2.53 2.50 2.37 2.48 2.53 2.35
Palm Kernel 2.10 2.22 2.20 2.43 2.75 2.95 3.11 3.17
Total 73.08 73.76 75.16 80.54 85.97 88.98 90.85 91.79
Production, Protein meal
Soybeans 89.08 90.82 98.84 107.54 107.74 116.47 124.71 129.58
Cottonseed 13.11 11.89 11.79 11.36 11.45 11.30 12.10 11.31
Rapeseed 18.58 17.53 18.85 19.12 22.27 21.18 19.99 18.64
Sunflowerseed 10.21 10.06 9.51 10.51 10.72 9.43 8.45 9.25
Fish 6.52 6.64 5.08 5.80 6.29 5.75 5.43 5.61
Peanut 5.73 6.01 5.41 5.76 5.27 5.52 6.13 5.79
Copra 1.74 1.97 1.74 1.44 1.77 1.90 1.68 1.70
Palm Kernel 2.54 2.70 2.67 2.93 3.32 3.56 3.75 3.82
Total 147.49 147.62 153.88 164.47 168.82 175.12 182.23 185.69
a
Forecast.
336 PALM OIL
Figure 1. Global vegetable oil stocks, 2002/03 preliminary. (Source: Foreign Agricultural
Service, USDA.)
Global palm oil production edged up just 1.0 million tons in 2001/02 to
24.9 million. After several years of large gains in production, the rate of increase
in new oil palm area in Southeast Asia was slowing. Last year, Malaysia imple-
mented a replanting program for older trees that covered nearly 200,000 hectares.
Malaysian oil yields were lower as palm trees showed signs of stress. Reduced fer-
tilizer application and very dry conditions beginning in February also hurt produc-
tivity. Waning yields from Malaysian plantations trimmed 2001/02 production to
11.7 million tons from 11.9 million in 2000/01. Indonesias younger plantations
helped its growth in palm oil production to exceed Malaysias, which rose from
7.9 million to 8.8 million tons. Global vegetable oil stocks for 19972002 are repre-
sented in Figure 1 (2).
Although Malaysian palm oil output in 2001/02 failed to keep up with the pre-
vious years level, relatively large beginning stocks sustained a stable export pace.
Both Malaysian and Indonesian exports benefited from Argentinas difficulties in
exporting soybeans and soybean oil in 2002. Malaysian palm oil exports for
2001/02 steadied around 10.35 million tons while Indonesian exports expanded
to 5.5 million tons. Like vegetable oil stocks in the United States and EU, palm
oil stock in Malaysian and Indonesia gradually declined. Malaysian palm oil stocks
were 1.1 million tons at the end of September 2002 compared to 1.5 million in early
2001. Tightening stocks buoyed the Malaysian palm olein price to $388 per ton by
September 2002. This was the highest price level since early 1999 and much higher
than the September 2001 value of $274 per ton (2).
1.4. Palm Oil Trade

Although many countries are involved in the production of palm oil, only a few are
net exporters of the commodity. The net exporting countries are those where oil
INTRODUCTION 337
palm can grow well to make it viable to produce the oil for export. Only Malaysia
and Indonesia are the major net exporters of palm oil, while the other exporters
have only a small share, each accounting for not more than 3% of the total export.
Thus, for some years, the sources of palm oil have been confined mostly to the two
major exporting countries as they account for more than 90% of the total exports of
palm oil.
Crude palm oil used to be the main form of export in the past. With the establish-
ment of refineries especially in Malaysia during the mid-1970s and 1980s, refined
palm oil products have replaced the crude as the main form of palm oil export. A
wide range of processed or semiprocessed products are exported, and these include
the different fractions of processed palm oil known as palm olein (liquid) and
palm stearin (solid). The availability of refineries also led to the production of spe-
cialty fats products aimed at the confectionery markets. A similar trend has been
seen in the export of palm kernel oil. Palm kernel oil is a coproduct to palm oil
produced at a ratio of 1013 tons of palm kernel oil for every 100 tons of palm
oil. Even the export of refined palm kernel oil has begun to decline as more is
being used locally by the oleochemical industry that has been established in recent
years.
Most major buyers of palm oil products use the NIOP or FOSFA contracts to
secure their palm oil supplies. Other major buyers such as India, Pakistan, and Chi-
na have their own trading specifications. Palm oil prices are quoted in the terminal
markets such as Rotterdam, New York, and Kuala Lumpur. There is a futures mar-
ket for palm oil in Kuala Lampur, and this is actively used as the reference point for
price determination. A network of brokers and dealers are involved in facilitating
trade in palm oil products. In addition, some major multinational buyers have estab-
lished their buying offices in the producing countries. Sellers also participate in
responding to tenders called by a number of importing countries for the supply
of palm oil. In this way, palm oil has been exported through many different chan-
nels and mechanisms to suit consumer needs.
Trade is facilitated by the existence of bulking installations at the major ports of
loading for the export of the palm oil products. Codes of practice for the handling
and shipment of palm oil have been formulated by the international trade associa-
tions to ensure the quality of the oil is protected. For example, the trading contracts
such as FOSFA and NIOP stipulate that the previous cargoes of the ship carrying
palm oil must not be any from the list of banned substances. Efforts are made to
continuously upgrade the quality of palm oil products through improvements in
standards, and these are discussed at international forum such as the Codex Alimen-
terius meetings. Many mills and refineries are also adopting the ISO 9000 to pro-
vide quality assurance for the products that they export.
The major importers of palm oil used to be the developed countries of the Eur-
opean Economic Community (EEC), the United States, and Japan. They accounted
for about 75% of the imports of palm oil in the early 1970s. With the increasing
exports of refined palm oil products, many developing countries, which did not
have refining capacities, were able to import processed palm oil for direct consump-
tion with minimal or no further refining. This helped to expand the market for palm
338 PALM OIL
oil in the developing countries. By the end of the 1980s, the developing countries
have become the major consumers of palm oil accounting for 75% of the import
trade. While the import share of palm oil by the developed countries has declined
to about 25%, the actual volume consumed by them has continued to expand. This
reflects the competitiveness of palm oil in terms of price and technical suitability
for all types of end uses both in the developed and developing countries.
Many countries are facing chronic shortages of oils and fats due to shortfall in
the domestic production in the face of increasing population and income. For these
countries, the relative availability of palm oil provides a convenient source of sup-
ply. Because of the rapidly expanding supply of palm oil, its price has been very
competitive. For the last 30 years, palm oil has been selling at a discount compared
to the other major oils in the world market, but prices of palm oil are highly corre-
lated with those of the other oils. This suggests that the market acknowledges the
high degree of substitutability of other oils and fats by palm oil.
Indias vegetable oil consumption still rose steadily in 2001/02, but moderated
from a robust 2000/01 growth rate of 11 percent. A larger domestic oilseed harvest
and a paring of stocks dampened import requirements. Total vegetable oil imports
by India (which surged by nearly one-fourth in 2000/01) declined to 5.2 million
tons in 2001/02 from 6.0 million. Imports of palm oil and soybean oil dipped to
3.4 million and 1.65 million tons, respectively. Neglible quantities of rapeseed
oil and sunflowerseed oil were imported, as they became less price-competitive
because of an import duty structure that favored soybean oil and crude plam oil.
India has not materially changed vegetable oil tariffs since October 2001, when
it cut the rate on crude palm oil from 75 percent to 65 percent.
Palm oil exporters had hoped that China would replace lagging Indian sales by
raising its import quota. China officially entered the WTO on December 11, 2001.
Chinas accession agreement stipulated that its 2002 tariff-rate quota (TRQ) on soy-
bean oil increase to 2.518 million tons and the within-quota tariff fall from 13 per-
cent to 9 percent. Tariffs on soybeans and soybean meal were bound at their
previous rates. But ample domestic production of soybean and rapeseed oils con-
tinued to limit Chinas need for vegetable oil imports.
China had originally set issuance of its vegetable oil import licenses by March 5
but had only begun distributing them in early April. Two-thirds of the annual 2.4-
million ton quota was to be allocated to private importers. Nevertheless, palm oil
imports by China surged in March. Before the quota, China had already imported
about 300,000 tons this year, some of which were waiting at ports in bonded ware-
houses for the licenses to be distributed. About half of the palm oil imports were
allowed to clear customs before April because importers could deposit a 52-percent
over-quota tariff for them. When the importers received their quotas, the differences
against the 9-percent-within-quota tariff was refunded.
China did not require foreign exporters to obtain separate safety certificates for
each cargo of soybean oil produced from biotech varieties. However, soybean oil
imports were temporarily handicapped by a requirement that safety certificates
be approved for biotech soybeans before the same applications for soybean oil
can be accepted.
CHEMICAL AND PHYSICAL PROPERTIES OF PALM OIL 339
A tightening of Chinas soybean and rapeseed supplies by mid-2002 created

opportunities for vegetable oil imports. Palm oil was the most favorably priced
and imports were unfettered by the countrys requirements for safety certificates,
inspections, and labeling of biotech oilseeds. Therefore, Chinas importers tried to
fill their increased 2002 palm oil TRQ (2.4 million tons) first. Palm oil imports by China
rose to 2.0 million tons from 1.6 million in 2000/01. For soybean oil, rising world
prices narrowed the differential to Chinas domestic prices, which limited its import
needs. Chinas soybean oil imports were 375,000 tons in 2001/02, still well below
the TRQ but substantially above the 80,000 tons imported the previous year (2).
The versatility of palm oil in terms of its presentation of various subproducts and
the wide range of technical properties increase the competitiveness of palm oil to
the consumers. Palm oil has become the major oil among the imported oils in most
countries. Even countries that are net exporters of oils and fats, such as the United
States, are importing palm oil in substantial quantities. For these countries palm oil
can provide certain technical advantages in some end uses, and palm oil usage gives
better margins of profit compared to the use of locally available oils and fats.
2. CHEMICAL AND PHYSICAL PROPERTIES OF PALM OIL
Palm oil, like all oils and fats, is made up mostly of glyceridic materials with some
nonglyceridic materials in small or trace quantities. It is this chemical composition
that defines the chemical and physical characteristics of palm oil, which in turn will
determine the suitability of the oil in various processes and applications.
2.1. Chemical Properties of Palm Oil

Triglyceride and Fatty Acid Composition Triglycerides form the major component
and bulk of the glyceridic material present in palm oil with small amounts of mono-
glycerides and diglycerides, which are artifacts of the extraction process. The fatty
acid chains present in the palm oil triglycerides could vary in the number of carbons
present in the chain (chain length) and in structure (presence of double bonds, i.e.,
unsaturation). It is the variations in the structure and number of carbons in
these fatty acid chains that largely define the chemical and physical properties of
palm oil.
The chain lengths of the fatty acids present in the triglycerides of palm oil fall
within a very narrow range from 12 to 20 carbons as shown in Table 3.
It can be seen that about 50% of the fatty acids present in palm oil are saturated
and about 50% are unsaturated. This even balance between saturation and unsatura-
tion determines the iodine value of the oil (about 53) and confers some stability
against oxidation to the oil as compared to other vegetable oils. The three fatty
acids in the triglyerides could be represented by the multitude of fatty acids listed
in Table 3. The different placement of the fatty acids attached to the glycerol mole-
cule can lead to a large number of different triglycerides. Subjecting lipolysis data
to statistical computer analysis, more details about the triglyceride composition of
340 PALM OIL
TABLE 3. Fatty Acid Composition of Malaysian

Palm Oil (3).
% of Total
Fatty Acids Chain Lengths Mean Range
12:0 0.23 0.11.0

14:0 1.09 0.91.5
16:0 44.02 41.846.8
16:1 0.12 0.10.3
18:0 4.54 4.25.1
18:1 39.15 37.340.8
18:2 10.12 9.111.0
18:3 0.37 00.6
20:0 0.38 0.20.7
palm oil could be obtained as shown in Table 4. These are in terms of the actual
acid chains present in the three carbon positions of the triglyceride molecule.
From the computation data, it can be seen that the triglyceride molecules could
be divided according to the number of saturated (S) and unsaturated (U) groups
that they contain. The computational results were found to be very close to the
analytical data obtained by Tan (6) shown in Table 5.
Carbon number analysis by high-temperature gas-liquid chromatography pro-
vides a rapid but less detailed analysis of the triglyceride composition in terms
of the total number of carbon atoms present in the three fatty acid chains of
the triglyceride molecule. Carbon number analysis data of palm oil are shown in
Table 6.
Knowledge about the detailed structures of the triglycerides present in palm oil
is important because they define some of the physical characteristics of the oil. The
melting points of triglycerides are dependent on the structures and position of the
component acids present. They also affect the crystallization behavior of the oil.
The semisolid nature of palm oil at room temperature has been attributed to the
presence of the oleo-disaturated fraction.
As mentioned previously, partial glycerides are artifacts of the extraction pro-
cess, especially the stages prior to sterilization. Oil obtained from unbruised steri-
lized fruits shows trace levels of partial glycerides. Random analyses of samples of
refined palm oil, palm olein, and palm stearin have shown the presence of about 2%
of 1,2-diglycerides and about 4% of 1,3-diglycerides with trace amounts of mono-
glycerides. These partial glycerides are important as they are known to affect the
crystallization behavior of the oil.
Minor Components. The carotenoids, tocopherols, sterols, phosphatides, triterpe-
nic, and aliphatic alcohols form the minor constitutents of palm oil. Though present
in less than 1% altogether in palm oil, nevertheless they play a significant role in the
stability and refinability of the oil, in addition to increasing the nutritive value of
the oil.
Crude palm oil contains between 500 and 700 ppm of carotenoids mainly in the
forms of a-and b-carotenes, the precursor of vitamin A. Unless extracted prior to
TABLE 4. Triglyceride Composition of Malaysian Tenera Palm Oil.a
No Double Bond 1 Double Bond 2 Double Bond 3 Double Bond 4 or More Double Bonds
A B A B A B A B A B
MPP 0.29 0.5 MOP 0.83 1.4 MLP 0.26 MLO 0.14 0.2 PLL 1.08 0.8
PMP 0.22 0.2 MPO 0.15 0.2 MOO 0.43 0.7 PLO 6.59 6.0 OLO 1.71 1.4
POP 20.02 23.7 PLP 6.36 6.3 POL 3.39 3.1 OOL 1.76 1.5
PPP 6.91 7.2 POS 3.50 3.1 PLS 1.11 0.8 SLO 0.60 0.4 OLL 0.56
PPS 1.21 1.0 PMO 0.22 PPL 1.17 1.0 SOL 0.30 0.2 LOL 0.14 0.1
PSS 0.12 0.1 PPO 7.16 6.9 OSL 0.11 OOO 5.38 5.1
PSP 0.7 PSO 0.68 0.6 SPL 0.10 0.1 OPL 0.61 0.5
SOS 0.15 POO 20.54 21.5 MOL 0.1
SPO 0.63 0.5 SOO 1.81 1.4
OPO 1.86 1.6
OSO 0.18 0.2
PSL 0.1
Others 0.16 0.34 0.3 0.19 0.6 0.15 0.22
Total 9.57 9.7 33.68 35.8 34.12 34.6 17.16 15.6 5.47 3.8
a
A: based on Kan-Ichi Hayakawa Computation: see Ref. (4); B: based on Vander Wals method: see Ref. (5).
342 PALM OIL
TABLE 5. Triglyceride Analysis of Tenera Palm Oil

(6) (based on saturationunsaturation criterion).
Triglyceride Type Composition (%)
Trisaturated (GS3) 10.2

Disaturated (GS2U) 48.0
Monosaturated (GSU2) 34.6
Triunsaturated (GU3) 6.8
TABLE 6. Carbon Number Analysis of Malaysian

Palm Oil (3).
Carbon Number Mean Range
C46 0.8 0.41.2

C48 7.4 4.710.8
C50 42.6 40.045.2 (POP, PPO)
C52 40.5 38.243.8 (POO)
C54 8.8 6.411.4
refining, these carotenoids are thermally destroyed during the deodorization stage in
order to produce the desired color for a refined oil. In crude palm oil, the presence
of these carotenoids appears to offer some oxidative protection to the oil through a
mechanism where they are oxidized prior to the triglycerides. Table 7 lists the car-
otenoid types that are present in crude palm oil.
Crude palm oil contains tocopherols and tocotrienols in the range of 600
1000 ppm. Refined palm oil retains about 50% of these products. Tocopherols
and tocotrienols are antioxidants and provide some natural oxidative protection
to the oil. Table 8 shows the types of tocopherols and tocotrienols present in
palm oil.
TABLE 7. Carotenoid Composition of Palm Oil (7).
Carotenoid Percentage
Phytoene 1.27
cis-b-Carotene 0.68
Phytofluene 0.06
b-Carotene 56.02
a-Carotene 35.16
cis-a-Carotene 2.49
x-Carotene 0.69
g-Carotene 0.33
d-Carotene 0.83
Neurosporene 0.29
b-Zeacarotene 0.74
a-Zeacarotene 0.23
Lycopene 1.30
Total carotene (ppm) 673
TABLE 8. Tocopherols and Tocotrienols in Crude

Palm Oil (8).
Type Percentage
a-Tocopherols 21.5
b-Tocopherols 3.7
g-Tocopherols 3.2
d-Tocopherols 1.6
a-Tocotrienols 7.3
b-Tocotrienols 7.3
g-Tocotrienols 43.7
d-Tocotrienols 11.7
From Table 8, it can be seen that a-tocopherols and g-tocotrienols account for
the major portion of the total tocopherols and tocotrienos present in palm oil. Gapor
(9) confirms the presence of the above-listed tocopherols and tocotrienols by
high-performance liquid chromatography (HPLC) and also indicated the probable
presence of the esterified forms.
The combined effects of the properties of the carotenoids, tocopherols, tocotrie-
nols and the 50% unsaturation of the acids confer on palm oil a higher oxidative
stability as compared to a lot of other vegetable oils.
In terms of sterols, palm oil contains far less cholesterol than many other vege-
table oils as shown in Table 9. Table 10 gives the sterol composition of crude and
TABLE 9. Cholesterol Levels in Crude Oils and Fats (10).
Oil Type Average (ppm) Range (ppm)
Coconut oil 14 524

Cocoa butter 59 n.a.
Palm kernel oil 17 940
Palm oil 18 1319
Sunflower oil 17 844
Soybean oil 28 2035
Rapeseed oil 49 2580
Maize oil 50 1895
TABLE 10. Sterol Composition of Crude and Refined Palm Oil and Their Products (ppm)
(11).
Sample Cholesterol Campesterol Stigmasterol Sitosterol Unknown
Crude palm oil 713 90151 4466 218370 218

Degummed, bleached 510 49116 2251 113286 Trace-8
RBD 15 1516 830 45167 Trace
Crude palm olein 68 57104 3051 149253 2428
Degummed, bleached 34 3643 2125 99123 Trace5
RBD 2 2630 1223 68114
344 PALM OIL
TABLE 11. Inherent Chemical Properties of Malaysian Palm Oil

(12).
Chemical Characteristics Mean Range
Saponification value (mg KOH/g oil) 195.7 190.1201.7

Unsaponifiable matter (%) 0.51 0.150.99
Iodine value (Wijs) 52.9 50.655.1
Slip melting point ( C) 34.2 30.837.6
refined palm oil and their products. From Table 10, it can be seen that the low-
cholesterol levels in crude palm oil and crude palm olein are further reduced
to even lower levels upon refining. This low-cholesterol level, together with the
antithrombotic and anticarcinogenic properties of some of the carotenoids, tocopher-
ols, and tocotrienols present add further to the nutritive value of palm oil and palm oil
fractions.
Inherent Chemical Properties of Palm Oil. Table 11 summarizes some of the
inherent chemical properties of Malaysian palm oil.
2.2. Physical Properties of Palm Oil

Table 12 shows some of the physical properties of palm oil. The apparent density is
an important parameter from the commercial point of view since it is used for
volume to weight conversions. It can also be used as a purity indicator.
The solid fat content of an oil is a measure (in percent) of the amount of solid fat
present in the oil at any one temperature. It is measured by means of wide-line
nuclear magnetic resonance (NMR) spectrometry after a standard tempering
procedure for the samples.
TABLE 12. Major Physical Properties of Palm Oil (13,14).
Property Mean (of 215 Samples) Range
Apparent density at 50 C (g/mL) 0.889 0.8880.889

Refractive index at 50 C 1.455 1.4551.456
Solid fat content
5 C 60.5 50.768.0
10 C 49.6 40.055.2
15 C 34.7 27.239.7
20 C 22.5 14.727.9
25 C 13.5 6.518.5
30 C 9.2 4.514.1
35 C 6.6 1.811.7
40 C 4.0 0.07.5
45 C 0.7
Slip melting point ( C) () 34.2 31.137.6
Figure 2. Crystallization thermogram of Malaysian palm oil (15).
Figure 3. Melting thermogram of Malaysian palm oil (15).

346 PALM OIL
The solid present in the oil at any one temperature is due to the process of crys-
tallization occurring in the oil as a consequence of its chemical properties. The dif-
ferent molecular triglyceride structures with their differing chemical characteristics
manifest their physical states at different temperatures, thus imparting certain crys-
tallization and melting behavior to the oil. These thermally associated processes can
be followed by means of differential scanning calorimetry (DSC). Figures 2 and 3
show the crystallization and melting thermograms of palm oil, respectively. In the
crystallization thermograms, points 1 to 2 define the olein crystallization peak while
points 2 to 3 define the stearin crystallization peak. These are defined by points 1 to
2 and 2 to 3, respectively, in the melting thermogram.
From the thermal characteristics considered above, it can be seen that palm oil
can be separated under controlled thermal conditions into two components, i.e., a
solid (stearin) and a liquid (olein) fraction. This fractionation process can be
affected either in the dry form in the presence of a detergent or solvent. The method
employed, to a certain extent, determines some of the chemical and physical prop-
erties of the oleins and stearins produced, especially the stearins. By varying the
fractionation methods and conditions used, a range of stearins with differing che-
mical and physical properties could be produced, yet keeping the chemical and
physical properties of the oleins to within a very narrow range of values as shown
in Tables 13, 14, and 15.
TABLE 13. Fatty Acid Compositions (%) (16).
Oleins Stearin
Fatty Acids Range Observed Mean Range Observed
12:0 0.10.5 0.2 0.10.6
14:0 0.91.4 1.0 1.11.9
16:0 37.941.7 39.8 47.273.8
16:1 0.10.4 0.2 0.050.2
18:0 4.04.8 4.4 4.45.6
18:1 40.743.9 42.5 15.637.0
18:2 10.413.4 11.2 3.29.8
18:3 0.10.6 0.4 0.10.6
20:0 0.20.5 0.4 0.10.6
Iodine value (Wijs) 56.160.6 58.0 21.649.4
TABLE 14. Triglyceride Composition by Carbon Number (16).
Oleins Stearin
Carbon Number Range Mean Range
C46 0.53.3
C48 1.34.0 2.3 12.255.8
C50 37.745.4 42.0 33.649.8
C52 43.351.3 45.7 5.137.3
C54 7.012.6 9.9 Trace8.4
TABLE 15. Melting and Solidification Characteristics (16).
Oleins Stearin
Tests Range Mean Range
Slip melting point ( C) 19.423.5 21.6 44.556.2
Cloud point ( C. crude) 6.614.3 10.4
Neutralized 5.411.9 8.1
Refined 6.011.5 8.8
Under normal fractionation conditions, soft stearins and oleins with cloud points
in the range of 810 C are produced. Where required, fractionation conditions
could be specifically altered to produce stearin or olein of a desired specification
for specialized application, but within the domain of the composition of palm
oil. For example, stearins of differing iodine values (IV) ranging from the hard
(IV of about 20) to soft (IV of about 50) could be produced, each with their
characteristic solid fat content curve as shown in Figure 4.
Figure 4. Solid fat content of crude stearin (15).

348 PALM OIL
Figure 5. Fractionation and palm midfraction.
TABLE 16. Palm Midfractions Solids Content by NMR.
Code Temperature ( C) Palm Oil 1 2 3 4 6 7 8 5A 5B 5C
10 50.3 81.3 58.4 60.1 61.5 65.6 66.4 53.8 71.9 51.8 71.1
15 35.2 71.1 37.1 48.5 45.2 48.3 55.2 41.6 63.2 32.9 61.8
20 23.2 59.5 18.5 34.3 26.8 30.5 42.0 27.1 43.3 17.5 45.0
25 13.7 29.7 1.7 22.8 8.7 8.2 28.7 15.8 22.1 9.3 28.0
30 8.5 8.6 14.2 2.4 2.9 19.6 9.6 11.4 4.6 16.8
35 5.8 3.6 10.4 1.2 15.1 5.6 6.6 12.0
40 3.5 7.1 10.9 2.6 2.5 7.3
45 3.4 5.7 3.0
50
For more specialized usage such as in the confectionery industry, a more specific
type of stearin is required or desired. This is catered to by using a double fractiona-
tion process as shown in the scheme in Figure 5. Table 16 shows the solid fat con-
tent profiles of the more common palm midfractions produced in Malaysia at
present.
Other characteristics of palm oleins and palm stearin are shown in Table 17. The
different types of palm oil products available for export are illustrated in Table 18.
2.3. Test Methods for Palm Oil and Products Analysis

In the palm oil trade, test methods for palm oil product analysis were traditionally
based on the American Oil Chemists Society (AOCS) test methods. Where test
methods are not available under the AOCS, other test methods were used, drawn
TABLE 17. Additional Analytical Characteristics (16).
Tests Oleins Stearins

Apparent density at 40 C (g/mL) 0.89650.8992 (n 21)
60 C (g/mL) 0.86590.8756 (n 40)
Refractive index nD 40 C 1.45861.4592 (n 21)
nD 60 C 1.44721.4511 (n 41)
Saponification value mg KOH/g oil 194202 (n 21) 193206 (n 41)
TABLE 18. PORAM Standard Specification for Processed Palm Oil.
Type of Palm Oil Specificationa Values
1. Neutralized palm oil *FFA (as palmitic) 0.25% Max.

Moisture and Impurities 0.1% Max.
IV (Wijs)
Melting point ( C) 5055
(AOCS Cc 325) 3339
2. Neutralized and
bleached palm oil *FFA (as palmitic) 0.25% Max.
IV (Wijs)Melting point ( C) 5055
(AOCS Cc 325) 3339
Color (5 14-inch Lovibond cell) 20 Red max.
3. Refined, bleached, and *FFA (as palmitic) 0.1% Max.
deodorized (RBD)/neutralized, Moisture and Impurities 0.1% Max.
bleached, and deodorized (NBD) IV (Wijs)Melting point ( C) 5055
palm oil (AOCS Cc 325) 3339
Color (5 14-inch Lovibond cell) 3 or 6 Red max.
4. Crude palm olein *FFA (as palmitie) 5.0% Max.
IV (Wijs)Melting point ( C) 56 Min.
(AOCS Cc 325) 24 Max.
5. Neutralized palm olein *FFA (as palmitic) 0.25% Max.
6. Neutralized and bleached
palm olein *FFA (as palmitic) 0.25% Max.
7. Refined, bleached, and *FFA (as palmitic) 0.1% Max.
deodorized (RBD)/neutralized, Moisture and Impurities 0.1% Max.
bleached, and deodorized (NBD) IV (Wijs)Melting point ( C) 56 Min.
palm olein (AOCS Cc 325) 24 Max.
350 PALM OIL
Type of Palm Oil Specificationa Values
8. Double fractionated palm olein *FFA (as palmitic) 0.1% Max.

9. Crude palm stearin *FFA (as palmitic) 5.0% Max.
IV (Wijs)Melting point ( C) 48 Max.
(AOCS Cc 325) 44 Min.
10. Neutralized palm stearin *FFA (as palmitic) 0.25% Max.
11. Neutralized and bleached palm *FFA (as palmitic) 0.25% Max.
stearin Moisture and Impurities 0.15% Max.
12. Refined, bleached, and deodorized (RBD)/neutralized, bleached,
and deodorized (NBD) palm
stearin 0.2% Max.
*FFA (as palmitic)
13. Palm acid oil *Total fatty matter 95% Min.
(basis 97%)
Moisture and Impurities 3% Max.
*FFA (as palmitic) 50% Min.
14. Palm fatty acid distillate Saponifiable matter 95% Min.
(basis 97%)
*FFA (as palmitic) 70% Min.
a
, Slip point, softening point, or rising point; *, molecular weight of palmitic acid is taken as 256.
Source: PORAM Technical Brochure, 1989.
from the British Standards Institute (BSI), International Union of Pure and Applied
Chemistry (IUPAC) methods, or the International Standards Organization (ISO).
However, the recent trend has been toward the adoption and standardization of ISO
methods. In early 1994, FOSFA have called for a harmonization of test methods
between the various test method organizations such as the American Official &
Analytical Chemistry (AOAC), AOCS, IUPAC, BSI, and other national standards
organizations. The aim of this exercise is to standardize the test methods for oils
and fats from all these organizations for trade purposes, to be placed under the ISO.
PRODUCTION PROCESS 351
3. PRODUCTION PROCESS
3.1. Origin of the Oil Palm

The oil palm, Elaeis guineensis Jacq., is grown commercially in Africa, South
America, Southeast Asia, and the South Pacific, and on a small scale in other tro-
pical areas. Until recent centuries the palm has been confined to West and Central
Africa where it existed in a wild, semiwild, and cultivated state. In Africa it
remained a domestic plant, supplying a need for oil and vitamin A in the diet,
and it was not until the end of the eighteenth and the beginning of the nineteenth
centuries that oil palm cultivation expanded to the Southeast Asian regions and
strengthened the entry of palm oil into the world oils and fats trade.
3.2. Oil Palm Plantations

The early development of the oil palm industry is well described by Hartley (17).
The export of palm oil and kernels from Africa began in the nineteenth century.
At this stage the only source of supply was the palm groves, the oil being extracted
by primitive and inefficient methods. The palm groves were sometimes developed
into peasant plantations (18) by deliberate planting of seedlings, but the first
large plantations were established in Sumatra and Malaysia in the early years of
this century. These were followed in the 1920s by plantations in the Belgian Congo
(now Zaire) and then in other parts of West Africa. In recent years very consider-
able further expansion of the industry has occurred, and oil palm products are now
an important component of world vegetable oil supplies.
Although the cultivation of the oil palm in plantations started in the Far East,
strangely there was no direct connection between the African groves and the estab-
lishment of this new industry. The earliest record of the introduction of oil palms to
the East Indies is of four seedlings, two from Bourbon (Reunion) or Mauritius and
two from Amsterdam, which were planted in the Botanic Gardens at Buitenzorg,
now Bogor, in Java in 1848. The foundation of the industry is generally attributed
to M. Adrien Hallet, a Belgian with some knowledge of oil palms in Africa, who
planted palms of Deli origin in 1911 in the first large commercial plantations in
Sumatra. Hallet recognized that the avenue palms growing in Deli were not only
more productive than palms in Africa but had a fruit composition superior to the
ordinary dura palms of the African west coast. Open pollinated selected tenera
seed was used on commercial plantations as early as 1924 (19). In the meantime,
however, M.H. Fauconnier, who had been associated with M. Hallet, had estab-
lished during 1911 and 1912 some palms of Deli origin at Rantau Panjang in the
Kuala Selangor district of Malaysia. These palms were in full bearing by 1917,
and in that year the first seedlings were planted on an area later to be known as
Tennamaran Estate.
The industry grew rapidly in Sumatra, but did not gain its full momentum in the
Far East until the 1930s. In 1925 there were 31,600 hectares planted in Sumatra and
only 3,348 in Malaysia (20) but by 1938 the areas had risen to 92,300 and 29,196
352 PALM OIL
TABLE 19. Planted Oil Palm Area (1000 ha).
1960 1970 1980 1990 1993
Malaysia 54.6 291.3 1,023.3 2,029.5 2,281.0

Indonesia n.a. 134.0 294.1 1,126.7 1,603.7
Papua New Guineaa n.a. n.a. 12.0 37 55
Ivory Coast 5.2 68.2 100.3 128a 158a
Nigeriaa n.a. n.a. 230.0 270 298
Columbia 0.4 19.8 36.7 81a 111a
a
Denotes mature area only.
Source: Oil World Weekly, April 4, 1985. Oil World Annual, 1993. H. A. J. Moll, Economics of Oil Palm (Ref. 2),
Pusat Penelitian Marihat, Indonesia.
hectares, respectively. With over 120,000 hectares, an industry of considerable

importance capable of producing more oil than was being exported from Africa
had in the span of about 20 years been established. The rate of planting in Malaysia
was rapid in the 1960s and 1970s. In 1960 there were only 55,000 hectares, but by
1975 half a million hectares had been added and a total of over 2 million hectares
was reached by 1990 (21). A major part of the planting was done by federal and
state land development authorities, government-sponsored settlement schemes,
which by the early 1980s accounted for half the total planted area in Malaysia.
In Indonesia, oil palm cultivation has expanded rapidly especially during the
1980s, where government estate enterprises, foreign private estates, private national
estates, and nucleus estates have been established. The expansion in planted area
for oil palm for this historical period is shown in Table 19.
3.3. Yield of Palm Oil

The oil palm is a highly efficient producer of vegetable oil. On per unit area basis
the oil palm is considerably higher yielding than any other vegetable oil crops.
Record yields for other crops such as soybean are about 2 tons of oil per hectare,
3 tons for rapeseed and olive, and 4 for coconut and sunflower. In contrast, thou-
sands of hectares of oil palm plantations in Southeast Asia regularly yield 5 tons of
oil per hectaare per year, and record yields are appreciably higher. The figures
shown in Table 20 represent some of the highest oil yields recorded in experiments;
higher yields may well have been obtained by other workers. Individual palms with
over 30% oil to bunch ratio exist, with a fruit yield of 40 t/ha, this would give an oil
yield of 12 t/ha.
Oil Yield Components. Palm oil and palm kernel oil are obtained from the oil
palm fruit. Yield of oil can be considered in terms of various components; the
two main components are yield of fruit bunches and oilbunch weight ratio (or
extraction ratio). Fruit yield can be considered in terms of the component bunch
number and mean bunch weight. Bunch weight increases with palm age while
bunch number decreases. The first yield of fruit bunches normally ripen during
TABLE 20. Yield Records Obtained from Some Trials in Malaysia.
Yield of
Fruit Bunches Yield of
Type of Trial (t/ha) Oil/Bunch (%) Oil (t/ha)
Progeny trial (best progeny, 1 year) 30.4 27.7 8.42

Spacing trial (best treatment, 1 year) 28.5 23.7 6.76
Spacing trial (best plot, 1 year) 30.9 23.7 7.32
Fertilizer trial (best plot, 1 year) 37.2 16.9 (dura) 6.28
Estate planting (best plot, 3rd year of production) 32.6 26.2 8.55
the third year after field planting. Yield rises to a maximum in the first few years
and thereafter usually tends to decline slowly.
OilBunch Ratio. The yield of oil depends on yield of fruit bunches and a
further component, oilbunch ratio. The oilbunch ratio is the product of a
number of components: these are fruitbunch ratio, mesocarpfruit ratio, and
oilmesocarp ratio. The oil content of the fruit of young palms is low: Hartley
(17) stated that it increases steadily until the fourth or fifth year of bearing.
However, Corley (22) reported that with tenera palms, an oilbunch ratio of
over 28% may be reached as early as 40 months after field planting.
Mesocarpfruit ratio is largely genetically determined and is little affected by
environmental factors. Fruitbunch ratio depends mainly on the efficiency of pol-
lination. Oilmesocarp ratio depends in part on the ripeness of the fruit, since oil is
only synthesized during the later stages of fruit development. There is also
considerable variation in oilmesocarp of bunches from the same progeny
harvested at different times of the year (23). It has also been shown that application
of potassium fertilizer leads to a reduction in the oilbunch ratio (24); but the
increase in fruit yield in response to potassium was more than enough to compen-
sate for the reduced oilbunch ratio.
Yield Variation. In seasonal climates the annual yield of oil palm usually has
only one peak, the time of the peak depending on the age and leaf production of
the palms and, in the mature plantation palm, on climatic conditions about 28
months before fruit ripening. In nonseasonal climates, there are occasionally two
peaks of production in the year, though one tends to be much more prominent; there
is considerable variation in the magnitude of the peaks.
3.4. Production Costs

Production cost of crude palm oil is made up of cost of production of fresh fruit
bunches (FFB) and the cost to mill the FFB. Refining costs are incurred when crude
palm oil undergoes refining to produce processed palm products.
Cost of FFB Production. To produce FFB, the costs covered involve felling and
land preparation, lining, terracing, drainage, planting of palms and leguminous cov-
ers, preparing of roads, pathways, and bridges, application of fertilizers, and pest
and disease control. The cost of establishing oil palm also depends on whether
354 PALM OIL
the area is undergoing replanting or new planting. Replanting refers to planting on

land that was formerly developed with other crops while new planting refers to
establishing an area formerly under jungle. In Malaysia, the cost of establishing
a hectare of oil palm area for the first three years is about MR 5013.26, or
$1856.75 (all dollar amounts are in U.S. dollars unless specified) for areas under
replanting. For new plantings, the cost of establishment would be 2030% higher
because new plantings would require more intensive land preparations such as new
terraces, new drainage systems, and new roads and pathways.
After three years, the palms are already mature and fresh fruit bunches could be
harvested monthly for the next 25 years or more. The direct items involved in the
production of FFB during the mature period include costs of fertilizers and their
applications, harvesting, control of pests, weeds, and diseases, maintenance of
infrastructures such as roads, harvesting paths, and bridges, soil and foliar analysis,
and agricultural equipment. On the whole, the direct cost of FFB production ranges
from $238 to $520 per hectare per year while the cost of producing a ton of
FFB ranges from $14.80 to $44.40 (Table 21). The range in cost is very much
dependent upon FFB yield, type of soil, drainage, FFB quality, and other factors.
Variations in costs are also due to the management systems of the oil palm, viz.
plantations, unorganized small holdings, or organized small holdings.
The indirect costs to produce FFB or to maintain mature oil palm hectarages are
categorized into two areas, personal emoluments and services and supplies. Perso-
nal emoluments include staff salaries, staff costs and benefits, wages, workers costs
and benefits, and other costs and benefits. Services and supplies covers traveling,
office expenses, maintenance, professional expenses, utilities, and sundries. The
TABLE 21. FFB Production for Estate Sector, Group Smallholders, Independent
Smallholders (U.S. $/hectare).
Independent
Smallholders
Sector Costs Estate Group Smallholders U.S. $/ha Ave.
Direct costs
Manuring 51.9 (22.5) 100.0 (32.1) 175.7 109.2
Weeding 26.0 (10.8) 1.9 (0.6) 3.3 10.4
Supply of damaged palms 26.4 (11.0) 15.6 (5.0) 27.4 23.1
Pest/disease 7.4 (3.1) 9.0 (2.9) 15.9 10.8
Pruning 10.6 (4.4) 14.0 (4.5) 12.3
Harvesting and collection 96.9 (40.3) 69.5 (22.3) 228.1 131.5
Transport of FFB 17.6 (7.3) 95.7 (30.7) 62.9 58.7
Miscellaneous 1.4 (0.6) 5.9 (1.9) 7.4 4.9
Total 238.2 (100.0) 311.6 (100.0) 520.7 356.8
Indirect costs
Salaries and wages 138.1 (59) 117.4 (46) 85.17
Supply 95.9 (41) 135.6 (54) 11.1 80.9
Total 234.0 (100) 253.0 (100) 11.1 166
Total direct and indirect 472.2 564.6 531.8 522.9
TABLE 22. Milling Costs.
Cost/Ton FFB
RM U.S. Percent
Mill management 1.51 0.56 5.04

Mill process staff 1.46 0.54 4.85
Mill process labor 3.11 1.15 10.37
Machinery upkeep 5.25 1.94 17.50
Building upkeep 0.11 0.04 0.36
Utilities 2.68 0.99 8.93
Kernel bagging 0.58 0.21 1.93
Head office costs 3.21 1.19 10.71
Depreciation 10.71 3.97 35.71
Insurance 0.38 0.14 1.25
Palm oil cess 1.00 0.37 3.33
Total 30.00 11.1 100
indirect cost per year per mature hectare of oil palm in Malaysia ranges between
$234.00 and $253.00. From studies of estates, schemes, and small holdings in
Malaysia, the indirect cost per ton of FFB averages at $8.30.
Milling Cost. After the fresh fruit bunches are harvested they are sent to palm oil
mills where the oil is extracted and the nuts separated. The amount of crude palm
oil (CPO) obtained from the bunch is in the ratio of 1824% depending on the
planting materials.
As the mill receives and processes fresh fruit bunches, it is logical to base
milling cost on per ton of FFB received. Presently, the average cost of milling a
ton of FFB is in the region of MR 30.00, or $11.10 (Table 22). Table 22 illustrates
that the largest cost is depreciation, some 35.71%, followed by machinery upkeep,
which is 17.5%, and mill process labor being just a little above 10%.
Production Cost of CPO. The costs outlined previously are summarized in
Table 23. The establishment cost was amortized over 25 years and allocated to
the cost of producing FFB. An average yield of 20 tons of FFB per hectare per
year and an oil extraction rate of 20% were utilized for data conversions.
TABLE 23. Product Cost of CPO in Malaysia (U.S. $).
Per ha/year Per ton FFB
Establishment cost amortized (1856 25) 74.24 3.71

Average direct cost 256.80 17.84
Average indirect cost 166.00 8.30
Total cost of FFB production 597.04 29.85
Milling cost 11.10
Cost of FFB production and milling 40.95
Cost/ton of CPO 204.75
356 PALM OIL
It can be seen that the cost to produce a ton of CPO is about $204.75. The largest
component in the cost of production of CPO is the cost of FFB at 73% while the
contribution of milling cost is at 27%.
Cost data on production of CPO is not available for most countries, but studies
conducted on Malaysia and Indonesia indicated that the cost of production is lower
than the market price of palm oil. However, cost of production in some countries is
higher than the international market prices of palm oil (2), and for these countries
exporting palm oil would not be viable.
Refining Cost. Crude palm oil can be further processed by refining. Presently,
most of the palm products obtained in the market are processed using physical or
steam refining. Crude or processed palm products may also undergo fractionation
where the solid and the liquid portions are separated. The total cost of refining a ton
of palm oil is about $25.92 while the cost of fractionation is about $5.55 per ton.
3.5. Palm Oil and Palm Kernel Extraction

Fruit Reception. In order to obtain good-quality palm oil, it is essential that the
damage to the fruit is minimal and therefore the handling of the bunches from
the field to the sterilizers must be carried out with the utmost care. At the mill
the FFB is generally discharged from the lorries, or trailers, onto a loading ramp
for the filling of sterilizer cages, which have a nominal capacity of 2.5 tons. A
flow diagram of the palm oil milling process is shown in Figure 6.
Sterilization. Sterilization is carried out by placing the sterilizer cages in hori-
zontal vessels at a steam pressure of 3 kg/cm2 (143 C) and the time under steam
is approximately 60 min. The objectives of sterilization are:
1. Prevention of further rises in the free fatty acid (FFA) of the oil due to
enzymatic reaction.
2. Facilitation of mechanical stripping.
3. Preparation of the pericarp for subsequent processing.
4. Preconditioning of the nuts to minimize kernel breakage.
Stripping. The objective of stripping is the separation of the sterilized fruit from
the bunch stalks. There are two basic actions involved in separating the fruits: (1) a
small vigorous shaking and (2) beating. Although many machines have been
evolved over the years, only the drum type is in general use.
The drum stripper consists of a long horizontal drum made up of small channel
section or T bars spaced far enough apart to permit the escape of the fruit yet close
enough to prevent the passage of the stalks or spikelets. Drum diameters vary from
1.8 to 2 m and lengths from 3 to 5 m and they usually rotate at about 2025 rpm.
The cage is fitted with lifting bars, and as the cage rotates the bunches are lifted up
and then drop back under the action of gravity, and by this action the fruits are
shaken out. As this action is repeated many times over, with the bunches turning
round and round as they pass along the drum, a good stripping is obtained.
Figure 6. Process flow diagram of a conventional palm oil mill.
Digestion. The object of digestion is to reheat the sterilized fruits and to loosen
the pericarp from the nuts and to break the oil cells before passing to the oil extrac-
tion unit. The best digestion conditions are obtained by mixing the fruits at a tem-
perature between 95 and 100 C for approximately 20 min. The digester is generally
a vertically arranged cylindrical vessel fitted with a central shaft carrying a number
of radial arms. Heating may be from a steam jacket or direct steam injection.
Oil Extraction. Oil extraction is generally carried out using continuous screw
presses comprising a perforated horizontal cage of a figure 8 cross section in which
two screws or worms run. A cone at the discharge end of the cage controls the pres-
sure to ensure a minimum of residue oil in the press cake with an acceptable amount
of broken nuts.
There are two products from the press: (1) a mixture of oil, water, and solids, and
(2) a press cake containing fibers and nuts.
Clarification. The crude oil from the press has an average composition of 66%
oil, 24% water, and 10% nonoily solids (NOS). Because of the high proportion of
358 PALM OIL
solids, it has to be diluted with water to obtain satisfactory settling. After dilution,
the crude oil is screened to remove fibrous materials and then pumped to a contin-
uous settling tank where it separates into two parts, i.e., oil and sludge. The top
oil is skimmed off and passed to a centrifugal purifier followed by a vacuum dryer
and finally a cooler before being pumped to the storage tanks. The sludge has an oil
content of approximately 10%, and this is reclaimed and fed back to the main set-
tling tank. The oil fed to the storage tanks has a moisture content between 0.1 and
0.12% and impurities less than 0.02%.
Oil Storage. It is recommended that storage tanks are internally coated with
epoxy materials to prevent iron pickup. To prevent damage by overheating of the
oil, the temperature of the oil during storage and transit is closely controlled
between 32 and 40 C. The unloading or loading temperature is between 50 and
55 C, and for heating to this temperature the maximum rate is 5 C per 24 hr.
Nut and Fiber Separation. When the oil is extracted from the digested fruit, a
cake of nuts and fiber is produced. This is fed, via a breaking conveyor, to a ver-
tical column having an upward airflow at a velocity of 6 m per second. At this velo-
city all the fiber is moved upward or held in suspension, and the nuts drop to the
bottom of the column. The fiber is led to a cyclone for use as a boiler fuel while the
nuts pass to a rotating polishing drum installed at the bottom of the column. This
drum can also be used to remove any large pieces of stalks, stones, or metal that
have gotten into the system.
Nut and Kernel Treatment. This treatment covers four distinct operations: (1)
nut conditioning, (2) nut cracking, (3) kernel and shell separation, and (4) kernel
drying.
Nut Conditioning. Nuts coming directly from the nut fiber separator are still
usually warm, and the kernels are still adhering to the shell. If an attempt is
made to crack them in this condition, many of the kernels will be damaged and
pieces of kernels will still be adhering to the shell. For ideal nut cracking it is neces-
sary to dry the nuts sufficiently to loosen the kernels and then cool the nuts to hard-
en the shell before cracking. This process is usually referred to as nut conditioning.
Nut Cracking. When a nut has been properly conditioned, its shell will crack
cleanly in two or more pieces, and the kernel will be released. Nut cracking
machines are almost invariably of the centrifugal type in which the nuts are given
velocity by being fed through a rotor and are caused to crack by being flung against
a stator ring.
However, the latest development has been the introduction of the ripple mill,
which consists of a balanced rotor of the squirrel cage design and two pieces of
semicircular ripple plates with ripple configurations. The performance of the ripple
mill is determined by the speed and clearance of the rotor, and the rotor provides
the velocity to force the nuts between the stationary ripple plate and the rotor. The
main advantage claimed for the ripple mill is that nut conditioning may not be
necessary.
Kernel and Shell Separation. This is normally achieved in two operations.
First, a winnowing system is used to remove the small pieces of shell and dirt fol-
lowed by hydrocyclones or claybaths.
The action of a hydrocyclone is somewhat similar to that of an air cyclone. By

imparting a circular motion to the fluid by means of the tangential entry, heavy par-
ticles are thrown by centrifugal force to the wall of the cylinder. After tracing a
helical path these particles find their way out through the bottom of the cyclone
while the lighter particles after taking part in an initial downard circular movement
gradually move toward the center of the cylinder and start moving upward leaving
the cyclone via the overflow tube. Although it is mainly the difference in densities
that enables the hydrocyclones to act as a shell and kernel separator, the size and
shape of the particles have some effect, i.e., flow resistance.
The specific gravity of undried kernels is about 1.07 and that of shell about 1.17.
Therefore in a clay and water mixture of specific gravity 1.12 (about 24 Twaddell),
the kernels will float and the shell will sink; this is the principle on which the clay-
bath separator works. Many models were developed from manually operated to
completely automatic versions. As suitable clays were not always readily available,
salt solutions and even dilute molasses were tried for the suspension. The claybath
is quite an efficient separator as long as the density of the suspension is maintained
at the correct value.
Kernel Drying. The moisture content of fresh kernels is about 20%. If bagged
or stored in this condition fresh kernels would soon become moldy. In addition
there would be a rapid increase in the FFA of palm kernel oil (the oil content of
dried kernels can be over 50%). Tests have shown that if the moisture content is
reduced to about 7%, kernels can be safely stored and transported without dete-
rioration due to mold growth.
Bagging and/or Storage of Kernels. After drying, the kernels are usually
bagged, approximately 12 bags to the ton, and stored in sheds awaiting transporta-
tion to the kernel crushing plants. However, with the increase in the costs of bags
and handling, there is a tendency to go over to bulk storage and transportation.
3.6. Mill Construction and Design

Crop Projection. To establish the mill-rated throughput, it is first necessary to have
details of the projected crop over say a 10-year period depending on the planting
program. The general practice is to consider the peak month crop at 12.5% of the
annual crop and that the mill will operate at 20 hr per day for 25 days during the
peak month.
Siting of Mill. The major points to be considered when siting the mill are:
1. The transport costs to bring the FFB to the mill: Ideally the mill should be at
the center of the planted area.
2. The costs to evacuate the produce: This depends on the distance from the mill
to the nearest main road.
3. The distance of the nearest reliable water supply.
4. The ground conditions: Poor ground conditions may involve piling, undulat-
ing areas requiring excavation and/or filling, etc.
360 PALM OIL
Therefore, a compromise has to be made between these factors in deciding the

mill siting.
Mill Design. From the crop projections it is possible to calculate the final capa-
city of the mill and whether phasing of the mill capacities is necessary. The avail-
ability of fuel has to be considered as nowadays there is only just sufficient fuel for
mills operating at over 10 tons of FFB per hour.
A typical layout of a palm oil mill is shown in Figure 7 where the main stations
of the process are also indicated. The equipment selected has to be carried out with
care in order to obtain the correct balance for throughput, steam consumption,
energy demands, and economics of the supply of stand-by equipment. Standardiza-
tion of equipment is an important point to consider when selecting machinery on
the basis of within mill and between mills when applicable. When a set range
and/or make of equipment can be chosen (e.g., valves, gearboxes, electric motors
and starts, etc.), considerable benefits can accrue by way of a reduction in the
amount of spare parts to be carried.
Once the main items of equipment have been selected, it is possible to proceed
with the design of the layout of the mill, design of the buildings, and some initial
work on the layout of the machinery. Final work on the machinery layout will have
to await working drawings from the suppliers, which will not be forthcoming until
after the orders have been placed for the equipment.
During the course of the design for the layout of the mill and its machinery,
thought must be given to the matter of safety and safe working practices.
For projects of this magnitude and complexity it is essential that a critical path
network (CPN) is established.
Mill Construction. Sufficient time should be allocated to carry out the actual
construction work for the mill; otherwise it could result in a host of contractors
and workers descending on the site, which makes good supervision virtually
impossible.
In a normal situation the civil works (foundations, building, etc.) are usually
completed before work commences on the erection of machinery.
The erection of the machines should follow a logical sequence of commencing
from in to out, i.e., threshing station erected first in the main bay, with the press-
ing, kernel, and clarification stations to follow. This sequence can only be imple-
mented if deliveries are planned accordingly. This, of course, depends on the
correct timing for orders, and this is where the CPN is invaluable. In most cases
the erection of the machinery is carried out by contractors. However, there are
many advantages in having the future mill engineer on site during the erection per-
iod. Besides being good training for the engineer for obvious reasons, it should
ensure that a better standard of work is carried out.
3.7. Treatment of Palm Oil Mill Effluent

Sources of Wastewater Production. Large quantities of water are required in the
palm oil milling operations. It is estimated that about 1 ton of water (including boi-
ler feedwater) is required to process 1 ton of fresh fruit bunches. Obviously, a great
361
Figure 7. Typical layout of a palm oil mill.

362 PALM OIL
proportion of the water will be discharged as wastewater, commonly known as palm

oil mill effluent (POME). Some water is lost as vapor (steam).
POME is mainly generated from sterilization and oil clarification processes in
which large quantities of steam/hot water are used. Another waste stream originates
from hydrocyclone operation where water (with clay or salt) is used as a medium to
separate shell and kernel. For a well-operated mill with good housekeeping prac-
tices, the amount of wastewater generated from the sterilization process (sterilizer
condensate), oil clarification process (separator sludge), and hydrocyclone are 0.9,
1.5, and 0.1 ton per ton of oil produced, respectively (25). Thus about 2.5 tons of
POME are generated for every ton of palm oil produced. Obviously water consump-
tion and wastewater production can be minimized through good housekeeping and
process control.
The sterilizer condensate and separator sludge are segregated into separate oil
pits for residual oil recovery before they are mixed again for treatment. The hydro-
cyclone waste contains very little residual oil and is discharged directly into the
treatment plant. The final mixed wastewaters are commonly known as palm oil
mill effluent.
The residual oil recovered from the oil pits are of poor quality. It is drummed and
sold as technical oil for nonedible applications. It is not advisable to recycle or mix
this poor-quality oil with the normal production oil as such practice will negatively
affect the normal production oil quality.
Characteristics of POME. POME, when fresh, is a thick brownish colloidal
slurry of water, oil, and fine suspended solids. It is hot (8090 C) and acidic (pH
45) and contains very high organic matter as indicated by its high biochemical
oxygen demand (BOD3, sample incubated at 30 C for 3 days) (Table 24). In terms
of BOD, POME is 100 times as polluting as domestic sewage. It also contains very
high suspended solids (SS), which are mainly oil-bearing cellulosic materials from
the fruits. The POME is nontoxic as no chemical is added to the oil extraction
process. The characteristics of the sterilizer condensate, separator sludge, and
hydrocyclone wastewater are also shown in Table 24.
TABLE 24. Characteristics of Palm Oil Mill Effluent (25).
Parametera Sterilizer Condensate Separator Sludge Hydrocyclone Water POME
pH 5.0 4.5 300 4.7

Oil and grease 4,000 7,000 300 6,000
BOD 23,000 29,000 5,000 22,000
COD 47,000 64,000 15,000 61,000
SS 5,000 23,000 7,100 13,000
DS 34,000 22,000 100 21,000
TN 500 1,200 100 800
AN 20 40 35
a
All in mg/L except pH: BOD, biochemical oxygen demand; COD, chemical oxygen demand; SS, suspended
solid; DS, dissolved soli; TN, total nitrogen; AN, ammoniacal nitrogen.
TABLE 25. Metal Content in Palm Oil

Mill Effluent (26).
Metal Conc. (mg/L)
P 180
K 2,270
Mg 615
Ca 439
C 25,440
B 7.6
Fe 46.5
Mn 1.98
Cu 0.89
Zn 2.30
In addition, POME also contains high metal content, which can be of importance
for other application like recycling as plant nutrients. Some of the essential metal
contents are given in Table 25.
Treatment Technology for POME. As palm oil milling processes require large
quantities of water, it is common to find palm oil mills located near rivers where
(free) water is readily available. Because of its high organic content (high BOD), if
discharged untreated into the watercourse, POME will soon undergo biological
oxidation that depletes the dissolved oxygen in the water system. Oxygen depletion
destroys aquatic life in the water and consequently the surrounding environment.
Therefore the industry has both the social and ethical obligations to reduce the
environmental impact caused by POME.
There are several options available to reduce the pollution problem created by
palm oil mills. These include complete treatment and disposal of POME or sys-
tematic utilization of POME for beneficial purposes. The choice depends very
much on the local environment.
POME, because of its high organic content (BOD) is easily amenable to biode-
gradation (27). Therefore biological oxidation is the most suitable process to break-
down the organic pollutants in POME. Biological treatment of wastewater is a
process in which a mixed population of microorganisms utilize as nutrients sub-
strates contaminating the water. Wastewater containing polluting substances is
brought into intimate contact with a dense population of microorganisms for a
duration sufficient for the microbes to break down and remove the pollutants to
the desired level.
The oxygen required for the microbial activities is supplied through dissolved
oxygen in water. Invariably the biological treatment system consists of a train of
anaerobic (absence of oxygen), aerobic (presence of oxygen), and/or facultative
(anaerobic and aerobic) processes to meet the required treatment efficiency. The
end products for anaerobic process are biogas [a mixture of methane (6070%),
carbon dioxide (3040%), and a small amount of hydrogen sulfide] and biosolids.
The end products for aerobic process are mainly carbon dioxide and some amount
of biosolids that need to be disposed of separately.
364 PALM OIL
The three most common and effective treatment systems developed for the
POME published so far are shown in Figures 9, 10, and 11. It has been shown
over the last decade that all these systems if operated according to design and main-
tained properly could meet the local discharge standards for POME as shown in
Table 26.
Ponding System. This is by far the most popular treatment system used by the
palm oil mills in Malaysia. It is the most economic system provided suitable land is
available at very low or no cost to the palm oil mill. Large land area is required for a
ponding system. Various designs and configurations of ponding systems are used.
Figure 8 shows a typical system used in Malaysia. It consists of essentially a num-
ber of ponds for different functions. Ponding systems are reliable, stable, and cap-
able of producing good-quality final discharge with a BOD of less than 100 mg/L.
The anerobic ponds are usually 57 m deep each and the facultative ponds are
11.5 m each. The hydraulic retention times (HRT) for the deoiling tank, acidifica-
tion, anaerobic, and facultative ponds are 1,4,45, and 16 days, respectively.
A ponding system is normally operated at low rate, with organic loadings ran-
ging from 0.2 to 0.35 kg per BOD per cubic meter per day. Because of the size and
configuration of the ponds, they are quite difficult to control and monitor. Further-
more, there is very little mixing. Mixing is only achieved through bubbling of the
biogas generated by the anaerobic process. It is hardly adequate to mix the digester
content. Also the rising biogas brings along very fine solids to the top of the ponds.
Figure 8. Schematic flow diagram for ponding system (28).

Figure 9. Schematic flow diagram of opened tank digester and extended aeration system (29).
365
366 PALM OIL
Figure 10. Schematic flow diagram of closed anaerobic digester and land application system (30).
Therefore it is very common to find islands of solids floating in the anaerobic

ponds. This often results in dead spots and short-circuiting in the ponds, which
reduces the treatment efficiency of the system. Obviously it is very labor intensive
and expensive to maintain the ponding system in very satisfactory conditions. It is
also imperative to ensure that as little oil as possible is allowed to get into the anae-
robic pond. Otherwise the oil will agglomerate with the rising solids brought up by
the biogas and form a sticky scum that is difficult to remove. It is not advisable to
allow excessive accumulation of the scum so that the effectiveness of the system is
not adversely affected.
TABLE 26. Discharge Standards for Palm Oil Industrial Effluents for Malaysia.
PORE and OIE Standard

a
Parameter POME A B

Temperature( C) 45 40 40
pH 5.09.0 6.09.0 5.59.0
BOD3 (mg/L) (3 days 30 C) 100 20 50
COD (mg/L)h 250 250
SS (mg/L) 400 50 100
O & G (mg/L) 50
AN (mg/L) 150
TN (mg/L) 200
a
BOD3, biochemical oxygen demand; COD, chemical oxygen demand; SS, suspended solids; O & G, oil and
grease; AN, ammoniacal nitrogen; TN, total nitrogen.
b
Ungazetted. POME, palm oil mill effluent; PORE, palm oil refinery effluent; OIE, oleochemical industrial
effluent. Standard A or B depends on locations.
Due to inadequate mixing by biogas, solid buildup at the bottom of the anaerobic
pond poses another maintenance problem to the palm oil mills. Excessive solid
buildup at the bottom of the ponds will reduce the effective digester capacity and
consequently shorten the hydraulic retention time. Thus the treatment efficiency
will be adversely affected.
Regular desludging of the ponds is recommended. Submersible slurry pumps can
be used for this purpose. The solids are removed at regular intervals to a series of
drying beds constructed besides the ponds. The dried solids, which contain high
plant nutrients, are used as fertilizers in the estates.
Open Tank Digester and Extended Aeration. In this system, after the oil recov-
ery pit, POME is treated in a two-phase anaerobic digestion process followed by
extended aeration in a pond. The digesters are open-top and unstirred. Figure 9
shows a schematic flow diagram of the system. The HRT for the acidification, anae-
robic, and aerobic process are 1, 20, and 20 days, respectively. The organic loading
of the anaerobic digester is in the range of 0.81.0 kg BOD per cubic meter per
day (27).
Similar to the ponding system, limited mixing is provided by the biogas gener-
ated. Hence such a system faces the same problem of solid buildup at the bottom of
the digester. In order to maintain sufficient HRT for effective digestion, the solids
have to be removed at regular intervals as in the ponding system. This can be easily
done by means of slurry pumps. The solids are carted away for land application in
the estates.
As compared to the ponding system, the scum and solid buildup in the digester
can be readily monitored and controlled.
Mechanical surface aerator is used to supply air/oxygen to the aerobic pond. In
aerobic systems it is important to ensure that enough oxygen is supplied to the aero-
bic microorganisms to do the job.
Close Tank Digester with Biogas Recovery and Land Application. The digesters
(see Figure 10) are operated as conventional high rate systems with organic loading
of 4.8 kg volatile solids (VS) per cubic meter per day. The HRT for anaerobic diges-
tion is about 10 days operating at a slightly elevated temperature of 4250 C. Good
mixing is ensured by recycling the compressed biogas through an emitter and a
draught tube. From the emitter, the biogas rises through the draught tube in large
bubbles. Thus the digester liquor is drawn into the bottom of the draught tube and
discharged from the top, causing effective circulation and hence mixing of the
digester content. As the content of the digester is well mixed, there is no problem
of solid buildup at the bottom of the digester.
In this system, the digester liquor, having a BOD of about 2000 mg/L, is applied
to the plantation nearby as fertilizer (31). Several systems have been developed for
land application of POME. The biogas generated can be harnessed for heat and
electricity generation. Excess biogas is flared off. About 0.59 m3 of biogas is pro-
duced per kilogram of VS added. Thus for a 60-ton FFB per hour mill operating 20
hr per day, about 20,000 m3 of biogas is obtainable. The biogas contains about 65%
methane, 35% carbon dioxide, and less than 2000 ppm of hydrogen sulfide. It has a
calorific value of about 5300 kcal/m3. It can be used as an energy source to supply
368 PALM OIL
heat or electricity to supplement the energy requirement if necessary (31,32).

Substantial saving in energy bill can be realized.
3.8. Oil Palm By-products

Apart from the production of CPO and palm kernel oil (PKO), the oil palm industry
also generates large quantities of by-products in the form of biomass. The bulk of
the by-products derived from the palm oil industry are basically lignocellulosic and
organic in nature and with a high plant nutrient content. With proper handling and
management, these by-products could be utilized and converted into value-added
products.
Biomass Production and Availability. The biomass production from the palm oil
industry is derived mainly from two sources, i.e., the plantations and the palm oil
mills. From the plantation the biomass produced per hectare of oil palm has been
estimated and shown in Figure 11. On an annual basis, about 0.4 tons of palm
Figure 11. Diagrammatic representations of actual quantity of oil palm wastes (33).
fronds per hectare are available through routine maintenance pruning and harvest-
ing. However, at the end of its economic life span (about 25 years), another
14.4 tons of palm fronds per hectare are available during replanting. In addition a
total of 74.4 tons of palm trunks per hectare are also available during the replanting
program.
At the palm oil mills, about 20% CPO and 1.6% of PKO are recovered from the
FFB thus leaving about 78.4% biomass, including palm kernel meal.
As shown in Figure 11 the processing of fresh fruit bunches generated about
1.5 tons of empty fruit bunches (EFB), 1.6 tons of palm press fibers, 0.9 ton of
palm kernel shell, 2.4 tons sterilizer condensate, and 0.7 ton of dry mill effluent
per hectare of oil palm annually.
Utilization of By-product. Oil Palm Trunks and Fronds. Under normal plan-
tation practices, the pruned fronds are placed along the palm interrows and act as
mulch. Besides conserving soil moisture and reducing soil surface erosion, the
fronds on decomposition return organic matter and slowly release plant nutrients
to the soil (Table 27).
During the replanting program, the oil palm trunks and fronds are chipped into
small pieces and pulverized using a special pulverizer. The biomass residues are left
in the field to allow for decomposition processing, which could then yield organic
matter and release of plant nutrients. The placement of trunk residues on field
terraces could also reduce soil erosion.
TABLE 27. Content of N, P, K, Mg, and Ca Obtained from Wastes of a Hectare

of Oil Palm (33).
Weight in kg ha1
Tissues and Dry Weight N P K Mg Ca
Felled palm trunk

75.460 kg ha1 368.2 35.5 527.4 88.3 146.4
Palm fronds
(a) At felling: Pinnae 114.0 7.5 109.4 8.4 7.1
14.467 kg ha1 Rachis 36.1 6.4 84.5 15.6 28.6
150.1 13.9 193.9 24.0 35.7
(b) Annual pruning Pinnae 819 5.4 78.7 6.0 5.1
10,400 kg ha1 Rachis 26.0 4.6 60.7 11.2 20.5
107.9 10.0 139.4 17.2 25.6
Empty bunches
1.546 kg ha1 5.4 0.4 35.3 2.7 2.3
Fiber
1,626 kg ha1 5.2 1.3 7.6 2.0 1.8
Shell
938 kg ha1 3.0 0.1 0.8 0.2 0.2
Effluent
13,604 kg ha1 Raw 12.9 2.1 26.6 4.7 5.4
Digested 4.4 0.9 20.7 3.9 3.1
370 PALM OIL
Empty Fruit Bunches and Fibers and Shells. Traditionally, the empty fruit
bunches generated at the palm oil mill are mostly incinerated to produce bunch
ash. Bunch ash is considered a good source of potassic fertilizer and is also useful
as liming materials because of its high alkalinity (pH 12). However, incineration of
EFB could cause air pollution, and this practice is not encouraged by the Depart-
ment of Environment.
Alternative uses of EFB have been investigated and results have shown that they
are suitable as a mulching material for oil palm. The EFB are fibrous in nature and
have a high moisture content (about 60%). The application of EFB in the interrows
of palm avenues has been shown to improve oil palm growth and yield perfor-
mance. When applied onto the soil surface, the EFB undergoes a degradation pro-
cess that will yield organic matter and slowly release plant nutrients for crop uptake.
Some of the oil palm fibers and shells are usually used as a surface mulching
material for oil palm seedlings at the nursery stage. These materials are beneficial
in conserving soil moisture and reducing fertilizer leaching in the polybags, and
thus they enhance the growth of the seedlings.
Palm Oil Mill Effluent (POME). Palm oil mill effluent is essentially organic in
nature and nontoxic but has a high polluting potential. In its raw state, POME has
an extremely high concentration of biochemical and chemical oxygen demand
(BOD and COD) and high in plant nutrient contents, particularly in nitrogen and
potassium. After treatment processes various types of POME are available, and
their chemical composition are shown in Table 28.
POME has proven to be a good source of organic fertilizer and is available in
large volume (Table 29). Applied at rates corresponding to the nutrient requirement
of crops, it will not have detrimental impact on the environment. The beneficial
effects of POME application on crop yield and performance have been investigated.
Several methods of land application systems are available.
Energy Potential from Oil Palm By-products. Apart from crude palm oil and
palm kernel, a palm oil mill also produces a large quantity of biomass as by-
products. In general, an FFB contains about 20% palm oil, 67% palm kernel,
TABLE 28. Types of POME Available and Their Chemical Compositions (3436).
Chemical Composition (mg/L.)

Total
Type of POME BOD N P K Mg
Raw 25,000 948 154 1,958 345

Digested (anaerobic)
Stirred tank 1,300 900 120 1,800 300
Supernatant 450 450 70 1,200 280
Supernatant 10% slurry 191 320 42 1,495 258
Bottom slurry 1,0003,000 3,552 1,180 2,387 1,509
Digested (aerobic)
Supernatant 100 52 12 2,300 539
Bottom slurry 150300 1,495 461 2,378 1,004
REFINING AND FRACTIONATION 371
TABLE 29. Annual Production of Raw Effluents for Mills with Capacity Ranging
from 10 to 60 tons FFB Per hour.
Mill Capacity (Ton FFB/Hr)

Annual Rate (ton) 10 20 30 40 60
Total FFB processed (capacity

16 hr 300 days) 48,000 96,000 144,000 192,000 288,000
Effluent production (FFB 0.67%) 32,160 64,320 96,480 128,640 192,960
TABLE 30. Heat Value of Biomass (37).
Biomass Moisture Content (%) Oil Content (%) Heat Value kcal/kg (dry)
Empty fruit bunches 65 5 3700

Fiber 42 5 4420
Shell 7 1 4950
1415% fiber, 67% shell, and 23% EFB. The heat value of each biomass is shown
in Table 30.
Fiber and Shell. The palm oil mill uses fiber and shell as boiler fuel to produce
steam for electricity generation and palm oil and kernel production processes. The
fiber and shell alone can supply more than enough electricity to meet the energy
demand of a palm oil mill. It is estimated that about 20 kWh (lower for higher-
capacity mill) of electrical energy is required to process 1 ton of FFB.
Empty Fruit Bunch. Apart from fiber and shell, EFB is another biomass that
can be readily converted into energy. However, this material has only been utilized
to a limited extent. This is because there is already enough energy available from
fiber and shell. Also due to its physical nature and high moisture content (5065%),
the EFB has to be pretreated to reduce its bulkiness and moisture content to below
50% in order to render it more easily combustible (37, 38).
Biogas from Palm Oil Mill Effluent. Biogas is generated from anaerobic treat-
ment of POME. It contains about 65% methane (CH4), 35% carbon dioxide (CO2),
and trace amounts of hydrogen sulfide (H2S). It has a calorific value of about 4740
6150 kcal/N m3. About 28 m3 of biogas are generated for every cubic meter of
POME digested. In a gas engine, about 1.8 kWh of electricity can be generated
from every cubic meter of biogas.
4. REFINING AND FRACTIONATION
4.1. Physical and Chemical Refining

Crude palm oil extracted commercially from the fresh fruit bunches contains a
small but variable amount of undesirable components and impurities. These include
some mesocarp fibers, moisture and insolubles, free fatty acids, phos-pholipids,
372 PALM OIL
trace metals, oxidation products, and odoriferous substances. As a result, palm oil is
normally refined to a bland, stable product before it is used for direct consumption
or for formulation of edible product. In Africa, however, crude palm oil is often
consumed in the crude form.
Two methods, namely physical refining and chemical refining, are available for
refining crude palm oil. They differ basically in the manner in which the free fatty
acids are removed. Physical refining has become the major processing route
because of its cost effectiveness, efficiency, and simple effluent treatment (39).
Both processes are able to produce refined, bleached, and deodorized (RBD)
palm oil of desirable quality and stability suitable for edible purposes (40). The
unit operations involved in these two processes and the components removed are
shown in Figure 12 and Table 31, respectively.
Physical Refining. Physical refining was introduced to palm oil processing in
1973 (41). Its unique feature is that the deacidification, deodorization, and thermal
decomposition of carotenoids are accomplished in one process in a stainless steel
deodorizer. It is a continuous processing consisting of a two-step operation of
pretreatment followed by steam distillation (42).
Pretreatment. Pretreatment refers to the initial degumming of crude palm oil
with concentrated phosphoric acid and the subsequent adsorptive cleansing with
bleaching clay. Crude palm oil is dosed with phosphoric acid (8085% concentra-
tion) at a rate of 0.050.2% (of the feed oil), heated to 90110 C, and given a resi-
dence time of 1530 min before passing to the bleacher where bleaching earth is
added as a slurry. The earth required ranges from 0.8 to 2.0%, depending on the
quality of the crude oil.
The purpose of the phosphoric acid is to precipitate the nonhydratable phospha-
tides while the function of the earth is fourfold: (1) to adsorb the undesirable impu-
rities such as trace metals, moisture, insolubles, and part of the carotenoids and
other pigments (43). (2) to reduce the oxidation products, (3) to adsorb the phos-
pholipids precipitated by the phosphoric acid, and (4) to remove any excess phos-
phoric acid present in the oil after degumming. The final residual color of the
pretreated oil alone is unimportant as the role of the bleaching earth is not so
much of color removal but more critically in its ability to act as an adsorptive
cleansing agent (44). Complete removal of residual phosphoric acid in the bleach-
ing stage is also critical as any slip through can result in the rapid rise of free
fatty acid content and color of the final RBD oil (39,42,45). As a further assurance,
a suitable quantity of calcium carbonate is often added after dosing of the bleaching
earth to the degummed oil, to help neutralize the residual phosphoric acid (46).
Bleaching is carried out under a vacuum of 2025 mmHg and at a temperature
of 95110 C with retention time of 3045 min (47). The slurry containing the oil
and earth is then filtered to recover a clear, light orange color pretreated oil. Usually
a small amount of diatomaceous earth is used to precoat the filter leaves to improve
the filtration process. As a quality precaution, the filtered oil is polished through
another security filter bag in series, to trap any earth particles that escape through
the first filter. This is essential as the presence of spent earth particles in the
pretreated oil reduces the oxidative stability of the final RBD oil (46). The spent
Figure 12. Flow diagrams of (A) physical refining and (B) chemical refining of crude palm oil.
373
374 PALM OIL
TABLE 31. Refining Crude Palm Oil: Unit Processes.
Stage Principal Impurities Reduced or Removed
Degumming Phospholipids, trace metals, pigments

Neutralization Fatty acids, phospholipids, pigments, oil insolubles, water solubles
Washing Soap
Drying Water
Bleaching Pigments, oxidation products, trace metal, traces of soap
Filtration Spent bleaching earth
Deodorization Fatty acids, mono-and diglycerides, oxidation products, pigment
decomposition products
Physical Refining Fatty acids, mono- and diglycerides, oxidation products, pigment
decomposition products
Polishing Removal of trace oil insolubles
earth from the filter normally contains about 2040% oil, and this is the major
source of oil loss in the refining process.
The pretreatment process can be carried out in batch, semicontinuous, or contin-
uous equipment, and the filters used are either plate and frame presses or verticle or
horizontal pressure filters with verticle stainless steel filter screens.
Deodorization. The pretreated oil is then ready for deacidification and deodor-
ization. The pretreated oil is first deaerated followed by heating to 240270 C in an
external heat exchanger before pumping into the deodorizer, which is kept under a
vacuum of 25 mm Hg. Traditionally thermal fluids are commonly used as the heat-
ing medium. However, to eliminate the risk of possible contamination of refined oil
with thermal fluid, superheated high-pressure steam is now commonly being used,
especially in new plants. Temperatures above 270 C are to be avoided to minimize
loss of neutral oil, tocopherols/tocotrienols, and also the possibilities of isomeriza-
tion and undesirable thermochemical reactions (48). Under such conditions and
with the help of stripping steam, the free fatty acids, which were still present in
the pretreated oil, are distilled together with the more volatile odoriferous and
oxidation products such as aldehydes and ketones, which otherwise would impart
undesirable odor and taste to the oil. At the same time, the residual carotenoids
present are also thermally decomposed (Figure 13), and the end result is the
production of a light-colored, bland RBD palm oil. To maximize the recovery of
thermal energy, the hot deodorized oil is heat exchanged against incoming pre-
treated oil to be cooled down to a temperature of 120150 C. Further cooling is
effected by water down to 5565 C prior to storage. Antioxidant and citric acid,
if required, are dosed into the RBD palm oil at this stage.
The desirable qualities of the pretreated and RBD palm oil are given in Table 32
(46.50).
Development in Palm Oil Deodorization. The main operation in the deodor-
izer is the stripping of volatile materials and thermal action due to the combined
effects of superheated steam, high temperature, and efficient vacuum. The older
deodorizers (prior to 1985) use bubble caps or sieve-tray designs to effect the
Figure 13. Thermal destruction of b-carotene (49).
countercurrent mixing action between the stripping steam and oil flow. However,
technological innovations have resulted in many plants changing over to new deo-
dorizers of packed column or the falling film types (5153). These new features
reduce pressure drop and improve the contact between the oil film and stripping
steam, thus enhancing mass-transfer efficiency. Incorporation of this new design
has resulted in deodorizers of larger capacities, faster throughput, lower loss of
neutral oil and thus lower steam consumption on a per ton basis (54).
Chemical Refining. Also called caustic refining, chemical refining involves three
stages: (1) gum conditioning and neutralization, (2) bleaching and filtration, and
(3) deodorization.
TABLE 32. Desirable Quality of Pretreated and RBD Palm Oil from the Factory.
Pretreat Palm Oil Refined Bleached

Parameter (Degummed/Bleached) and Deodorized Palm Oil
Free fatty acids (as C16:0) (%) Same as crude feed 0.10% max.
Peroxide value (mEq/kg) 0 0
Moisture and impurities (wt %), max. 0.1 0.1
Iron (mg/kg), max. 0.12 0.12
Copper (mg/kg), max. 0.05 0.05
Phosphorus, (mg/kg), max. 4 4
Lovibond color, max. (5 14-inch cell) 3.0R
376 PALM OIL
Gum Conditioning and Neutralization. The crude oil is heated to a tempera-

ture of 8090 C. Phosphoric acid of 8085% concentration is then dosed in at a
rate of 0.050.2% (of the feed oil). This serves to precipitate the phospholipids.
After this, the degummed oil is further treated with a caustic soda solution of about
4 N (or 20 Be) concentration with a calculated excess (based on free fatty acid content
of the crude oil) of about 20%. The reaction between caustic soda and the free fatty
acids in the degummed oil results in the formation of sodium soap, which is readily
removed by a centrifugal separator. The lighter phase discharged consists mainly of
neutralized oil containing 5001000 mg/kg of soap and moisture while the heavy
phase is mainly soap, insoluble impurities, gums, phosphatides, excess alkali, and a
small quantity of oil loss through emulsification. As an excess of alkali is used, it is
unavoidable that a slight loss of neutral oil through saponification also occurs.
The neutralized palm oil (NPO) is then washed with 1020% hot water to
remove traces of soap still present. After another stage of centrifugal separation,
the washed oil is then dried under vacuum to a moisture level below 0.05%.
Bleaching and Filtration. The neutralized palm oil is treated with bleaching
earth in a similar manner as that described in physical refining. However, in this
case, the earth also removes traces of soap that are present.
Deodorization. The neutralized and bleached oil is then channeled to the deo-
dorizer in a similar manner to that in the physical refinery. The oil is subjected to
distillation at a temperature of 240260 C and a vacuum of 25 mm Hg with direct
steam injection. Under such conditions, residual free fatty acids, volatile oxidation
products, and odoriferous materials are removed together with thermal decomposi-
tion of carotenoids (Figure 14). The final product, called neutralized, bleached, and
deodorized (NBD) palm oil is then cooled down to 60 C and passed through pol-
ishing filter bags before pumping to the storage tanks. The desirable quality char-
acteristics of intermediate and final products are given in Table 33.
Refining Factor. The efficiency of the refining process is estimated by the use
of a refining factor (RF):
% Total oil loss

RF
% FFA incrude oil
TABLE 33. Desirable Quality of Freshly Produced Intermediate and Final Products
in Alkaline Refining of Palm Oil (48,50).
Neutralized,
Neutralized Neutralized and Bleached, and
Parameter Palm Oil Bleached Palm Oil Deodorized Palm Oil
Free fatty acids (as C16:0) (%) max. 0.15 0.15 0.10
Peroxide values (mEq/kg) 0 0
Moisture and impurities (wt %) max. 0.1 0.1 0.1
Iron (mg/kg) 0.15 0.12
Copper (mg/kg) 0.06 0.05
Phosphorus (mg/kg) 4 4
Soap content (mg/kg) 20 0 0
TABLE 34. Desirable Quality of RBD/NBD Palm

Kernel Oil.
Parameter Value
Free fatty acid (as % C12:0), max. 0.05

Moisture and impurities (wt %), max. 0.1
Lovibond color (5 14-inch cell), max. 1.5R
Iron (mg/kg), max. 0.5
Copper (mg/kg), max. 0.1
Phosphorus (mg/kg), max. 1.0
Both figures are corrected for moisture and insoluble impurities in the crude oil. In
alkaline refining, the term acid oil factor (AOF) is sometimes used:
%Acid oil
AOF
%FFA
The AOF is used for monitoring losses of neutral oil in the neutralization
process.
Values of RF for chemical refining range from 1.5 to 2.0, while lower figures of
1.21.4 are usually recorded in physical refining.
Refining of Other Palm Products. Beside crude palm oil, crude palm olein,
crude palm stearin, crude kernel oil, crude palm kernel olein, and crude palm kernel
stearin can also be refined by either chemical or physical processes described
before. The basic unit operations and processing conditions for crude palm olein
and stearin are similar to those of palm oil. However, in refining palm kernel pro-
ducts, due to the virtual absence of carotenoids, the earth dosage required in the
bleaching stage is lower, usually less than 1.0%. Furthermore, due to the presence
of shorter chain (C8C14) fatty acids, the deodorization temperature required is
about 230250 C. Typical achievable quality of RBD/NBD palm kernel oil is given
in Table 34.
By-products. Chemical Refining. The neutralization of free fatty acid in the
crude palm oil with caustic soda results in the formation of soapstock, which is trea-
ted with dilute sulfuric acid of pH 2.03.5 at 110130 C for 30 min. A by-product
called palm acid oil is then separated from the aqueous phase by centrifugation
followed by hot-water washing. It consists mainly of free fatty acids, neutral oil,
and partial glycerides. A small amount of unsaponifiable matter is also present.
Characteristics and properties of palm acid oil (derived from chemical refining of
crude palm oil, stearin, and olein) are given in Table 35 (55).
Physical Retining. The by-product in the physical refining of crude palm oil is
the palm fatty acid distillate (PFAD). It is obtained as a condensate of the volatile
matters carried over from the deodorizer by the action of the stripping steam. It
consists of 8090% of free fatty acid. It has often been used as a raw material
for soap making, feed compounding, and oleochemical feedstock. An important
and valuable constituent of PFAD is vitamin E in the form of tocopherols and
378 PALM OIL
TABLE 35. Characteristics of Malaysian Palm Acid Oil.
Parameter Range Mean
Free fatty acid (% as C16) 66.988.7 72.8

Moisture content (wt %) 0.100.68 0.28
Iodine value (Wijs) 41.864.4 53.2
Titer ( C) 38.847.1 44.6
Unsaponifiable matter (wt %) 0.401.95 0.79
Fatty acid composition (wt %)
C12:0 0.10.5 0.1
C14:0 1.01.6 1.2
C16:0 31.856.0 47.1
C18:0 4.15.2 4.6
C18:1 29.948.9 36.3
C18:2 6.312.0 9.6
C18:3 0.30.8 0.7
TABLE 36. Characteristics of Palm Fatty Distillates from the Physical Refining of Palm
Oil Products (55).
PFAD from PFAD from PFAD from

Palm Oil Palm Olein Palm Stearin
Parameter Range Mean Range Mean Range Mean
Free fatty acid (% as C16:0) 72.389.4 83.3 71.898.6 85.5 77.789.5 85.9
Moisture content (wt %) 0.030.15 0.08 0.030.12 0.07 0.040.16 0.09
Iodine value (Wijs) 51.257.4 55.3 45.659.1 57.4 44.352.6 44.8
Titer ( C) 40.749.0 46.3 36.947.8 45.5 44.352.6 49.8
Unsaponifiable matter (wt %) 1.53.4 2.5 1.63.7 2.3 1.32.5 1.9
Fatty acid composition (wt %)
C12:0 0.10.3 0.2 0.010.6 0.2 0.10.3 0.1
C12:0 0.91.5 1.2 0.31.5 1.2 1.21.6 1.3
C16:0 42.951.0 47.1 39.149.0 44.1 47.661.3 57.0
C18:0 4.14.9 4.5 3.85.1 4.5 4.25.4 5.1
C18:1 32.839.8 36.6 29.342.6 39.0 25.236.3 29.0
C18:2 8.611.3 9.6 7.112.8 10.2 6.07.4 6.8
C18:3 0.20.6 0.5 0.30.9 0.5 0.20.5 0.4
tocotrienols (56). A process for the economical recovery of vitamin E from PFAD
has been developed by the Palm Oil Research Institute of Malaysia (PORIM) (57).
Characteristics and properties of PFAD from the physical refining of palm oil are
given in Table 36.
4.2. Fractionation
The triglycerides of palm oil consist of a combination of fatty acids with different
chain length as well as degrees of unsaturation. This results in the presence of sub-
stantial quantity of both low- and high-melting triglycerides. Crystallization of the
oil under controlled cooling followed by separation will yield a low-melting liquid
phase (olein) and a high-melting solid phase (stearin). Factors affecting the crystal-
lization process are oil composition, polymorphism, and cooling condition (58,59).
Oil Composition. Palm oil contains about 48% of diglycerides (50), which can
form a eutectic mixture with the triglycerides resulting in lower solid content. This
can slow down the rate of crystallization. The monoglycerides, present at less than
1% in palm oil, have no significant effect on the crystallization.
Polymorphism. Palm oil triglycerides are polymorphic and thus can crystallize
in several forms. The polymorphic forms are designated as a, b0 , and b in the order
of increasing stability and melting points. Upon cooling, palm oil initially crystal-
lizes in the a form, which gradually transforms in the order of a!b0 !b form. To
have good separation, it is desirable to have b0 -form crystals because b0 crystals
agglomerate into large, firm clusters resulting in good subsequent filtration.
Cooling Rate. Cooling rate affects the nucleation and crystal growth of the oil.
As the oil is cooled, it becomes supersaturated. When the temperature is sufficiently
low (about 3236 C), saturated glycerides will crystallize, and these crystals act as
nuclei for further crystallization of the lower melting glycerides, resulting in forma-
tion of larger clusters of crystals. Slow cooling rate and proper stirring speed is
essential for the formation of the desired crystal form.
Process Description. There are three commercial methods for fractionating
palm oil: dry, detergent, and solvent process.
Dry Fractionation. This is usually carried out semicontinuously using neutra-
lized, neutralized and bleached, or fully refined palm oil. It does not require the
use of any chemicals or additives. The oil is kept homogenized at about 70 C to
destroy any presence of crystal in order to induce crystallization in a controlled
manner during subsequent cooling. Crystal formation and growth occur as the oil
is agitated and cooled using chilled-water circulation. The cooling program is con-
trolled by setting the temperature differential between the oil and chilled water, and
also the time of cooling. When the temperature reaches the desired temperature,
which is dependent on the quality of olein required (but usually about 20 C), the
cooling is stopped and the thick partially crystallized mass is ready for filtration.
The different filtration systems now used in the industry are drum rotary filters,
stainless steel belt florentine filters, and membrane filters. Over the last decade,
the membrane filter, which actually is a filter press equipped with a membrane
plate, is increasingly used because it gives a higher yield of olein (about 70
75%) and a harder stearin compared to that of about 65% obtainable from florentine
or rotary drum filters (60).
Detergent Fractionation. Also known as the Lanza or Liprofrac process, deter-
gent fractionation is normally carried out on crude palm oil. The oil is first cooled in
the crystallizer with chilled water to allow the crystallization of the higher melting
glycerides. When the desired temperature is reached (usually about 20 C), the crys-
tallized mass is mixed with an aqueous detergent solution containing about 0.5% of
sodium lauryl sulfate and magnesium sulfate as an electrolyte. The stearin crystals
are wetted by the detergent solution and separate out into a suspension in the aqu-
eous phase. On centrifuging, the olein is discharged as the lighter phase, and the
380 PALM OIL
stearin forms part of the aqueous phase. The olein phase is then washed with hot
water to remove excess detergent and vacuum dried before storage. The aqueous
phase is heated to 95100 C to break the emulsion for recovering the stearin, which
is again washed with hot water and dried under vacuum before storage. Yield of
olein is about 80%.
Solvent Fractionation. This process is the most expensive because of solvent
loss, solvent recovery equipment, much lower temperature requirement, and strin-
gent safety features. The process involves the use of solvents such as hexane or
acetone. The oil is first dissolved in the solvent followed by cooling to the desired
temperatures to obtain the desired crystals. Cooling is effected by brine if very low
temperature is required. The miscella containing the partially crystallized oil and
solvent is then filtered under vacuum suction in an enclosed drum filter. The olein
miscella and stearin miscella are then separately distilled to remove the solvent and
recover the fractions. Yield of olein is about 80%. The solvent process nowadays is
only viable in the production of high value products such as cocoa butter equivalent
or other specialty fats.
Double Fractionation. Double fractionation is carried out for the production of
palm olein with higher iodine value of above 60 or for the production of palm-
midfraction (PMF), which contains a high proportion of oleodipalmitin used for
production of palm-based cocoa butter equivalent (60, 61). Usually the first olein
obtained is recycled back to the plant for further cooling, crystallization, and
filtration. The second stearin otained is termed palmmidfraction. Special and skill-
ful control of the crystallization of both stages is critical in achieving the desired
quality of the products.
Fractionation of Palm Kernel Oil. As in palm oil, palm kernel oil can also be
fractionated via the dry, detergent, and solvent processes (62). The principles
applied are quite similar. The conditions of operation, however, are quite different
because of the different triglyceride composition and crystallization behavior of
palm kernel oil. In the dry fraction process, the separation of palm kernel olein
from the palm kernel stearin is effected by hydraulic pressing under high pressure.
In this case, the palm kernel stearin, which is an important material for production
of lauric-basic cocoa butter substitute, is the premium product. Its yield ranges from
25 to 40% depending on the process used.
4.3. Quality Assurance

General. Palm oil is one of the most stable vegetable oils, and this can be attributed
to the presence of natural antioxidants, and also to the balanced ratio of saturated to
unsaturated fatty acids. Nevertheless palm oil, whether crude or refined, is still sus-
ceptible to quality deteriorations. Stringent preventive measures are necessary to
ensure the production of refined palm oil products of superior quality and accept-
ability.
Quality Chain. The processing chain of palm oil begins with the harvesting of
fresh fruit bunches (FFB) from the estates followed by processing of the FFB into
crude palm oil. Thereafter, the crude palm oil is sent to the refinery for processing
into various grades of refined products that are then transported to the bulking
installations for export. Because any quality problem that may arise at any point
in the chain will affect the other stages down the line, it is necessary that the right
quality be attained right from the beginning of the process chain. The two main
quality problems associated with palm oil are hydrolysis, leading to formation of
fatty acids, and oxidation, leading to rancidity.
Hydrolysis. The hydrolysis of palm oil is promoted by the presence of free
moisture and heat and also by lipolytic enzymes endogenous to the plant tissue
(63,64). This mode of deterioration occurs during the bruising of fruits in the har-
vesting and transportation of the FFB to the mill and also their extended storage
under unfavorable conditions. In the case of oil during storage, the hydrolysis is
attributed to a chemical reaction that is autocatalytic (65). The presence of high
FFA in crude palm oil is undesirable as it (1) reduces the yield of RBD palm oil
through higher loss of PFAD by-product, (2) reduces the capacity of refining,
and (3) results in poor bleachability of crude oil and poor stability in the final pro-
duct (50). Crude palm oil with high FFA content invariably also contains a high
amount of partial glycerides, especially diglycerides. Interactions between the
diglycerides and the triglycerides often lead to formation of eutectics resulting in
poor crystal formation during fractionation, difficulty in separation of olein and
stearin by filtration, and also in production of olein with poor cold stability (66, 67).
Oxidation. Oxidation of oils and fats is due to prolonged exposure to air. By
virtue of the low polyunsaturated fatty acid content, palm oil is relatively more
stable to oxidative deterioration than the polyunsaturated vegetable oils. However,
in the presence of trace metals such as iron and copper, excessive oxidation at the
olefin bonds of the oleic and linoleic acids can occur, resulting in rancidity. Highly
oxidized crude palm oil is known to have poor bleachability and thus requires more
bleaching earth and more severe refining conditions, and the final product will
likely be of poor stability (44, 45, 68).
Quality Assurance Measures in Plantation and Milling. Availability of good-
quality crude palm oil is a prerequisite for the production of good-quality refined
palm oil products. The criteria for good-quality crude palm oil are:
Low free fatty acid content

Low in oxidation characteristics
Good bleachability
Low in trace metals and insoluble impurities
Moisture content of about 0.150.20%
High in deterioration of bleachability index (DOBI)
[DOBI, which is defined as the ratio of the uncorrected absorbance values at

446 nm to that at 269 nm, was introduced as a quality parameter to differentiate
the refinability of good- and poor-quality crude palm oil. The relation to quality
is DOBI >3, good; 2.42.9, fair, and <2.3, poor (45, 69).]
382 PALM OIL
TABLE 37. Crude Palm Oil Quality.
Component Ripe, Fresh, Unbruised Fruit Average Traded PO
Triglycerides (%) 98 <98

Diglycerides (%) 23 48
Monoglycerides (%) 0.1 0.2
FFA (%, as C16:0) 0.1 3.5 (max 5)
Phosphorus (ppm) 23 2030
Tocopherols (ppm) 800 600800
Carotene (ppm) 550 550
Totox 1 >5
Iron (mg/kg) 0.10.3 510
Copper (mg/kg) 0.01 0.05
In the palm oil industry, it is often said that good quality is made in the field,
not in the mill. This statement clearly emphasizes the importance of maintaining
good harvesting practices of fruit bunches in the plantation. A good harvesting
practice is one that gives the best compromise between oil yield, oil quality, and
harvesting cost. Field factors that determine quality of the palm oil include the
degree of ripeness of the fruit bunches, the severity of bruising of the harvested
crop, delays between harvesting and sterilization, and contamination of FFB by
sand, dirt, or stones (70). Data in Table 37 serve to indicate that the oil extracted
from fresh unbruised fruits can have very low FFA and oxidative characteristics
compared with that normally traded (50). Precautionary measures taken by mills
to minimize hydrolysis, oxidation, and contamination of the crude palm oil are
summarized below:
FFB handling: minimize bruising and sterilize as soon as possible ( 24 hr)

Sterilization: optimize conditions, avoid overheating, do not mix boiler
condensate with crude palm oil
Clarification: eliminate water and impurities; use hermetic system
Drying: reduce moisture to 0.170.2% before storage
Processing of Crude Palm Oil. The ultimate aim of the processing of crude palm
oil is to obtain various products such as RBD palm oil, RBD palm olein, or RBD
palm stearin that meet the requirement of the end users. An effective and efficient
quality assurance program in a processing plant is essential and should consist of
the following monitoring activities:
Raw material: Each and every delivery must be carefully inspected to ensure that
specifications are met and that the shipment is free from contamination. Good
raw material is a prerequisite to good-quality product.
In-process materials: Regular analyses of important quality parameters will
serve as a check that proper processing has been achieved. Good commu-
nication between laboratory and production personnel is of great importance

to ensure success of the quality assurance program.
Finished products: Regular checks on the finished products must be carried out
to ensure compliance to quality specifications before the products are
permitted for discharge to storage or shipment.
Process control: Proper processing conditions (dosage of processing aids,
temperature, pressure/vacuum, flow rate, etc.) must be closely adhered to
and monitored to ensure the oil is processed correctly and to minimize
undesirable side reactions. In order to assess process efficiency, the oil losses
through spent clay or soapstock (as in the case of alkaline refining only) must
also be monitored.
4.4. Palm Oil Refinery Effluent Treatment

The characteristics of palm oil refinery effluent vary according to the type of refin-
ery operation (chemical or physical refining, fractionation process, etc.), process
control, and housekeeping program. It is quite difficult to derive general character-
istics for raw effluent. Therefore the choice of treatment system will depend very
much on the complexity of the raw effluent, i.e., its flow and characteristics.
However, there has been a trend among refiners over the last two decades to
reduce effluent and other forms of pollution by:
1. Changing from chemical refining to physical refining

2. Automation and strict process control to prevent spillage and product loss
3. Installation of new equipment that is based on low energy and water
consumption
In Europe and the United States, there are two basic ways of providing effluent
treatment facilities for the edible oil refining industry. One way is for the industry to
construct a treatment system at the manufacturing plant site to treat its effluent to an
acceptable level for discharge directly to rivers or other public water courses. The
second is to dicharge the untreated or partially treated effluent to sewers of a local
government agency providing wastewater conveyance and treatment services or
publicly owned treatment works (7173). The latter practice is generally termed
joint treatment. The industry practicing joint treatment is required to provide
control and pretreatment to various degrees in order to use the publicly owned
facilities.
The main reason for the joint treatment practice is that it costs less than the alter-
native. The other treatment costs are fairly shared among the users and beneficiaries
of the system. Another important advantage is minimizing the space necessary for
the treatment facilities. This saving of space is particularly important for crowded
industrial estates. Obviously, the joint treatment plant is expected to operate to
achieve better efficiency because its operation is carried out by specialized, full-
time, and well-trained professionals. Understandably, in any commercial operation,
384 PALM OIL
an effluent treatment plant is way down on the priority list. The industry is quite
reluctant to spend money on pollution control equipment. Therefore joint treatment
is the most economical and practical choice unless such facility is not available.
Treatment Method. Oils and Fats Recovery. The oils and fats recovery system
adopted obviously depends on the local circumstances. Typically, the first stage of
pretreatment is the use of a physical process to recover the free oils and fats. The
most commonly used physical separation process for the removal of free oils
and fats are fat traps, tilted-plate separators, and dissolved air flotation units. In
addition, centrifuge and electroflotation systems are occasionally used (73).
pH Control and Chemical Treatment. pH adjustment is often necessary to
prepare the effluent for subsequent treatment processes, especially the biological
ones. Chemical treatment involves the use of coagulants and flocculants such as fer-
ric chloride, aluminum sulfate, lime, and polyelectrolytes to reduce the total fatty
matters prior to the separation by flotation and sedimentation processes. pH adjust-
ment is often required for optimum results. Chemical treatment is usually required
for effluent from a chemical refining process.
Aerobic Treatment Process. Effluent from edible oil refineries has been shown
to be amenable to biological treatment, both anaerobic and aerobic processes
(74,75). The application of activated sludge process or aerated lagoon in this con-
text is well established in the edible oil industry (71, 73, 76). Figure 14 shows the
process flow of a typical activated sludge process.
Aerated lagoon treatment was very popular in the United States (73). The main
disadvantage of the aerated lagoon as compared to the activated sludge process was
that it required large land area and long hydraulic retention time (520 days). The
long retention time was required because of the low mixed-liquor suspended solids
(MLSS) concentration in the aerated lagoon. There was no recycle of the MLSS
from the discharge to the aerated lagoon. On the other hand the MLSS in the acti-
vated sludge process is maintained at optimum level by recycling the MLSS, which
is a standard feature of the process. The MLSS is normally maintained between
2000 and 5000 mg/L depending on the process requirement. Thus the retention
time can be very short, a fraction of a day.
The effluent from edible oil refinery tends to be deficient in nitrogen for aerobic
biological treatment. Nitrogen has to be added to fulfill the nutrient requirement. In
general, a ratio of BOD : N of 100 : 5 is required for biological treatment. Phos-
phorus is generally present in adequate amounts in the effiuent. This is because
phosphoric acid is used in the refining pretreatment process. A ratio of BOD : P
of 100:1 is adequate for the same purpose.
Figure 14. Conventional completely mixed.

Total fatty matter (TFM) concentrations in the raw effluent can lead to poor
performance of the treatment process. Therefore, care should be taken to ensure
efficient removal of TFM in the pretreatment (physical/chemical) process. TFM
is one of the main contributors to BOD.
Another important design criterion that affects the treatment process perfor-
mance is the organic loading rate. Typically, good treatment efficiency can be
achieved when the organic loading rate is less than 0.15 kg BOD per kilogram
MLSS per day. Under good maintenance and operation conditions, the discharge
BOD and suspended solids (SS) concentration of 20 and 30 mg/L, respectively,
can readily be obtained.
The treatment system for those refineries employing the chemical refining pro-
cesses consists of a train of processes with balancing tanks with pH adjustment,
chemical and physical treatment facilities (coagulation and flocculation as well
as air flotation), and the activated sludge process (40). The main problems encoun-
tered in the operation of the activated sludge plant are the high fluctuations of the
loading rates (both hydraulic and organic) and the requirement of close system
monitoring and supervision by skilled operators. This was much lacking in the
industry. Thus the process seldom achieved the expected treatment efficiency.
There has been very little publication/information on the treatment of palm oil
refinery effluent. Osenga (41) introduced a treatment process consisting of a cross
flow interceptor (CFI) for oil separation, physical and chemical treatment, and air
flotation units to remove the floes followed by a batchwise activated sludge process
for the liquid effluent treatment. This process also requires close supervision in
order to achieve the desired treatment efficiency.
Chin and Wong (74) attempted to treat palm oil refinery effluent by conventional
activated sludge process with limited success. The treated effluent was highly
colored with over 800 Hazen units.
Sequencing Batch ReactorA New Treatment Process for Palm Oil Refinery
Effluent. Since the early 1970s, an alternative aerobic process called the sequen-
cing batch reactor (SBR) has gained much popularity in the treatment of various
types of wastewaters. It is very similar to the old fill and draw (batch) system. There
are essentially five modes of operation in the SBR process: fill, react, settle, draw,
and idle.
All these operations take place in a single reactor instead of two as in the con-
ventional AS process (Figure 15). These operations have been made simple by the
advent of reliable and inexpensive microprocessor-based controllers. The process
has been found to be very efficient in the treatment of a variety of wastewaters.
Arora and co-workers (77) have reported that SBR has many advantages over the
conventional AS process. These advantages include equalization, ideal settling,
simple operation, compact layout, and cost saving (capital and running costs).
Irvine and co-workers (7880), Palis and Irvine (81), and Melcer and co-workers
(82) have used SBR to treat wastewater from small communities. Excellent removal
efficiency of BOD and nitrogen were recorded. Lo and co-workers (83,84) and Ng
(85) have also successfully applied the laboratory-scale SBR treatment to milking
parlor wastewater and piggery wastewater, respectively.
386 PALM OIL
Feed water
(raw effluent)
Feed/react React
slow air
Idle React fast

air
Decant Settle
Figure 15. Schematic representation of SBR cycle.
SBR Treatment Plant Characteristics. The flow diagram of the SBR treat-
ment plant used is shown in Figure 16. It consists of one holding tank and two reac-
tors. The effluent, consisting of circuit bleed water from the barometric condenser,
cooling water, floor washing and cleaning water, is collected in a sump where any
oil and fatty matter is recovered. It is then pumped to the holding tank where nutri-
ent and pH adjustments are carried out, if necessary, before it is fed to the one of the
reactors by a centrifugal pump. Air is supplied by a compressor through cone-
shaped diffusers installed at the bottom of the reactor. Discharge of the treated efflu-
ent is controlled by the operation of a valve. Time for each mode of operation is
predetermined. All the operations including the switching on and off of the pumps
and compressor, opening of discharging valve, etc., are controlled by a micropro-
cessor-based sequencing controller as shown in Figure 16. The program can be
easily changed on site to meet the operation requirements.
Figure 16. Schematic diagram of full-scale SBR process.
Characteristics of Palm Oil Refinery Effluent. Table 38 shows the character-

istics of the effluent from a typical palm oil refinery employing a physical refining
process. The effluent is slightly milky and is acidic in nature. It can be seen that the
characteristics of the effluent vary quite widely. Nutrient (N and P) contents seem
to be sufficient for biological process according to BOD : N : P of 100 : 5 : 1 ratio.
It contains low suspended solids and fatty matter, which are mainly dirt and free oil
from washwater and oil spillage.
TABLE 38. Characteristics of Palm Oil Refinery

Wastewater (physical refining and dry fractionation
process).
Parametera Range Mean
Temperature ( C) 3045 35
pH 3.87.0 5.3
BOD 501500 530
COD 1503000 900
TS 1002000 580
SS 50100 80
TN 020 10
P 1.010 4
O&G 25600 200
a
All parameters in milligrams per liter (mg/L) except pH and
temperature. BOD, biochemical oxygen demand: COD, chemical
oxygen demand: TS, total solids:SS, suspended solids; TN, total
nitrogen; P, phosphorus; O&G. Oil and grease.
388 PALM OIL
Process Performance. The performance of the SBR process is shown in

Figure 17. It can be seen that very stable and consistent performance can be
achieved. The process could sustain high fluctuations in feed chemical oxygen
demand (COD). With the feed COD varying between 240 and 1000 mg/L, the
SBR could produce highly purified final discharge. The COD and SS were consis-
tently less than 100 and 50 mg/L, respectively. It has been established that when the
COD is less than 250 mg/L, the BOD is consistently less than 50 mg/L.
Very good settling characteristics of the sludge were also observed. The MLSS
settled very well in less than 30 min (Figure 18). The SVI of the sludge was less
than 50 mL/g. Random checking on the viability of the sludge showed that the
MLSS consisted of over 80% of MLVSS.
5. END USES
Palm oil is used in both edible and nonedible applications (Figure 19). Ninety per-
cent of palm oil and its products are used for edible purposes. Currently palm oil is
used in food preparation or food manufacture worldwide. The remaining 10% of
palm oil and its products are used for nonedible applications, mainly in the soap
industry and in the manufacture of oleochemicals.
5.1. Food Applications

The use of palm oil in food dates back 5000 years. For edible and nonedible uses,
palm oil is normally refined. However, even today, unrefined palm oil is still used for
cooking in certain African villages much the same way as it used to be. Examples
demonstrating the range of palm oil applications in food are shortening, margarine,
vanaspati, deep frying fat, and specialty fats.
Shortening. At 20 C, palm oil has 2225% solid fat content and is a valuable
ingredient for shortening formulations. Unlike margarine, which is an emulsion of
about 80% oil and 20% water, shortening is pure (100%) oil and fat.
There are many types of shortenings, each tailor-made for a particular applica-
tion. There are also general-purpose shortenings that are used in the preparation of
many foods: in cooking and frying and in the manufacture of bakery products such
as cakes, cookies, rusks, wafer, pastries, and bread (86, 87).
Other related bakery products include cream fillings and icings. One important
function of a shortening is to incorporate and hold air, whether beaten in a cake
batter or creamed with sugar (87, 88). This ability to trap air enables the formation
of a porous structure and increases the volume of the cream or the baked cake. This
in turn influences texture: shortening contributes to tenderness of various baked
products. For optimum creaming ability and to be functional in cakes, the shorten-
ing must be stable in the b0 form. The b0 form refers to tiny fat crystals that are
responsible for the smooth texture of the shortening and aid in incorporation of
numerous air bubbles during the creaming process. In this respect palm oil short-
ening is at an advantage because the crystals exist in the b0 form (87, 89).
Figure 17. Performance of sequencing batch reactor.
389
390 PALM OIL
Figure 18. Setting characteristics of MLSS.
In a study on some shortening formulations based on palm products in combination

with other vegetable oils, Idris and co-workers (88) reported that a high-palmitic-
acid content was good for aeration of fat/sugar mixtures. The results of the study
indicated that palm stearincottonseed oil (3 : 2) shortening was best for aerated
cream filling. Table 39 shows baking performance of some palm-based shortenings.
A blend of palm stearin and low-erucic-acid rapeseed oil was very economical and
performed excellently in cakes. For application both in cream fillings and baking,
interesterified palm olein was the suitable material (88).
Palm oil can be blended with butterfat for use as a biscuit shortening or alterna-
tively diacetyl flavoring can be added to palm oil shortening to give the desirable
buttery taste (90). Idris and co-workers (90) reported that biscuits made with short-
enings containing palm oil and butterfat were not significantly different in flavor
from those made with 100% butterfat. In another study, Idris and co-workers
(91) reported the texture characteristics of pressed cookies made with oil, hydroge-
nated palm oil, and interesterified palm oil shortenings. With higher solid fat con-
tent and firmer consistency, shortening based on hydrogenated palm oil produced
firmer dough and harder as well as crispier cookies.
Margarine. Margarine is a type of emulsion consisting of fat and water.
Although the original purpose in developing margarine was to imitate butter, there
has since been a considerable diversification of margarine products, which now
include:
Table margarine in tubs

Table margarine in block form
Cream/cake margarine
Margarine for tropical climates
Puff pastry margarine
391
Figure 19. Palm oil utilization chart.

392 PALM OIL
TABLE 39. Baking Performance of Shortenings Based on Palm Products in Combination

with Other Vegetable Oils (88).
Specific Cake Volume

Shortening Composition Experimental/Control100 (%)
18% Hydrogenated palm oil (MP 41.5 C) 95

42% Palm stearin (IV 44)
40% Soybean oil
40% Cottonseed oil
40% Low-erucic-acid repeseed (LEAR) oil
50% Palm stearin (IV 44) 96
50% Soybean oil
50% Cottonseed oil
50% Low-erucic-acid rapeseed (LEAR) oil
40% Soybean oil
40% Cottonseed oil
50% Low-erucic-acid rapeseed (LEAR) oil
100% Interesterified palm olein 99
There are also low-calorie spreads that are similar to margarine in their physical
behavior but have a much higher water content. The physical properties of margar-
ines are largely determined by the fat component, and these properties vary with the
type of product. Thus tub margarines are soft and are spreadable straight from the
refrigerator. Table margarines in packets are not as soft but are spreadable at room
temperature, while cake or cream margarines are a little firmer than table margar-
ines. At the extreme end, pastry margarines are much firmer, in order to give the
flaky texture to the end product. Palm olein is suitable as the liquid component
of margarine blends, while palm stearin or hardened palm oil can be used as the
solid component (92). Ward (93) recommended that at least 10% palm oil be incor-
porated in canola-based margarines. Palm oil and palm oil products have also been
found to be very good ingredients for puff pastry margarine (94).
Vanaspati. Vanaspati, or vegetable ghee, is a major commodity in countries such
as India, Pakistan, Egypt, Saudi Arabia, Iraq, and Iran. In India and Pakistan, con-
sumers prefer products with a granular texture. In Iraq and Iran a smooth texture is
preferred. Kheiri (95) reported that vanaspati from India contained between 5 and
20% of palm oil products. A higher percentage, more than 50% of palm oil pro-
ducts, has been reported in Pakistan vanaspati formulations (95).
END USES 393
TABLE 40. Summaries of Published Work on Frying in Palm Oil.
Authors Country Conclusion
Von Zeddelman and Wurziger (96) Germany Hardened groundnut oil and palm oil
products best
Faur (97) France Palm olein and palm oil excellent for catering
Toregard and Eriksson (98) Sweden Palm oil and palm olein superior to hardened
soybeam oil
Herendi and Bethke (99) Germany Palm olein performed as well as groundnut oil
Deep Frying Fat. Palm oil is the most widely used industrial frying fat because
it has no unpleasant room odor, has high resistance to oxidation, does not polymer-
ize to gums, and has a nutritionally good fatty acid composition (50% unsaturated
and no trans acids). A number of published reports show palm oil products in a
favorable light when compared with alternative frying media (Table 40). The
good frying properties of palm oil are due to its moderate degree of unsaturation,
the absence of linolenic acid, and the presence of tocopherol. The tocopherol (380
890 ppm) acts as an effective natural antioxidant (100). For industrial frying of
instant noodles, palm oil is very suitable (101).
The liquid fraction of palm oil, palm olein, is also widely used for frying. In fact,
in Malaysia, it is now the main cooking oil used in most households. During frac-
tionation the tocopherols are somewhat concentrated in the palm olein, so that
refined palm olein typically has 500600 ppm total. The longer frying life and its
reduced tendency to foam and polymerize make it a better frying oil than corn or
soybean oils (102). In another frying study, Augustine and co-workers (103) found
that palm olein was comparable in terms of oxidative stability during frying with
the hydrogenated vegetable oils, namely hydrogenated soybean, hydrogenated sun-
flower, and hydrogenated cottonseed oils. The ability of palm olein to produce fried
foods of acceptable quality without the need for hydrogenation can be considered
an advantage.
5.2. Specialty Fats

Palm oil and palm kernel oil are also ideal raw materials for the production of
specialty fats. Specialty fats are particularly suitable for confectionery products,
especially chocolates. Specialty fats can be classified according to their chemical
composition into three types: (1) symmetrical, (2) lauric, and (3) high trans.
Symmetrical-type specialty fats contain predominantly SOS-type triglycerides.
The major triglycerides in cocoa butter are POSt, StOSt and POP. These triglycer-
ides, comprising about 75% of the total, are often summarized as SOS or symme-
trical triglycerides. Solvent fractionation (104) of palm oil (Figure 20) produces a
midfraction with a high content of the POP triglyceride but deficient in StOSt and
POSt. This deficiency can be corrected by adding Illipe fat (or Borneo tallow),
which contains these glycerides. Therefore the physicochemical characteristics
394 PALM OIL
Figure 20. Acetone fractionation of palm oil.
of cocoa butter can be closely matched by correctly blending palm midfraction with
Illipe fat.
Because of the similarity in the chemical compositions of the symmetrical-type
fats and cocoa butter, they are compatible with each other in almost any propor-
tions, and for this reason these specialty fats are usually called cocoa butter equiva-
lents (105). In certain countries, legislation allows up to about 15% of the cocoa
butter in chocolate to be replaced by symmetrical-type specialty fat and the product
may still be described as chocolate. In terms of texture and flavor these products
are very close to real (cocoa butter) chocolate.
Lauric-type specialty fats are produced from oils containing mainly lauric and
myristic acids. The simplest lauric-type fats can be made by hardening palm kernel
oil to slip melting points between 32 and 41 C. Palm kernel oil can also be fractio-
nated to give a stearin with much better melting properties than the hardened palm
kernel fats. The palm kernel stearin with physical properties resembling that of
cocoa butter, is called cocoa butter substitute, or hard butter. It is usually hydroge-
nated to further improve its meeting profile.
Hydrogenated palm kernel oil or olein is used as a cheaper alternative toffee fat
to replace the more expensive dairy butter, either completely or in combination with
butter. Hydrogenated palm kernel oil is also a good general-purpose coating fat.
END USES 395
High-trans-type specialty fats can be produced by a combination of selective

hydrogenation and fractionation from liquid oils. These high-trans-type fats can
be produced by selectively hydrogenating blends of soybean oil and palm olein
or palm olein alone. They are more compatible with cocoa butter than the lauric-
type cocoa butter substitutes, thus they are sometimes called cocoa butter partial
replacers.
5.3. Recent Food Applications

The newer applications of palm oil in foods include its use in emulsion-based, pow-
dered, and convenience food products. Butterfat has been traditionally used in ice
cream, but palm oil and palm kernel oil are now used commercially to replace it.
Similarly, palm oil can also replace butterfat in the manufacture of milk, to give a
product known as filled milk.
Palm oil is used because it is more economical than other oils and is easily avail-
able. In addition, it is more stable to oxidation than butterfat. Filled-milk powder
can be made from skimmed-milk powder recombined with refined palm oil.
Another use of a palm oil product is in infant food formulations. The low-melt-
ing olein has been found to be very suitable for use in infant food formulations
when blended with other vegetable oils. Low-melting olein contains 1015% pal-
mitic acid in the 2-position of the glycerol chain. This contributes to the high
digestibility of the product (106).
Apart from the products mentioned, there are many other foods that contain
palm oil and palm kernel oil products. These include soup mixes, cake and dessert
mixes, rendang or curry mixes, sardines, baked beans, breakfast cereals, shrimp-
paste powder, bouillon, peanut butter, and beverages. Palm oil products have also
been used as a spray oil on biscuits.
5.4. New Potential Food Applications

An important future application of palm oil in food is the use of refined red palm oil
in cooking. Refined red palm oil is a highly nutritious oil rich in vitamin E and
b-carotene. Nor Aini (107) reported that the deep red color of the oil blends well
with ingredients such as chili and curry, making the dishes more attractive and
appealing. The use of refined red palm oil is a possible alternative means of
combatting vitamin A deficiency, which is prevalent in many countries.
Another promising application of a palm oil product is the use of RBD palm
olein of high IV as salad oil. The use of palm olein as salad oil can be made possible
by blending it with other vegetable oils (108,109). Yet another potential use of RBD
palm oil is as a barbecue oil. Its high stability and bland taste makes it a good
choice for this application. The oil acts as a flavor carrier, and it also prevents
the barbecued meat from drying out, so that one gets a juicy and tasty end product.
Palm oil is indeed a versatile oil. Its applications are varied and it can be used in
almost any food.
396 PALM OIL
5.5. Nonedible Applications

For simplicity the nonfood uses of palm oil and palm kernel oil and their products
will be divided into two categories, i.e., those where products are made directly
from the oils (direct route) and those where they are obtained via the oleochemicals
route (Figure 21).
Direct Route. Soaps. Soaps are derived from oils or fats by reacting them with
caustic soda at 80100 C in the process known as saponification. The use of soap as
a laundering agent and for cleansing the skin is many centuries old. Although mod-
ern detergents have almost eliminated the use of soap for home laundry purposes,
soap is still the main ingredient in toilet bars for personal use. In 1990 the world
consumed 8.9 million tons of soaps, and consumption is expected to grow at 2.2%
Figure 21. Nonfood applications of palm oil and palm kernel oils.
END USES 397
TABLE 41. Fatty Acids Compositions of Selected Oils/Fats.
Weight Percentage
Palm Palm
Fatty Palm Palm Kernel Kernel Palm Soybean
Acids Oil Stearin Tallow Oil Olein Coconut Olein Oil
C6 0.3 0.4 0.2

C8 4.4 5.4 8.0
C10 3.7 3.9 7.0
C12 0.2 0.3 48.3 41.5 48.2 0.2
C14 1.1 1.3 2.5 15.6 11.8 18.0 1.0
C16 44.0 55.0 26.6 7.8 8.4 8.5 39.8 6.5
C18 4.5 5.1 21.8 2.0 2.4 2.3 4.4 4.2
C18:1 39.2 29.5 42.8 15.1 22.8 5.7 42.5 28.0
C18:2 10.1 7.4 2.3 2.7 3.3 2.1 11.2 52.6
Other 0.8 0.7 4.0 0.1 0.1 0.9 8.0
IV 53.3 35.5 3548 17.8 25.5 9.5 58.4 133
SAP. V 196 199 195 245 256 198 192
per annum, slightly faster than the world population (110). The developing coun-
tries are expected to show higher growth rates in soap consumption than the devel-
oped countries.
The incorporation of C16C18 and C12C14 fatty acids in soaps is important as
they provide the cleaning, solubility, and foaming properties required. Tallow and
coconut oil have been the traditional sources of these fatty acids. A comparison
between the fatty acid compositions of palm oil, palm stearin, tallow, palm kernel
oil, palm kernel oleins, and coconut oil (Table 41) indicates that the first three
are rich in C16C18 fatty acids while palm kernel and coconut oils are rich in
C12C14 fatty acids. However, for palm products to establish a niche in the market
as raw materials, soap manufacturers have to be convinced that apart from price
competitiveness, they will yield soaps with properties and performance comparable
if not superior to those from tallow and coconut oil.
Palm stearin (POs) and palm kernel olein (PKOo) are produced, along with palm
olein (POo) and palm kernel stearin (PKOs), when palm oil and palm kernel oil
(PKO) are fractionated. While palm olein and palm kernel stearin have higher
added value because of their specific food applications, POs and PKOo are nor-
mally sold at discount prices. Several studies carried out by Kifli and co-workers
(111) revealed that POs and tallow can be formulated together with PKO to
give soaps that are comparable with tallow-PKO blends (Tables 42, 43, and 44).
Since POs is cheaper than tallow, the resulting soaps are expected to be cheaper.
Perfume retention of palm-based soaps has also been found to be better than that
of soaps made from tallow (112, 113). More interesting are the observations of
Kifli and co-workers (111) on POs and PKOo blends: Soaps based on these were
found to have comparable foaming power and better color.
398 PALM OIL
TABLE 42. Characteristics of Soaps Based on Tallow-Palm Stearin-Palm Kernel Oil

Blends.
Blends
Parameters 00 : 80 : 20 60 : 20 : 20 40 : 40 : 20 20 : 60 : 20 80 : 00 : 20
Acid value 223 222 220 219 216

Titer ( C) 49 44 41 39 39
IV 29 29 27 31 34
Free caustic (%) 0.3 0.3 0.2 0.2 0.1
Moisture 9 10 8 8 9
Hardness 23 22 22 23 21
Foamability 255 255 245 265 260
Whiteness (Hunter) 81 82 84 81 80
TABLE 43. Characteristics of Palm Oil-Palm Stearin-Palm Kernel Fatty Acids

Blends in Soap.
Blends
Parameters 80 : 00 : 20 60 : 20 : 20 40 : 40 : 20 20 : 60 : 20 00 : 80 : 20
Acid value 214 215 217 219 219

Tiler ( C) 42 44 46 47 48
IV 38 34 33 30 34
Free caustic (%) 2.1 0.1 0.8 6.7 2.3
Moisture 7 7.6 8.6 8.2 7.8
Hardness 13.5 12.0 12.3 10.5 16.2
Foamability 320 355 340 150 295
TABLE 44. Characteristics of Soaps Based on Palm Stearin-Palm Kernel Olein Blends.
Blends
Parameters 90 : 10 80 : 20 70 : 30 60 : 40 50 : 50
Acid value 179 185 188 191 196

Titer ( C) 47 44 42 41 40
IV 28 30 28 28 25
Free caustic (%) 0.12 0.05 0.2 0.09 0.1
Moisture 16.3 22.7 25.2 17.2 27.6
Hardness 10 12 11 16 7
Foamability 340 270 315 340 365
Whiteness (Hunter) 94 93 92 90 93
Poor color and discoloration are common complaints expressed by soap manu-
facturers attempting to use palm products for the production of white soaps. Sapo-
nification color value (SCV), which represents the color of the saponified oil, will, to a
limited extent, indicate the whiteness of the soap produced from the oil. For white
soap the SCV of the oil has to be lower than 3R. Palm products have SCVs greater
END USES 399
TABLE 45. Saponification Color Values of NBD and RBD Palm Oil and Palm Oil Products.
Original Color Saponification Color

Samples No. Ave. Range Ave. Range
RBD palm stearin 11 2.1R 1.6R2.5R 6.4R 5.2R6.5R

NBD palm stearin 5 1.6R 1.5R 5.6R 5.4R6.1R
RBD palm oil 16 2.0R 1.2R2.8R 6.8R 5.5R9.0R
NBD palm oil 2 1.7R 1.7R 6.9R 6.7R7.2R
RBD palm olein 10 2.5R 2.1R2.9R 8.1R 6.7R8.9R
NBD palm olein 4 2.1R 2.0R2.3R 8.5R 7.5R8.9R
than 5R (Table 45) and are therefore unsuitable for the production of white soaps.
However, bleaching the oil with hydrogen peroxide has been found effective in
reducing the SCV and producing stable white soaps (114). Since the reaction
with hydrogen peroxide is exothermic, extra care must be exercised. Endogeneous
minor components present in palm oil could be one of the factors causing the color
and discoloration. Preliminary effort by Ooi and co-workers (115) identified hydro-
xy-a-carotene-5,8-epoxide and chrysanthemaxanthin as two of the possible minor
components causing the yellow color in palm oil soaps. Besides minor components,
soap manufacturers know that the presence of trace metals and synthetic antioxi-
dants contribute to the discoloration.
Diesel Substitute. Vegetable oils were used as motor fuel by Rudolf Diesel
in 1900 when demonstrating his compression engines (116). Since then many
publications referring to similar usage of cracked products of oils and fats have
been published.
Recent research (117) has demonstrated that crude palm oil can be used directly
as a fuel to run cars fitted with suitable diesel Elsbett engines. The exhaust fumes
from crude palm oil engines were found to be cleaner than those from diesel
engines, with essentially no sulfur or nitrogen oxides. It is also cheaper and safer
to transport crude palm oil than diesel because of the higher flash point (crude palm
oil at 240 C vs. diesel at 52 C).
Cost is always the main factor that determines large-scale utilization. However,
the initial results suggest that the use of crude palm oil as an engine fuel would be
30% more costly compared with petroleum diesel under Malaysian conditions.
Epoxidized Palm Oil and Products, Polyols, Polyurethanes, and Polyacrylates.
Epoxidized palm oil and palm oil products (EPOP) can be produced by reacting
palm oil, palm stearin, or palm olein with peracids. Preformed peroxyacetic and
peroxyformic acids, as well as peroxyacetic acid and peroxyformic acid generated
in situ, were studied by Ahmad and co-workers (118,119) to find suitable methods
for the production of EPOP. The best procedures were found to be preformed
peroxyacetic acid and peroxyformic acid generated in situ (Table 46).
Epoxidized oils, especially epoxidized soybean oil (ESBO), are used extensively
as plasticizers and stabilizers for plastics, particularly polyvinyl chloride (PVC). A
plasticizer increases the workability of a plastic while a stabilizer reduces the rate
400 PALM OIL
TABLE 46. Reaction Conditions for the Production of Epoxidized Crude Palm Oil.
Preformed Peroxyacetic Preformed Peroxyformic

Peroxyacetic Generated Peroxyformic Generated
Reaction Parameters Acid in situ Acid in situ
Mole ratio acid : H2O2 2.06 : 1 1.1 : 1 4.6 : 1 0.35 : 1

Catalyst (%) 1.52 0.02
Temperature ( C) 4045 RT
% Peracid formed 13.7 11.0
Epoxid temperature ( C) 6065 6070 5060 5060
Oxirane oxygen content (%) 2.62 1.87 Low 1.82
2.08 (toluene)
1.86 (CH2Cl2)
of degradation of a plastic by heat, light, or microorganisms. Epoxidized oils can

fulfill both functions, and their compatibility with a plastic increases with their
epoxide content.
Because palm oil and its products have lower iodine values, the epoxide contents
of EPOP are lower than that of ESBO. As plasticizers and stabilizers, EPOP are
therefore not expected to perform better than ESBO, but their performance could
be made comparable by slight modifications of the formulations. PVC jungle and
rain boots plasticized and/or stabilized with EPOP have been produced that are
comparable in performance to those plasticized and/or stabilized with ESBO (120).
The value of epoxidized oils lies in the versatility of the epoxide rings. Being
labile, they can easily be converted to other useful functional groups, thus diversi-
fying end uses. EPOP can be converted to various polyols by reacting them with
short-chain polyhydric alcohols in the presence of catalysts. By changing the ratio
of EPOP to polyhydric alcohols and the types of polyhydric alcohols, polyols with a
range of hydroxyl values and viscosities can be produced (121123) (Table 47).
Polyols when reacted with isocyanates produce polyurethane foams. The water
formed in the reaction acts as an internal blowing agent, thus avoiding the need
to use environmentally unfriendly blowing agents such as chlorofluorocarbons.
Polyols from EPOP react with isocyanates at a slower rate than do polyols based
on petrochemicals. The resulting foams, however, have regular cell structures and
exhibit good hydrophobicity. With suitable formulations these properties could be
fully exploited to give rise to interesting products (124).
TABLE 47. Properties of Palm-Based Polyols.
Ratio of EPOo-Polyhydric Alcohol
Parameters 1:1 2:1 4:1
OH value 350450 200300 150200

Viscosity (mPa S) 9801300 13002100 35004700
END USES 401
TABLE 48. Characteristics of Some Polyacrylates Coating Based on Epoxidized Palm

Oil/Products After uv Curinga
Formulations
Additives 1 2 3 4 5
EPOLA 60 60 60 60 80
PUA 5
HPA 10 10 10 10 10
NPGDA 15
TMMTA 25
TMPTA 25 10
PETA 25
Viscosity (cps/25 C) 614 423 705 228 947
Pencil hardness 3H 4H 4H 4H <2B
<2B 2B <2B <2B
Gel fraction (%) 86 87 86 88 84
a
In all the formulations 5% benzophenone was used as photoinitiator. EPOLA, epoxi-dized palm olein
acrylates; PUA, polyurethanes acrylates; HPA, hydroxypropyl acrylate; NPGDA, neopentyl glycol diacrylates;
TMMTA, trimethlolmethane triacrylates; TMPTA, trimethlolpropane triacrylates; PETA, pentacrythritol
triacrylates.
Polyacrylate resins can be produced from EPOP by reacting them with acrylic
acids. These resins can be applied on solid surfaces and when they are cured by
ultraviolet irradiation, clear and glossy finishes result. The hardness and the tacki-
ness can be increased or reduced by varying the amount and types of crosslinkers
and the strength of the irradiation used (125127). Characteristics of some polya-
crylates based on epoxidized palm olein are listed in Table 48.
Oleochemical Route. Oleochemicals. Oleochemicals are chemicals derived
from oils or fats. They are analogous to petrochemicals, which are chemicals
derived from petroleum. The hydrolysis or alcoholysis of oils or fats form the basis
of the oleochemical industry. The hydrolysis of the triglycerides composing oils and
fats produces fatty acids and glycerol. If oils or fats are made to react with an alco-
hol instead of with water, the process is alcoholysis, and the products are fatty acid
esters and glycerol.
RCO OCH2 RCOOR CH2 OH

ROH
CHO OCR RCOOR + CH2 OH
RCO OCH2 RCOOR CH2 OH
Oils or fats Fatty acids or esters Glycerol
In hydrolysis, R0 H.
In alcoholysis, R0 alkyl group.
Fatty acids or their esters can be used as the starting materials for making fatty
alcohols and fatty nitrogen compounds. These products can be further modified to
produce various derivatives. Hence oleochemicals are often divided into at least two
categories: basic oleochemicals and derivatives. The five basic olcochemicals are
402 PALM OIL
TABLE 49. Production of Basic Oleochemicals from Pacific Regions

Compared to the World Production.
Countries 1990 1995 2000
Malaysia 262,200 806,950 1,200,000

Philippines 172,470 285,000 480,000
Indonesia 62,700 199,500 400,000
Thailand 11,000 22,000 44,000
Total 508,370 1,313,450 1,124,000
World (%) 4,417,000 5,264,000 6,098,000
12% 25% 35%
fatty acids, esters, alcohols, nitrogen compounds, and glycerol. There are various
types of derivatives that can be produced from these through different chemical
modifications. Figure 22 shows some of them.
Raw Materials for Oleochemicals. Oleochemicals or derivatives based on
C12C14 and C16C18 chain lengths have a variety of uses. Tallow and coconut
oil have been the traditional raw materials used for the production of C16C18 and
C12C14 chain lengths, respectively. While tallow is produced by the developed
countries such as the United States, the world has to rely on the Pacific region
for the supply of lauric oils (C12C14 source). The Philippines has been the
main supplier of lauric oils.
Palm-Based Oleochemicals. Palm kernel oil, like coconut oil, is a lauric oil. Its
fatty acid composition is in fact very similar to that of coconut oil (Table 41).
Table 49 shows the present volume of production of basic oleochemicals in the
Pacific region and forecasts of production up to the year 2000 (127). The oleochem-
icals produced by Malaysia and Indonesia will be based mostly on palm and palm
kernel oils, while those from the Philippines will be based mostly on coconut oil.
According to this forecast, the ASEAN region will be producing 35% of the worlds
basic oleochemicals by the year 2000. Malaysia alone will account for nearly 20%.
Uses of Oleochemicals Based on Palm Oil and Palm Kernel Oil. (a) Fatty Acids. The
most common method for the production of fatty acids adopted by the oleochem-
icals industry is high-temperature and high-pressure fat splitting. The fatty acid
mixture produced is separated into broad cuts or pure fatty acids by simple or frac-
tional distillations. Tables 50 and 51 list examples of fatty acids derived from palm
products. The exact specifications of the various fatty acids produced vary slightly
depending on the exact raw materials and process used. The specifications could
also change due to continuous upgrading of processes.
One of the traditional raw materials used for the production of stearic acid is
tallow, and very often, consumers or customers will ask for products equivalent
to stearic acid from tallow. Single, double, and triple pressed stearic acids from
palm oil are in fact produced via distillation processes, but similar terminologies
were used to indicate their similar characteristics. Besides their light color, fatty
acids derived from palm products have a low content of unsaponifiables, indicating
excellent purity.
END USES 403
TABLE 50. Palm-Based Fatty Acids.
Fatty Acid Compositions

Fatty Acids C8 C10 C12 C14 C16C18 C18 : 1 C18 : 2 C18 : 3
Distilled PKO 3 3 50 15 9 16 2
Stripped PKO 52 17 10 16 3
Distilled PO 0.5 1.5 45.5 37.5 9.5 0.5
Distilled POs 0.5 1.5 61.6 25 6 0.5
Single pressed 0.5 1.5 50.0 10
Double pressed 0.5 1.5 53.0 3
Triple pressed 0.5 1.5 51.0
Lauric 70% 0.2 >70 >22 2
Lauric 92% 25 >92 >2
Lauric 98% 01 >98
TABLE 51. Palm-Based Fatty Acids.

Quality Parameters
Color
Acid Sap. Iodine Unsap. Titer
Fatty Acids Value Value Value Matter ( C) R Y
Lauric 70% 270275 270275 0.11 0.5 3335 0.5 2

Lauric 92% 277281 277281 0.10.5 0.5 4142 0.2 2
Lauric 98% 278282 278282 0.10.2 0.2 4344 0.2 1
Rubber g. stearic >195 >196 <10 >52 5.0 50
Candle g. stearic 208214 209215 1.0 0.5 5356 0.4 4
Cosmetic g. stearic 206211 207212 0.5 0.5 5556 0.3 3
Without further chemical modification, fatty acids are used in rubber processing
and in the manufacture of candles and cosmetic products.
(b) Fatty Acids for Rubber Processing. In rubber processing, fatty acids are
added as processing aids with softening effect, as external lubricants and as vulca-
nization accelerators. The chain lengths of the fatty acids have no effect on the
performance of the fatty acids. However, a high degree of unsaturation will
interfere with the process (128).
(c) Fatty Acids for Candles. Fatty acids or mixtures of fatty acids and petroleum
waxes can be used for the production of candles. For maximum shrinkage in order
to ensure easy removal from the mold, about 7:2 ratio of C16:C18 fatty acids is
required (129). This ratio favors fatty acids from PP since they have higher palmitic
acids content.
(d) Fatty Acids for Cosmetics Products. Only good grades of fatty acids can be
used for the production of cosmetic products. The normal types of fatty acids used
are myristic, palmitic, and stearic acids. They are used for various purposes such as
lather improver, conditioners, and to provide luster and sheen (128).
404 PALM OIL
(e) Fatty Acids for Soaps. The most important application of fatty acids is
for the production of soaps via a neutralization process. As discussed, white soaps
cannot be prepared directly from PP due to their high SCVs. When palm fatty
acids are distilled, part of the impurities are removed and the SCVs of palm fatty
acids are usually lower than 3R. Good-quality soaps can therefore be derived
from palm fatty acids. Besides the ease in production, the use of fatty acids allow
soap formulators to blend their own ratio of fatty acids, thus allowing greater
flexibility.
Lately, purely for aesthetic reasons, transparent and translucent soaps are gaining
popularity, especially in South America. Stearic acid and triple pressed stearic acids
can be used to produce soaps having good transparency. High palmitic acid content,
however, appears to reduce transparency (130). The crystals of transparent/
translucent soaps based on palm kernel oil and palm fatty acids are found to be
in the b form (131).
(f) Fatty Acids for the Production of Metallic Soaps. Another important applica-
tion of palm fatty acids is for the production of metallic soaps other than sodium
soaps. The most common are the Ca and Zn palmitates or stearates. They can be
prepared either via the fusion or the precipitation method. During the processing of
rubber, the processability is improved regardless of the fatty acids used. However,
Zn soaps were found to provide better internal lubrication (132).
(g) Fatty Acids for Medium-Chain Triglycerides. When palm kernel oil is used
as the starting raw material, the medium-chain fatty acids, i.e., C6C10, present are
normally stripped off since these acids are known to cause skin irritation. Originally
considered as waste products, these medium-chain fatty acids can be resynthesized
into a new class of oil known as medium-chain triglycerides (MCTs). MCTs have
many applications such as in the flavor and fragrance industries, in surface treat-
ment of confectionery products, as release agent in the baking industry, and for
the lubrication of machines (133).
(h) Fatty Esters. Esterification of fatty acids with alcohols and alcoholysis of
triglycerides are two of the most common methods used for the production
of fatty esters based on palm products. Table 52 lists some of the properties of
the various types of palm methyl esters in comparison to fatty acids. Fatty esters
are found in several industries such as textiles, cosmetics, pharmaceuticals, plas-
tics, and lubricants. As synthetic lubricants, fatty esters are getting closer atten-
tion (128, 134, 135) due to their good lubricity, minimum viscosity change with
temperature, low-temperature fluidity, and high thermal and oxidative stability
(128).
(i) Fatty Esters for Soaps Production. Fatty esters are increasingly being used
for the production of pure white soaps (136). In contrast to fatty acids, soaps
produced from fatty esters are normally better in quality since the fatty esters
can be better purified. In the process, alcohols (usually methanol) will be produced.
Complete removal of alcohol is necessary before the soaps can be certified fit
for use.
(j) a-Sulfonated Methyl Esters. a-Sulfonated methyl ester (SME) is a new class
of anionic surfactant. SME has received a lot of attention as an active ingredient for
TABLE 52. Comparison Between Fatty Acids and Fatty Methyl Esters.
Color
Acid Sap. Iodine Unsap. Titer
Product Value Value Value Matter ( C) R Y C12 C14 C16 C18 C18:1 C18:2
PKO FA 225 256 15 1 22 1 10 4753 1519 811 13 1219 24

265 266 20 26
PKO FME 1 238 14 0.5 9 0.3 3 45 14 7 1 12 2
248 19 50 18 10 3 19
POs FA 206 207 28 1 48 3 30 0 0 55 3 20 4
216 217 38 54 1 3 70 7 30
POs FME 0.5 196 22 1 21 0.5 5 0 0 55 3 20 5
208 45 1 3 70 7 30 10
Stearic A 188 188 1 1 65 1 10 6 92 0 5
98% 195 196 3 66 8 94 1 10
Methyl stearate 92% 1 187 1 1 36 0.2 2 6 92
195 191 8 94
406 PALM OIL
the production of washing and cleaning products due to several factors that include
(136, 137):
1. Easy production procedures

2. Good detergency especially in hard water and in the absence of phosphates
3. C16C18 have been found to exhibit good detergency
4. Good biodegradation characteristics
Distilled fatty methyl esters with low iodine values are used as the starting raw
material for the production of SME. The fatty methyl ester is first reacted with
sulfur trioxide at 8090 C in a falling-film reactor. The dark product obtained
from this process is bleached using hydrogen peroxide. After bleaching, the
lighter color product is neutralized with alkali to produce an a-sulphonated methyl
ester.
Due to the good detergency properties of C16C18 fatty methyl esters, palm
stearin provides a suitable and cheap source of raw material for the production
of SME (138). The detergency properties of SMEs derived from palm stearins
have been found to be comparable to linear alkyl benzene sulfonates (LAS), the
workhorse of the detergent industry. It was interesting to note that SME based
on palm fatty acids distillates (PFAD), a by-product of the physical refining
industry (Figure 22), performs as well as those derived from fractionated esters.
These findings indicate that SME could very well be an important anionic surfac-
tant for the future, and palm fatty acids distillates can be a cheap source of raw
materials.
(k) Diesel Substitute. Palm-based methyl esters have been extensively tested as
diesel substitutes in taxis, buses, lorries, tractors, and stationary engines (139, 140).
Methyl esters from crude palm oil and crude palm stearin have very similar fuel
properties as petroleum diesel (Table 53). Data available to date indicate that
cold starting is easy, engines run smoothly with less unburnt hydrocarbon, CO,
SO2, and black smoke in the exhaust fumes. No dilution of the lubricating oil
was observed and ignition lag was shortened (139). In contrast to crude palm oil,
the use of palm methyl esters as diesel substitute does not require any modification
of the engines. The economic viability of palm methyl esters as diesel substitute
will depend on the cost of diesel, crude palm oil, and glycerol.
(l) Fatty Alcohols. The most important application of fatty methyl esters is for
the production of fatty alcohols. For economic reasons, three technologies for the
production of fatty alcohols have gained worldwide acceptance:
1. High temperature and pressure hydrogenation of oils/fats

2. High temperature and pressure hydrogenation of fatty methyl esters
3. High temperature and pressure hydrogenation of fatty acids
To date the most common method for the production of fatty alcohols is via
high-temperature and high-pressure hydrogenation of fatty methyl esters using a
END USES 407
Figure 22. Processing of palm oil/palm kernel oils.
copper-chromite catalyst in a fixed-bed reactor. If the unsaturation present in the

molecule needs to be preserved, then a Zn-containing catalyst is used.
Novel catalysts (141) such as RuSn and ReSn can be used to hydrogenate
fatty acids or methylesters to fatty alcohols. Using these catalysts, the hydrogena-
tion can be carried out at the same temperature (250 C) but lower pressure (50 bars),
and the unsaturations present remain unaffected. The presence of tin has been found
to be instrumental in the preservation of the unsaturation during hydrogenation.
RuSn supported on alumina was found to be most selective when they were
prepared via the solgel method.
Fatty alcohols as such find limited uses. Cetyl and stearyl alcohols are used as
suppressors of water evaporation in dry areas. Unsaturated alcohols are used as
emulsifiers and textiles auxiliaries. More than 90% of the fatty alcohols produced
worldwide are used for the production of anionic (fatty alcohol sulfates and fatty
408 PALM OIL
TABLE 53. Fuel Characteristics of Malaysian Diesel, Methyl Esters from Crude Palm Oil
and Crude Palm Stearin.
Malaysian Methyl Esters Methyl Esters

Product Test Conducted Diesel of CPOa of CPSa
Specific gravity ASTM

D-1298 ( F) 0.8330 at 60.0 0.8700 at 74.5 0.871 at 78.0
Sulfur content (% wt) IP 242 0.10 0.04 0.002
Viscosity at 40 C (cST)
ASTM D-97 4.0 4.5 4.6
Pour point ( C) ASTM D-97 15.0 16.0 17.0
Distillation D 86 ( C)
I.B.P. 228.0 324.0 320.0
10% 558.0 330.0 331.0
20% 270.0 331.0 332.0
50% 298.0 334.0 335.0
90% 376.0 343.0 343.0
F.B.P. 400.0 363.0 349.0
Final recovery (%) 98.0 98.5
Cetane index ASTM D-976 53 50 52
Gross heat of combustion
(kJ/kg) 45,800 40,135 39,826
Flash point ( C) ASTM D-93 98 174 165
Conradson carbon residue
(% wt) ASTM D-189 0.14 0.02 0.02
a
CPO, crude palm oil; CPS, crude palm stearin.
alcohol ether sulfates) and nonionic (fatty alcohol ethoxylates) surfactants. These
derivatives are extensively used in the production of washing and cleaning products
(128).
(m) Fatty Nitrogen Compounds. The most common fatty nitrogen compounds
are fatty amides, nitriles, amines, and quarternary ammonium. The most important
of these are the quarternary ammonium compounds, better known as quats. Lately
manufacturers in the developed countries are voluntarily reducing or stopping the
use of quats and imidazoline derivatives in softeners and conditioners in view of
the findings that they may not be completely biodegradable and may involve the
possible formation of nitrosamine (142) in products containing them.
(n) Glycerol. Glycerol is a valuable coproduct of the oleochemical industry.
Although glycerol can be produced synthetically, natural glycerol (i.e., glycerol
derived from oils or fats) are preferred by the customers or consumers of today.
During the production of fatty acids via fat splitting or fatty ester via alcoholysis,
mixtures of 1030% glycerol and water, known as sweet waters are produced.
These sweet waters can be processed to pure glycerol via distillation or ion-
exchange methods (116). To achieve the pharmaceutical grade further treatment
with activated carbon is carried out. Table 54 gives some of the important
characteristics of glycerols produced from palm oil.
END USES 409
TABLE 54. Types of Glycerol Derived from Palm Oil and Products.
Glycerol
Quality Parameters Crude 99.5% 99.8%
Ref. Density 20/20C 1.2623 1.2631

Ref. index nD20 1.4731 1.4737
Glycerol % 88 99.5 99.8
a Color 5 5
Acid content 0.1 0.1
Saponification eqv. 1.0 1.0
Ash (%) 1.0 0.01 0.01
Cl (organic cl) (ppm) 2(5) 2(5)
Heavy metals (ppm) 2 1 1
Glycerol is a polyhydric alcohol that finds wide uses in several areas of applica-
tions. These include; as solvent or drugs carrier in pharmaceutical products; as
humectants in cosmetics and tobacco; as ingredients for the production of explo-
sives; as plasticizer/stabilizer for less polar polymers; as antifreeze or heat transfer
agent; as hydraulic fluid; for the production of polyesters that can be used in grease
and/or lubricants; and for polyols and polyurethanes and mono and diglycerides,
which are useful food emulsifiers.
Ability to reduce the surface or interfacial tension of water and oil is one of the
properties required of an emulsifier. Different chain length fatty acids in monogly-
cerides were found to have little effect on the interfacial tension between palm oil
and water, in contrast C18:2 (linoleic acid) monoglyceride (143) reduces the inter-
facial tension to greater than 50% (Table 55).
Prospects. The amount of palm oil/palm oil products used in the production
of soaps is expected to increase in the near future especially in the developing
countries. Besides being price competitive and exhibiting good performances, soaps
from palm oil/palm oil products are acceptable by all religions.
There is a strong competition between products derived from oleochemicals and
those derived from petrochemicals. With the current awareness on environmental
issues and preference for environmentally friendly products, the utilization of
palm oil/palm oil products for nonfood applications via the oleochemicals route
is also expected to increase. Due to the ready availability of raw materials, technol-
ogy, capital, and market demand, the nonfood applications of palm oil/palm oil
products are expected to have a bright future.
TABLE 55. Interfacial Tensions of RBD Palm Oil and Water at 1% Concentration
of Monoglyceride of Various Fatty Acids.
Fatty Acids C12 C14 C16 C18 C18:1 C18:2
Interfacial tension (mN/m) 11.3 12.2 12.1 13.1 14.8 6.5

410 PALM OIL
6. POTENTIAL DEVELOPMENTS
6.1. Environmental Trends: Toward Zero Waste Operation

With the growing awareness for the protection of the environment, there is a greater
need for producers to improve the environmental profile of their products. Consu-
mers and regulatory bodies expect more information on the effects of products on
the environment during their manufacture, use, and disposal. The Life Cycle
Assessment (LCA) is used as a holistic approach to assess the impact of a product
throughout its life cycle. An important aspect of LCA is recycling and waste
management.
In oil palm plantations, the main residues that must be disposed of are the fronds
that are pruned regularly and the biomass comprising the palm trunk and fronds at
the end of the crops economic cycle (about 25 years). In commercial practice,
pruned fronds are redistributed to the fields white the biomass, senescent palms,
are recycled with the zero burning technique of replanting. Without burning the
palm residues, this approach does not pollute the air and it enhances soil fertility
through recycling of organic matter and plant nutrients.
Figure 23. Current use and recycling of by-products and wastes from oil palm plantings and
conventional palm oil milling (144).
NUTRITIONAL EFFECTS OF PALM OIL 411
The main by-products and wastes produced from the processing of palm oil are
the empty fruit bunches (EFB), palm oil mill effluent (POME), palm fiber, and palm
kernel shell. EFB and POME have been used extensively as mutch and organic fer-
tilizers in oil palm plantations while palm fiber and shell are used as fuel, making
the palm oil mill self-sufficient in energy (Figure 23). Excess shell has been used
for road surfacing in estates.
The underutilized wastes from the oil mill are biogas generated by anaerobic
digestion of POME and clinker and boiler ash; however, the quantities of the latter
two are insignificant to cause any environmental impact. In segments of the palm
oil industry, biogas is being used to generate heat and electricity to supplement the
energy demand by the subsidiary factories.
Although most of the by-products and "wastes" from oil palm planting and palm
oil processing are recycled or utilized, there is still room for improvement, the ulti-
mate aim being toward zero waste. Concerted efforts by all sectors are being direc-
ted toward improving the efficiency of oil palm cultivation and processing with a
view to optimizing the use of inputs and energy and reducing the production of
wastes/effluents. Currently, the by-products from the palm oil industry are used
mainly as organic fertilizers, soil ameliorants, and fuel. Research and development
effort has shown that these resources can be made into value-added products such
as fiberboards, furniture, and single-cell proteins. Table 56 shows the potential uti-
lization of oil palm by-products.
7. NUTRITIONAL EFFECTS OF PALM OIL
7.1. General Nutritional Properties

Archeological evidence shows that palm oil has been consumed by humans for
more than 5000 years (145). Its digestion and absorption rates in the human
body are in excess of 97%, which is very similar to other common edible oils
and fats. In many communities palm oil is an important source of dietary energy
and provides sufficient quantities of the essential fatty acid, linoleic acid (18:2,
n6) for normal healthy metabolic functions. Like all other edible oils of vegetable
origin, palm oil is considered cholesterol free.
7.2. Effect of Palm Oil on Blood Lipids

Being arbitrarily etassified as a saturated fat, palm oil has been postutated to
increase serum cholesterol levels and hence enhance the risk of coronary heart
disease (CHD). Recent findings, besides earlier published literature, however, indi-
cate that this hypothesis is not uniformly true. For instance, Kris-Etherton and
co-workers (146) demonstrated that feeding a palm oil diet to rats did not raise plas-
ma cholesterol in comparison to a highly polyunsaturated corn oil diet. Similarly,
Sugano and co-workers (147) were unable to establish significant differences in
plasma cholesterol in rats fed a palm olein diet compared to other polyunsaturated
oils. Sundram and co-workers (148) compared the effect of palm oil and its
TABLE 56. Utilization of Oil Palm/Palm Oil by-Products and Wastes (144).
Present Level
Waste/By-product Quantitya Where Utilized of Utilization Status Potential New Uses
Oil Palm Plantations

Pruned fronds 10.4 tons/ha Recycled in plantation Very high By-product Vitamin E extraction, fiber-board, etc.
Palm trunks and fronds at
replanting 89.9 tons/ha (i) Recycled in plantation Very high By-product Wood product, pulp, paper, animal feed,
palm heart, glucose, cellulase, fuel etc.
(ii) Furniture Very low By-product
Palm Oil Mill
EFB 2023% Mulching in plantation Very high By-product Fiber board: MDF
Fiber 1213% Fuel to boiler Very high By-product Fiberboards
Shell 68% Fuel to boiler Very high By-product Activated carbon; potting medium
Decanter solid 23% Land application as
fertilizer Moderate-high By-product Animal feed
Boiler ash 0.40.6% Surface landfill in
plantation/fertilizer Low By-product/waste Fertilizer/soil ameliorant
Clinker Small quantity Landfill/disposal Very low-low Mainly waste Surface landfill
Sterilizer condensate 1220% (i) Feed to ETP Very high Waste/by-productb Cellulase, single-cell protein
(ii) Recycle for dilution Low By-product Crude oil dilution
Centrifuge waste 4050% Feed to ETP Very high Waste/by-productb Oil recovery for acid oil production
Decanter effluent 3040% Feed to ETP Very high Waste/by-productb
Hydrocyclone/claybath water 511% Feed to ETP Very high Waste/by-productb Recycling to reduce quantity
Factory washing 48% Feed to ETP High Waste/by-productb De-oiling-oil recovery for acid oil produc-
tion
Effluent Treatment Plant
Sludge cake (i) Land application as
fertilizer Moderate-high By-product
(ii) Animal feed Very low-low By-product
Anaerobic solid 510% Land application as fertilizer Very high By-product
Aerobic solid <5% Land application as fertilizer High By-product
Biogas 28 m3/1 EFB Biogas engine Very low Mainly Heat and power generation
a
Figures in percentages refer to % to FFB.
b
These are mainly waste products in the mill but as they are reused immediately after treatment in the Effluent Treatment Plant (ETP), they are considered by-products.
fractions with two commonly used polyunsaturated oils, namely soybean and corn
oil. It was demonstrated that palm oil feeding did not elevate plasma cholesterol
whereas high-density lipoprotein cholesterol (HDL-C) tended to be raised on the
palm oil diet relative to the corn oil diet. Similar observations have also been recorded
in other animal models including the rabbit, chicken, and hamsters (149151).
7.3. Human Studies Evaluating the Effect of Palm Oil on Blood Lipids
The effects of palm oil on serum lipids and lipoproteins recorded in animal studies
have similarly been observed in several human studies. In some early human studies
(152, 153), it was reported that subjects on a palm oil diet had elevated plasma and
low-density lipoprotein cholesterol LDL-C levels compared to a diet containing a
polyusaturated fat. However, on a critical reassessment of these and other relevant
studies (154), it was found that plasma cholesterol levels after the palm oil period
were actually lower than at the point of entry of the experiments when the subjects
were on their habitual diets.
Sundram and co-workers (155) performed a dietary intervention trial in a free
living European (Dutch) population, consuming a diet that was traditionally high
in fat content. The habitual fat intake of this population was maximally replaced
with palm oil (up to 70% replacement). The consequence of this fat replacement
was carefully monitored over an experimental duration of 6 weeks using a dou-
ble-blind crossover design. It was found that compared to a Western-type diet,
the palm oil diet did not raise serum total cholesterol (TC) and LDL-C. Maximum
substitution with palm oil, however, resulted in an elevation of the beneficial
HDL2-C while significantly lowering the triglyceride content in the atherogenic
LDL fraction. The apolipoproteins (apo AI and apo B), which are increasingly
being recognized as better indicators of atherogenic risk, were also regulated by
the diet wherein a net beneficial effect (lower Apo B/Apo AI ratio) was evident
on consumption of palm oil. Thus palm oil, when used to replace the habitual fat
content in a Western-type diet, had no deleterious effects on serum or lipoprotein
cholesterol and triglyceride levels. In fact, as demonstrated in this study, the use of
palm oil caused a slight improvement of the cardiovascular risk indicators asso-
ciated with lipoproteins and apolipoproteins.
A human study by Marzuki and co-workers (156) using young volunteers eval-
uated the effect of consuming foods containing either palm olein or soybean oil. In
normal healthy volunteers the level of blood cholesterol was not changed by the
palm olein or soybean oil diets. Similarly both LDL-C and HDL-C levels were
unaffected by these diets. When the same diets were fed to volunteers having
high blood cholesterol levels (hypercholesterolemia), the soybean oil diet was
found to induce higher cholesterol levels than the palm olein diet. Similarly
LDL-C was also raised by the soybean oil diet.
In a similar experiment conducted on a Malaysian population (157), diets con-
taining palm olein, corn oil, and coconut oil were evaluated for their potential to
modulate serum lipids. A reduction in serum cholesterol was observed on admin-
istering a palm olein or corn oil diet relative to a coconut oil diet. A second study
(158) evaluated the effects of palm olein and olive oil on serum lipids and lipopro-
414 PALM OIL
teins in comparison to a coconut oil diet. Each test oil was served as the sole cook-
ing fat and contributed 23% of the total dictary energy or two thirds of the total fat
intake. The coconut oil diet significantly raised all the serum lipid and lipoprotein
parameters measured, i.e., TC, LDL-C, and HDL-C. However, the one-to-one
exchange between palm olein (rich in 16 : 0) and olive oil (rich in 18 : 1) resulted
in identical TC (192. 193 mg/dL). LDL-C (130, 131 mg/dL), and HDL-C (41,
42 mg/dL). This indicates that in healthy normocholesterolemic humans, palm olein
can be exchanged for olive oil without affecting the serum lipoprotein concentration
or distribution.
In a study of 30 middle aged men, six different fats were used as ingredients of a
normal American diet, forming 50% of the total fat intake (159). When palm oil
was the test fat, there was no significant effect on TC but HDL-C and apolipopro-
tein AI was increased while apolipoprotein B was decreased as compared with the
baseline diet.
Heber and co-workers (160) evaluated diets enriched in palm oil, coconut oil, or
hydrogenated soybean oil for three 3-week test periods in healthy American males.
No significant changes in TC, LDL-C, or apolipoprotein AI or B were apparent fol-
lowing consumption of the palm oil diet. They therefore concluded that enrichment
of the diet of normal healthy individuals with palm oil does not increase cardiovas-
cular risk factors related to lipids and lipoproteins. Truswell and co-workers (161)
compared the effect of palm olein and canola oil on plasma lipids and reported that
the mean 3% rise in TC on palm olein compared with a normal Australian diet was
predominantly due to a 10% rise of HDL-C.
7.4. Possible Mechanism for the Cholesterol-Lowering Potential

of Palm Oil
It has long been recognized that the cholesterol-raising potential of the saturated
fatty acids is variable. Thus, it has been shown that stearic acid (C18 : 0) does
not raise serum cholesterol (162). The major saturated fatty acid in the human
diet as well as in palm oil is, however, palmitic acid. This, together with lauric
(C12 : 0) and myristic (C14 : 0) acids, is considered hypercholesterolaemic (163).
Hayes and co-workers (164) recently reexamined this hypothesis in nonhuman pri-
mates (monkeys), using dietary fats containing predominantly lauric and myristic
acids (coconut oil) or palmitic acid (palm oil). They showed that compared to diets
rich in lauric and myristic acids, diets containing palmitic acid were actually neutral
in their effect on both total serum and LDL-C. Hayes and Khosla (165) have
advanced a hypothesis postulating that the LDL receptor activity is modulated by
an energy threshold effect of the different saturated fatty acids in the presence of
linoleic acid (18 : 2). Above a threshold of 6.5% energy as 18 : 2, saturated fatty acids
of any kind have minimal effects. Between 3 and 6.5% energy as 18 : 2, myristic
acid (14 : 0) is the only fatty acid to increase LDL-C while below 3% energy as
18 : 2, 14 : 0 is highly hypercholesterolemic and 16 : 0 only moderately so.
These observations have been validated in a recent human study (166) in which
5% energy was exchanged between 16 : 0 and 12 : 0 14 : 0, whereas all other fatty
acids were held constant. Resident male volunteers received diets (30% as fat) on
4-week rotations. Compared with the 12 : 0 14:0-rich diet, the 16:0 diet produced
a significant 9% lower serum cholesterol concentration reflected primarily by a
lower (11%) LDL-C concentration.
In a follow-up study (167), diets enriched by 16:0 (palm olein), 18:1 (rapeseed
oil), or the American Heart Association (AHA) step-one diet were compared by
feeding these diets in rotation to 23 volunteers. TC and LDL-C levels were found
to be unaffected by these diets, despite the exchange of key fatty acids common in
human diets. The AHA diet, however, significantly increased HDL-C while lower-
ing the LDL/HDL cholesterol ratio. There was hardly any difference in the lipid and
lipoprotein concentrations of subjects following consumption of the 16 : 0 and
18 : 1 enriched diets.
These human and animal studies provide strong evidence that the lipemic effects
of the different saturated fatty acids are not equal. 16 : 0 is hypothesized to behave
as a neutral fatty acid (does not raise cholesterol) in normocholesterolemic indivi-
duals (<5.2 mmol/L) and when dietary cholesterol intake is low (<300 mg/day). In
such situations 14 : 0 appears to be the unique cholesterolraising fatty acid. The lack
of 14 : 0 in palm oil and the hypothesized neutrality of 16 : 0 gives credence for the
use of palm oil as a dietary oil suitable for the majority of the worlds populations.
7.5. Nutritional Properties of Minor Components in Palm Oil

Palm oil, both crude and refined, is a rich source of vitamin E, which consists of
a mixture of tocotrienols and tocopherols. A technology for the preparation of a
locotrinol-rich fraction (TRF) from palm fatty acid distillate has been developed.
Palm Vitee (TRF in superolein and encapsulated) has been evaluated in a number
of nutritional studies in both animals and humans. The nutritional properties of TRF
are as follows:
Qureshi (168) first isolated tocotrienols from barley and proved that they could
suppress the hepatic production of cholesterol through their ability to suppress the
activity of the enzyme HMG-CoA reductase, which regulates cholesterol synthesis
in the liver.
Subsequently, Qureshi (169) extended his investigations to TRF (Palm Vitee)
from palm oil in both animal and human models. In a double-blind crossover study
involving 20 hypercholesterolaemic human subjects (serum cholesterol >294 mg/dL),
Palm Vitee supplementation was found to cause a significant drop in serum TC and
LDL-C. The LDL-associated apolipoprotein Apo B was also decreased by 911%.
Moreover, Palm Vitee supplementation resulted in a significant decrease (25%) in
serum thromboxane and platelet factor PF4 by 16%. Similar cholesterol-lowering
effects of Palm Vitee have also been indicated in genetically hypercholesterolemic
swine (170).
In a similar study Tan and co-workers (171) fed volunteers one Palm Vitee cap-
sule per day for 30 consecutive days. Each capsule contained 18 mg tocopherol and
42 mg tocotrienols. In these volunteers. Palm Vitee lowered both serum TC and
LDL-C. The magnitude of reduction for serum cholesterol was up to 36% while
reduction in LDL-C ranged from 0.9 to 37% when compared to their respective
starting values.
416 PALM OIL
These studies indicate that Palm Vitee is most effective in reducing cholesterol
when subjects have elevated blood cholesterol levels. However, these observations
have not been uniformly reproduced by different workers. For example. Wahlqvist and
co-workers (172) have reported that Palm Vitee has no effect on blood cholesterol
levels in their hypercholesterolemic subjects. As a result, it has been suggested that
certain population groups behave as responders and others as nonresponders
when given Palm Vitee to manage their hypercholesterolemia. Studies are presently
in progress to evaluate the underlying mechanisms associated with these observations.
The structural differences between d-a-tocopherol and tocotrienol, viz, the unsa-
turated side chains in the latter, account for some differences in their physiological
activities. Serbinova and co-workers (173) have reported that in membranes palm
oil tocotrienols had 4060 times higher antioxidant potency than a-tocopherol lar-
gely due to a higher recycling efficiency and uniform distribution in membrane
bilayers. Under oxidative stress, tocotrienols protected human LDL against oxida-
tion and their protective potency was greater than that of a-tocopherol in the pre-
sence of ascorbate. This may be a key factor in protection against the onset of
degenerative atherosclerotic disease. A tocopherol-tocotrienol mixture in a ratio
similar to that present in palm oil has also been shown to depress the systolic blood
pressure and increase the aortic production of prostacyclin in spontaneously hyper-
tensive rats (174).
7.6. Effect of Palm Oil on Experimental Carcinogenesis

Sundram and co-workers (175), using a rat model treated with the chemical carci-
nogen DMBA, evaluated the effect of palm oil on the progression of mammary
tumors. Both crude and refined palm oils were evaluated against the polyunsatu-
rated corn oil and soybean oils. Rats fed either 20% soybean oil or corn oil devel-
oped tumors first at 9 weeks following DMBA administration. The appearance of
tumors was more rapid and enhanced in the polyunsaturated corn and soybean oil
fed rats compared to the palm oil groups. At the time of sacrifice tumor incidence
was 90% in the soybean fed rats, 85% in with corn oil, and only 65% with crude
palm oil maintained rats. At the same level of fat intake, tumor incidence in animals
fed polyunsaturated oils was significantly higher than in rats fed palm oil diets.
Tumor yield in the palm oil groups was significantly lower than that in the corn
or soybean oil diet.
In an earlier study Sylvester and co-workers (176) found that palm oil induced
lower tumor numbers and tumor load per rat comparable to a low-fat control (5%
corn oil). In comparison to a palm oil diet (20% by weight), diets containing 20%
corn oil, beef tallow, or lard all resulted in significantly higher tumor numbers and
tumor load.
These studies indicate that palm oil exerts an inhibitory effect on the progression
of chemically induced carcinogenesis.
The 10% level of linoleic acid in palm oil seems ideal in meeting the nutritional
requirements of this essential fatty acid, without eliciting growth responses in the
tumor cells in comparison to more polyunsaturated oils. However, when compared
FUTURE PROSPECTS OF PALM OIL AND MARKET REQUIREMENTS 417
to animal fats such as lard, which contain similar levels of the linoleic acid, it
appears that the inhibitory effect of palm oil cannot be attributed to its fatty acid
composition alone. The answer may lie in the various minor components present in
both crude and refined palm oil. It is postulated that these minor components, more
so the tocotrienols, which have antioxidant activity, may be involved in exerting the
inhibitory effect on tumor development.
The effect of tocotrienols on cancer progression was evaluated by Komi-yama
and Yamaoka (177). The antitumor activity of tocotrienols was evaluated in terms
of the increase in the lifespan of mice inoculated with tumor cells. a-Tocotrienols
and g-tocotrienols were effective against the sarcoma cancer cell lines and Ehrlich
carcinoma. When human lung carcinomas were challenged with these tocotrienols,
a cytotoxic activity due to the tocotrienols was exhibited. Similarly. DMBA-treated
rats responded with lower tumor numbers when their diets were supplemented with
palm tocotrienols (178). Recently, a-carotene isolated from palm oil has been shown to
have antitumor activity against mouse lung cancer and against skin cancer (179).
8. PROSPECTS OF PALM OIL AND MARKET REQUIREMENTS
In the past decade, palm oil has become internationally well known as a vegetable
oil suitable for various applicationsboth edible and nonedible. This is brought
about by it being a versatile oil for the production of various products, with tech-
nical and economic advantages over other oils and fats. Its price competitiveness
and readily available supply is able to serve the needs of oils and fats consumers
worldwide.
8.1. Versatility of Palm Products

Palm oil has the flexibility to be used as it is or in fractionated forms to produce a
very wide range of products. Interesterification can further significantly modify its
properties including crystallization behavior. It has good oxidative stability. It has
long been known as a good heavy-duty frying medium because of its relatively low
polyunsaturation and the slip melting point, which is low enough to avoid excessive
waxiness in most applications (180).
Margarine blends can be developed containing higher levels of palm products
and having solids content profiles close to those of popular commercial brands.
Standard quality table margarine can contain palm products as high as 70%
(50% palm oil and 20% hardened palm oil). Table margarines (packet) can contain
as high as 63% palm oil, 30% palm kernel oil, and 7% palm stearin. Blends for
random interesterification could utilize as much as 60% of hardened palm oil or
palm stearin. Tub margarines could be formulated containing 50% palm oil.
Another major area of use for palm oil due to it being a semisolid fat is in the
manufacture of shortenings and vanaspati. Shortenings include a variety of products
such as the fats used domestically for cooking, frying, and flour confectionery and
those used industrially in cake baking and in large-scale frying operations for pro-
ducts such as potato crisps and doughnuts.
418 PALM OIL
In ambient temperature range of 2035 C there is a close similarity between

butterfat, palm oil, and European shortening. It has, so to speak, been hydrogenated
by nature to be very close to many specifications for shortenings and vegetable ghee.
The world consumption of oils and fats can be classified into two equally impor-
tant categories, i.e., 50% of solid fats and 50% of liquid oil markets. The supply
pattern of vegetable oils and fats is not equally divided between the solid and liquid
oils and fats. Only palm oil is semisolid accounting for 20% of the vegetable oils
market. Under current pattern of consumer demand, there is a shortage in the supply
of solid fats and oversupply in the liquid oil market. Most of the liquid oils have to
be hydrogenated to turn them into solid fats as shortenings, margarines, and ghee.
Palm oil is naturally placed in an advantageous position with respect to the pattern
of large consumer demand for solid fats. With skillful formulation, no hydrogena-
tion is required for palm products. Interesterification methods may be used to
improve the ability of different components of palm oil to be used to meet customer
specifications in the fat products.
Palm oil can also yield liquid fractions for the liquid oil market, i.e., palm olein
and stearin can be obtained through fractionation of palm oil. Palm olein, the liquid
fraction of palm oil, behaves as a liquid oil in hot climates. It is a very stable frying
medium that is comparable to any other frying fat for resistance to breakdown.
However, in temperate climates or during cold nights where temperatures are below
1820 C, palm olein begins to solidify. This problem can be overcome by blending
palm olein with a more unsaturated oil. In Japan, a blend of 50% palm olein and
50% rapeseed oil is being marketed successfully.
Palm stearin, the solid fraction of palm oil, has the edge over tallow because of
the assurance of expanding supply, while world production of tallow has stagnated.
Applications where stearin could replace tallow are in shortenings, frying fats, and
soaps. Lard could also be substituted by palm stearin or RBD palm oil in most of its
applications.
Palm stearin can be the cheapest source of C16C18 fatty acids for soap. Palm
stearin alone has a very high titer value (4750 C) such that when a high proportion
is incorporated into toilet soap formulation, the soap becomes hard and cracks
easily. Experiments conducted indicated that 3050% of palm stearin could be
incorporated with tallow and 20% of lauric (palm kernel) fatty acid to obtain titer
of the finished product of between 40.5 and 44 C (181).
8.2. Technical and Economic Advantages

Hence, the utilization of palm oil in products requiring a proportion of solid fats in
their formulation would offer technical and economic advantages. Substantial cost
savings can be achieved when palm oil is used in place of other oils and fats in
various applications. Savings are made through the lower costs of raw materials,
through the reduction in use of chemicals to process the oil, reduction in the cost
of hydrogenation, reduction in costs due to minimal process losses, and savings
resulting from the long life of palm oil during the frying applications.
The amount of chemicals required to process oils and fats depends on the level
of FFA, color, and other impurities in the oils. Normally about 2 kg of phosphoric
FUTURE PROSPECTS OF PALM OIL AND MARKET REQUIREMENTS 419
acid, 2.5 kg of caustic soda, and 30 kg of bleaching earth are required to refine 1 ton
of crude palm oil. Since processed palm products are available from the markets,
importers need not have to refine crude palm oil but can instead import and use
processed palm products directly. If unrefined vegetable oils were to be imported,
these oils need to be processed by the importers. In addition, liquid oils would
require a catalyst, usually nickel, for hydrogenation. Hence, the cost of chemicals
would be minimal when RBD palm oil is used.
Palm oil, being a semisolid or consistent fat, results in a reduction in cost if uti-
lized in products requiring a proportion of solid fats in their formulation because
hydrogenation is not necessary. If liquid oils are used, they have to be hydrogenated
to obtain the solid consistency, which in turn leads to the formation of both cis and
trans isomers (182). Hydrogen is produced by electrolysis, and with the increasing
cost of energy, hydrogenation is increasingly becoming expensive. To hydrogenate I
ton of liquid oil from an IV of 130 to 70 would require 348kW of electricity. The
additional cost incurred by using liquid oil depends on the electricity rates of the
countries involved.
The availability of semiprocessed to fully processed palm oil products for trade
provides benefits especially to countries that are lacking in refining capacities.
If further refining is still needed, losses incurred would be relatively negligible.
Refining of crude oils results in losses ranging from 4.2 to 5.5%, depending on
whether the oils undergo continuous or batch processes.
Another advantage in the use of imported refined palm oil is the avoidance of
having to deal with effluent treatment. When palm oil is refined at the source,
such as in Malaysia, physical refining is used instead of alkali refining, and the
amount of effluent generated is much lower and can be easily treated.
8.3. Meeting World Oils and Fats Demand

Numerous forecasts on the demand for vegetable oils indicated that the world
would require at least 105 million tons of oils and fats from 83 million tons by
1992. This was equivalent to an additional demand of 2.75 million tons per year.
In the long run, production of by-products from oilseeds and meat is not expected to
rise more sharply than the demand for the main products. Similarly, production of
other oils and fats such as groundnut, sesame, olive, coconut, and fish oils including
butter are increasing at a rate that is below the demand growth for all oils and fats.
Hence, the additional requirement of the world for oils and fats has to be met by the
above average growth of only four oils and fats: sunflower, rape, palm, and palm
kernel oils. However, oils of sunflower and rape are subject to meal demand and the
per hectare returns of competing commodities (grains, pulses, rise, etc.). These thus
leave palm oil to be the main oil to cater for the majority of the additional world
requirements.
8.4. United States Use of Palm Oil

Table 57 gives U.S. supply and disappearance of edible oils (2).
TABLE 57. Edible Fats and Oils: U.S. Supply and Disappearance, 106 lb (2).
Item 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a 2003b
Stocks Octobera
Coconut 277 188 251 164 163 84 150 393 152 136 260 227 148
Corn 138 196 150 118 241 116 129 102 135 267 117 104 114
Cottonseed 137 78 81 106 82 94 66 79 76 49 93 39 40
Lard 24 27 26 34 24 23 20 40 21 18 14 10 5
Palm 53 44 33 35 15 31 46 35 48 48 61 70 42
Palm kernel 53 49 88 73 55 22 51 64 73 49 155 128 50
Peanutc 25 51 50 25 40 65 86 41 40 32 31 32 50
Safflower 28 28 18 31 21 44 27 38 48 36 21 17 19
Soybean 1,786 2,239 1,555 1,103 1,137 2,015 1,520 1,382 1,520 1,993 2,767 2,359 1,486
Sunflower 47 100 56 65 82 147 93 60 121 157 136 23 25
Canola 41 71 67 137 54 77 65 112 169 206 110 52 55
Tallow, edible 41 33 41 36 52 34 48 46 43 40 49 24 35
Imports
Coconut 841 1,163 999 1,100 874 1,188 1,438 791 926 1,115 1,093 860 970
Corn 5 7 7 10 11 14 28 42 18 27 61 65 65
Cottonseed 18 38 26 0 0 0 0 48 8 0 0 22 0
Lard 2 3 3 2 2 1 2 2 2 3 6 10 10
Olive 216 253 262 260 227 304 333 355 397 455 455 485 540
Palm 220 267 368 218 236 322 282 284 345 399 490 425 440
Palm kernel 342 302 304 280 262 392 359 401 393 351 330 470 475
Peanutc 1 0 11 4 5 14 10 73 12 79 39 70 70
Canola 815 861 902 938 1,086 1,075 1,088 1,060 1,139 1,193 1,108 929 1,215
Safflower 22 15 16 26 35 30 51 51 33 34 40 43 45
Soybean 1 10 68 17 95 53 60 83 83 73 46 50 85
Sunflower 9 0 7 1 2 22 8 5 4 8 36 60 5
Tallow, edible 6 10 15 18 8 5 2 3 10 32 7 11 10
Production
Corn 1,821 1,878 1,906 2,227 2,139 2,231 2,335 2,374 2,501 2,403 2,461 2,453 2,650
Cottonseed 1,280 1,126 1,119 1,312 1,229 1,216 1,224 832 939 847 876 725 865
Lard 1,016 1,011 1,015 1,052 1,013 979 1,065 1,106 1,069 1,050 1,080 1,075 1,100
Peanutc 356 286 212 314 321 221 176 145 229 179 230 286 219
Canola 32 49 406 299 355 342 451 548 617 641 585 541 629
Safflower 69 87 111 115 127 103 115 111 91 88 76 89 91
Soybean 14,345 13,778 13,951 15,613 15,240 15,752 18,143 18,078 17,825 18,420 18,898 18,435 17,020
Sunflower 911 730 580 1,165 860 840 959 1,177 1,046 873 673 320 595
Tallow, edible 1,515 1,414 1,535 1,550 1,559 1,407 1,517 1,677 1,792 1,764 1,932 2,075 2,000
Exports
Coconut 22 0 19 18 12 12 6 11 14 8 7 8 10
Corn 566 712 717 865 977 988 1,118 989 970 951 1,172 890 900
Cottonseed 269 184 248 329 221 232 208 111 141 131 150 110 115
Lard 131 129 119 140 94 103 122 140 189 93 90 105 100
Olive 20 15 11 21 24 21 19 15 12 9 10 12 12
Palm kernel 2 9 4 2 2 2 2 2 2 2 2 2 2
Palm 7 7 7 13 20 9 11 11 11 11 10 11 10
Peanutc 151 52 61 97 108 21 13 10 18 14 8 42 19
Canola 15 16 76 153 147 295 349 272 284 187 255 166 157
Safflower 73 65 75 93 122 83 83 92 51 35 37 37 40
Soybean 1,644 1,461 1,531 2,683 992 2,033 3,079 2,372 1,375 1,401 2,519 2,250 850
Sunflower 471 586 450 978 628 709 815 800 630 545 453 110 200
Tallow, edibled 333 306 316 277 241 181 236 322 224 338 475 485 490
Domestic disappearance
Coconut 910 1,084 1,067 1,083 941 1,111 1,189 1,021 927 983 1,119 930 958
Corn 1,202 1,220 1,228 1,250 1,298 1,244 1,271 1,394 1,417 1,630 1,363 1,618 1,804
Cottonseed 1,088 975 873 1,007 996 1,012 1,004 772 833 672 780 636 750
Lard 885 886 890 924 922 880 925 987 886 964 1,000 985 990
Olive 216 253 262 260 227 304 333 355 397 455 455 473 528
Palm 223 271 359 225 201 298 282 260 335 375 471 425 427
Palm kernel 344 254 315 295 293 362 344 390 414 243 355 511 458
Peanut 179 236 187 206 193 194 217 208 233 244 260 296 275
TABLE 57 (Continued )
Item 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a 2003b
Canola 801 898 1,162 1,165 1,271 1,134 1,143 1,287 1,435 1,744 1,496 1,301 1,687
Safflower 15 47 40 57 17 67 73 59 86 102 89 93 95
Soybean 12,248 13,012 12,939 12,913 13,465 14,267 15,262 15,652 16,059 16,318 16,833 17,108 16,522
Sunflower 396 188 129 171 168 207 186 320 385 357 370 268 385
Tallow, edible 1,197 1,109 1,239 1,275 1,345 1,218 1,286 1,360 1,581 1,449 1,488 1,590 1,515
a
b
c
August-July year beginning 1982.
d
Disappearance, as defined by the USDA-ERS, means beginning food stocks, production, and imports minus exports, shipments to U.S. territories, and ending stocks.
REFERENCES 423
REFERENCES
1. FAO/WHO. Fats, Oils and Related Products, Vol. 8, Codex Alimentarius Commission,
Rome, 1993, p. 61.
2. Oil Situation and Outlook Yearbook, Ecconomic Research Service, USDA, accessed
Oct. 2004.
3. W. L. Siew, T. S. Tang, C. H. Flingoh Oh, and A. A. Rahman, PORIM Rep. PO(172)
(1990).
4. H. Kifli, Malaysian Oil Palm Growers Council Technical Bulletin No. 3. 1981.
5. H. P. Sow, Ph.D. Thesis. University Science Malaysia, 1979.
6. B. K. Tan, paper presented at the Conference on Chemical Research in Malaysia, held in
Kuala Lumpur Mar. 1921. 1979, Kuala Lumpur, Malaysian Institute of Chemistry.
7. S. C. Yap, Y. M. Choo, C. K. Ooi, A. S. H. Ong, and S. H. Goh, Elaeis 3(2) (12), 375
(1991).
8. F. Mordret and A. M. Laurent, Rev. Fr. des Corps Gras 25, 245 (1978).
9. Ab. Gapor and co-workers. International Conference on Palm Oil Product Technology in
the Eighties, Kuala Lumpur, June 2224, 1981.
10. M. J. Downes, Leatherhead Res. Rep. Nos. 781, (1982), 411 (1983), 487 and 445 (1984),
518, 516, and 519 (1985).
11. W. L. Siew, private communication, 1982.
12. Anon, Malaysian Standard MS 814: 1983.
13. A. H. G. Chin, C. H. Oh, and W. L. Siew, MARDI Res. Bull. 10(1), 80 (1982).
14. B. K. Tan and Flingoh C. H. Oh, PORIM Tech. (3), (1981).
15. C. H. Oh, Flingoh, private communication, 1982.
16. B. K. Tan and C. H. Oh, PORIM Tech. No. (4), (May 1981).
17. C. W. S. Hartley, The Oil Palm, Longman, London, 1967.
18. A. C. Zeven, Thesis, University of Wageningen, 1967.
19. R. Godding, Bull. Agric. Congo Bedge. 21, 1263 (1930).
20. R. B. Jagoe, Malay. Agric. J. 35, 3 (1952).
21. Anon, Statistics on Commodities. Ministry of Primary Industries, Malaysia, 1991.
22. R. H. V. Corley in R. L. Wastie and D. A. Earp, ed., Advances in Oil Palm Cultivation,
Incorporated Society of Planters, Kuala Lumpur, 1973, pp. 3749.
23. H. A. M. Van der Vossen, Thesis, University of Wageningen (1974).
24. H. L. Foster, M. Tayeb Dolmat, and Gurmit Singh, Proceedings of International Oil
Palm Conference, PORIM, Kuala Lumpur, 1987, pp. 294304.
25. D. Whiting, Market Development of Palm Oil Products, Geneva, International Trade
Center, UNCTAD/GATT, 1982, pp. 108113.
26. K. H. Lim, Proc. of 1987 International O. P. P. O. Conf.-Agriculture, 1992, p. 595.
27. A. N. Ma, PORIM Bull. (4), (May 1982).
28. C. F. Chooi, workshop, Proc. Palm Oil Res. Inst., Malaysia, (9), 5362 (1984).
29. K. H. Lim, workshop, Proc. Palm Oil Res. Inst., Malaysia, (9), p. 1141 (1984).
30. K. H. Lim, S. K. Quah, D. Gillies, and B. I. Wood, workshop, Proc. Palm Oil Res. Inst.
Malaysia, (9), 4252 (1984).
424 PALM OIL
31. S. K. Quah and D. Gillies, Proc. of Nat. Workshop on Palm Oil By-Product Utilization,
1981, pp. 119126.
32. N. S. Chua and H. L. Giam, paper presented at National Workshop on Recent
Development in Palm Oil Milling and Pollution Control Aug. 56, 1986.
33. K. W. Chan, I. A. Watson, and K. C. Lim in E. Pushparajah and S. C. Chin, eds., Soil
Science and Agriculture Development in Malaysia, 1980, pp. 213242.
34. J. Kanagaratnam, A. L. Lai, K. H. Lim, and B. J. Wood, Proc. of Nat. Workshop on Oil
Palm By-Product Utilization, 1981, pp. 1623.
35. C. H. Lim, T. C. Ping, K. W. Chan, and S. Y. Chooi, Proc. of Land Application of Palm
Oil and Rubber Factory Effluent, Malaysia Society of Soil Science, 1983, pp. 7189.
36. K. H. Yeow and Gurmit Singh. MOPGC/TRC Internal Report, 1983.
37. N. S. Chua, paper presented at Seminar on Developments in Palm Oil Milling
Technology and on Developments in Palm Oil Milling and Technology and Environment
Management, May 1617, 1981.
38. H. K. Jorgensen, AOCS, 60(2), 282 (1985).
39. M. S. A. Kheiri, J. Am. Oil Chem. Soc. 62, 210 (1985).
40. L. R. Strecker, J. M. Hasman, and A. Mazer, Proceedings of World Conference in
Emerging Technologies in the Fats and Oils Industries, AOCS, 51, (1985).
41. F. E. Sullivan, J. Am. Oil Chem. Soc. 53, 358 (1976).
42. T. S. Tang, K. T. Teoh, and Y. Y. Lee, Palm Oil Product Technologies in the Eighties,
Incorp. Soc. Planter, Kuala Lumpur, 1983, p. 195.
43. M. S. Fraser and G. Frankl, J. Am. Oil Chem. Soc. 58, 926 (1981).
44. P. A. T. Swoboda, J. Am. Oil Chem. Soc. 62, 287 (1985).
45. K. G. Berger, PORIM Bull., (4), 57 (1982).
46. F. V. K. Young, PORIM Tech. (1), 31 (1981).
47. P. D. Howes, T. C. Soon, S. H. Lim, D. B. Shaw, and P. K. Stemp, Proceedings of PORIM
International Conference on Progress, Prospects and Challenges Towards the 21st
Century. Module III: Chem. & Tech., PORIM, Kuala Lumpur, 1993, p. 55.
48. M. G. A. Wilhems, J. Am. Oil Chem. Soc. 62, 454 (1985).
49. M. Loncin, Fette Seifen Anstrichm 64, 531 (1962); 76, 104 (1974).
50. B. Jacobsberg, PORIM Occasional Paper, No. 12, 11 (1983).
51. P. Sjoberg, Proceedings of PORIM International Oil Palm Conference, Progress and
Prospects Conference II: Technology, PORIM, 1987, p. 219.
52. A. Athanassiadis, Ibid, 440 (1987).
53. M. M. Sein, Proceedings of PORIM International Palm Oil Conference, Progress,
Prospects & Challenges Towards the 21st Century, Module III: Chem. & Tech., 1991,
p. 78.
54. H. L. Gian and N. S. Chua, Proceedings of 1987 International Oil Palm/Palm Oil
Conference: Progress & Prospects (Technology), PORIM/ISP, 1988, p. 177.
55. Hamirin Kifli, PORIM Rep. PO(5a), 83 (1983).
56. M. T. Ab, Gapor and co-workers, Palm Oil Product Technology in the Eighties, PORIM,
Kuala Lumpur, 1983, p. 145.
57. M. T. Ab, Gapor and co-workers Austral. Pat. 618,791 (1993).
58. R. Timms, Chem. Ind., 342 (1991).
REFERENCES 425
59. F. V. K. Young in F. D. Gunstone, Critical Reports on Applied Chem., Vol. 15, John
Wiley & Sons, Inc., 1987, p. 62.
60. S. Wong, T. H. Goh, L. C. Tan, and P. P. Kerk, Proceeding of PORIM International Palm
Oil Conference, Module III: Chem. & Tech., 1993, p. 90.
61. E. Diffense, J. Am. Oil Chem. Soc. 62, 376 (1985).
62. J. B. Rossell, J. Am. Oil Chem. Soc. 62, 385 (1985).
63. D. R. Abigor, F. I. Opute, A. R. Opoku, and A. U. Osagie, J. Sc. Food Agrie. 36, 599
(1985).
64. S. Ravigadevi, C. L. Chong, K. C. Oo, K. H. Yeo, and R. Premavathy, J. Expt. Bot. 42,
1199 (1991).
65. M. Loncin, Oleagineaux 7, 695 (1952).
66. D. A. Okiy, Oleagineaux 33, 625 (1978).
67. K. G. Berger, Oil Palm News 22, 10 (1977).
68. V. K. Lal and A. Gasper, Proceedings of 1991 PORIM International Palm Oil Con-
ference Module III, 1993, p. 243.
69. W. L. Siew, PORIM Bull., (24), 41 (1992).
70. S. B. Ng and A. Southworth, Proceedings of Workshop on Quality in the Palm Oil
Industry, PORIM, Kuala Lumpur, 1985, p. 7.
71. G. N. McDermott, JAOCS, 53, 449 (1976).
72. W. C. Seng, JAOCS, 926A (1980).
73. Donnelly, Wastewater from Edible Oil Refining and Margarine Processing (1989).
74. K. K. Chin and K. K. Wong, Water Res. 15, 1087 (1981).
75. K. K. Chin, W. J. Ng, and A. N. Ma, Biological Wastes 20, 101 (1987).
76. G. Choffel, JAOCS 53, 446 (1976).
77. M. L. Arora, E. F. Barth, and M. B. Umpbres, JPWCF 57, 867 (1980).
78. R. L. Irvine, L. H. Ketchum, Jr., M. L. Arora, and E. F. Barth, JWPCF 59, 847 (1985).
79. R. L. Irvine, G. Miller, and Ajit Singh Bhamrah, JWPCF 51, 224 (1979).
80. R. L. Irvine, D. V. S. Murthy, M. L. Arora, J. L. Copeman, and J. A. Heidman, JWPCF
59, 132 (1987).
81. J. C. Palisand and R. L. Irvine, JWPCF 57, 82 (1985).
82. H. Melcer, W. K. Bedford, B. H. Topnik, and N. W. Schmidtke, JWPCF 59, 19
(1987).
83. K. V. Lo, N. C. R. Bulley, and E. Kwong, Agricul. Wastes 13, 131 (1985).
84. K. V. Lo, J. P. Tam, and P. H. Liao, Biolog. Wastes 25, 193 (1988).
85. W. J. Ng, Biolog. Wastes 22, 285 (1987).
86. K. C. Okkyung and Y. Pomerang, J. Am. Oil Chem. Soc. 60, 1848 (1983).
87. N. A. Idris, PORIM Information Series No. 9, Palm Oil Research Institute of Malaysia
(1992).
88. N. A. Idris, K. G. Berger, and A. S. H. Ong, J. Sci. Food Agric. 46, 481 (1989).
89. N. A. Idris, Ph.D. Thesis, Univeriti Kebangsaan Malaysia, 1991.
90. N. A. Idris, A. Abdullah, M. S. Embong, and A. H. Hassan, Elaeis 3, 302 (1991).
91. N. A. Idris, A. Abdullah, and M. S. Embong, ASEAN Food J. 7, 202 (1992).
426 PALM OIL
92. K. G. Berger, PORIM Occasional Paper No. 2, Palm Oil Research Institute of Malaysia
(1981).
93. J. Ward, J. Am. Oil Chem. Soc. 65, 1731 (1988).
94. Y. K. Teah and A. S. H. Ong, Palm Oil Develop., (4), 1718 (1986).
95. M. S. A. Kheiri, PORIM Occasional Paper No. 5, Palm Oil Research Institute of
Malaysia (1982).
96. H. Von Zeddelman and J. Wurziger, Fette, Seifen Anstrichminel 75, 18 (1973).
97. L. Faur, Rev. Frane. des Corps Gras 78, 77 (1975).
98. B. Toregard and A. Eriksson, Proceedings of 10th Scandinavian Symposium on Lipids,
Nyborg, Denmark Sec. A. 1979, p. 54.
99. M. E. Herendi and R. Bethke, Susswaren 12, 410 (1982).
100. A. M. T. Gapor, Palm Oil Develop. No. 12, 2527, (1990).
101. S. Mashashi, T. Yoshikazu, and S. Masanori, J. Am. Oil Chem. Soc. 62, 449 (1985).
102. M. A. Augustine, L. K. Heng, and I. Nor Aini, Pertanika 11, 393 (1988).
103. M. A. Augustine, L. K. Heng, and I. Nor Aini, Pertanika 12, 167 (1989).
104. S. Torrey, Edible Oils and FatsDevelopments Since 1978, Royes Data Corp., Park
Ridge, N. J., 1983.
105. J. A. Wolfe, Manufacturing Confectioner, 53 (Apr. 1977).
106. H. Traither and A. Dieffenbacher, J. Am. Oil Chem. Soc. 62, 417 (1985).
107. I. Nor Aini, Palmsearch Circular No. 2, 8792, (1990).
108. I. Nor Aini, H. Hanirah, F. C. H. Oh, and N. Sudin, J. Am. Oil Chem. Soc. 69, 1206
(1992).
109. I. Nor Aini, H. Hanirah, N. Sudin, F. C. H. Oh, and T. S. Tang, JAOCS (1994).
110. H. J. Richtler and J. Knaut, Seifen-Ole-Fene-Waschse 117(15), 545 (1991).
111. H. Kifli and S. Krishnan, Proceedings of the PORIM International Oil Palm/Palm Oil
Conference: Progress and Prospects, PORIM, Kuala Lumpur, 1987, pp. 304315.
112. A. Kuntom and H. Kifli, Proceedings of the PORIM International Oil Palm/Palm Oil
Conferences: Progress and Prospects. PORIM, Kuala Lumpur, 1987, pp. 328331.
113. A. Kuntom and H. Kifli, in T. H. Applewhite, ed., Proceedings of the World Conference
on Oleochemicals into the 21st Century, American Oil Chemists Society Champaign,
III., 1990, pp. 298304.
114. S. Ahmad and H. Kifli, in A. R. Baldwin, ed., Proceedings of the Second World
Conference on Detergents: Looking Toward 90s. American Oil Chemists Society,
Champaign, III., 1987, pp. 250256.
115. T. L. Ooi, A. Kuntum, W. L. Siew, Y. A. Tan, and B. Muhammad, in Proceedings of the
1993 PORIM International Palm Oil Conference, PORIM, Kuala Lumpur, 1993.
116. G. Dieckelmann and H. J. Heinz, The Basics of Industrial Oleochemistry, Oleochemicals
Consulting International, D-4330 Mulheim (Ruhr)-Saarn, 1988.
117. Y. Basiron and A. Hitam, PORIM Information Series No. 4, 1992.
118. S. Ahmad, Y. C. Leong, I. Mohammad, and E. L. Ong, Proceedings of the PORIM
International Oil Palm/Palm Oil Conferences: Progress and Prospects. PORIM, Kuala
Lumpur, 1987, pp. 291303.
119. S. Ahmad, A. Rafiei, and Z. Ismail, PORIM Rep. PO (125a) 87 General (1987).
REFERENCES 427
120. S. Ahmad, Y. C. Leong, C. S. Wong, and A. Rafiei, PORIM Rep. PO (134B)88 General
(1988).
121. H. Hassan, S. Ahmad, and A. Ibrahim, PORIM Information Series No. 6 (1992).
122. H. Kluth and A. Meffert, U.S. Pat. 4,508,853, (Apr. 2, 1985).
123. H. Kluth, B. Gruber, A. Meffert, and W. Huebner, U.S. Pat. 4,742,087, (May 3, 1988).
124. H. Hassan, S. K. Yeong, and S. Ahmad, Proceedings of the PORIM International Palm
Oil Conference, PORIM, Kuala Lumpur, 1993.
125. H. M. Nor, M. H. Mahmood, H. Kifli, and M. A. Rahman, in T. H. Applewhite, ed.,
Proceedings of the World Conference on Oleochemicals into the 21st Century, American
Oil Chemists Society, Champaign, III., 1990, pp. 311314.
126. H. M. Nor, M. H. Mahmood, H. Kifli, M. A. Rahman, and A. Rafiei A, Nuc. Sci. J.
Malaysia 8, 149 (1990).
127. M. H. Mohammad, H. Nor, H. Kifli, M. A. Rahman, and A. Rafiei, Nuc. Sci. J. Malaysia
9(2), 95 (1991).
128. H. Kifli and S. Ahmad, PORIM Rep. (1985).
129. T. L. Ooi and A. S. H. Ong, Proceedings of the PORIM International Oil Palm/Palm Oil
Conferences: Progress and Prospects. PORIM, Kuala Lumpur, 1987, pp. 322327.
130. A. Kuntom, H. Kifli, and C. H. Oh, Transparent Soaps Derived from Palm Oil, ABISA
Congress, Olindo, Brazil, 1992.
131. A. Kuntom, unpublished results, Palm Oil Research Institute of Malaysia, Kuala
Lumpur.
132. I. Mohamed, E. L. Ong, H. Kifli, and M. A. Rahman, PORIM Rep. 1989.
133. F. Timmermann, Medium Chain TriglyceridesThe Unconventional Oil, Chemicals
Fabrik Grunau Gmbh, Germany, 1991.
134. R. M. Speed, Proceedings of 1989 PORIM International Palm Oil Development
Conference, PORIM, Kuala Lumpur, 1989, pp. 127129.
135. M. A. Margeson and J. S. Schwartz, Proceedings of 1991 PORIM International Palm Oil
Conference, PORIM, Kuala Lumpur, 1991, pp. 193198.
136. T. Ogoshi and Y. Miyawaki, J. Am. Oil Chem. Soc. 62(2), 331 (1985).
137. T. C. Inagaki, in T. H. Applewhite, ed., Proceedings of the World Conference on
Oleochemicals into the 21st Centuries, American Oil Chemists Society, Champaign,
III., 1990, pp. 269278.
138. S. Ahmad, Proceedings of the PORIM International Palm Oil Conference, PORIM,
Kuala Lumpur, 1993.
139. A. S. H. Ong, A. H. Hassan, Y. Basiron, Y. M. Choo, and A. Gapor, Proceedings of the
Symposium on New Developments in Palm Oil, 1989.
140. A. N. Ma, and co-workers, Proceedings of the PORIM International Palm Oil
Conference, PORIM, Kuala Lumpur, 1993.
141. K. Y. Cheah and co-workers, J. Am. Oil Chem. Soc. 69(5), (1992).
142. G. Eisenbrand, M. Blankart, H. Sommer, and B. Weber, N-Nitroalkanolamines in
Cosmetics, IARC Science Publication, 1991, pp. 238241.
143. M.C. Chow and C. C. Ho, Proceedings of the PORIM International Palm Oil
428 PALM OIL
144. C. H. Teoh and C. S. Chia, paper presented at 1993 PIPOC, Sept. 2025, 1993, Kuala
Lumpur, Malaysia.
145. N. MacFarlene, A. A. Swetman, and D. G. Coursey, Proceedings of Second African Palm
Oil Processing Workshop, Nigeria (1983), FAO, Rome, 1984.
146. P. M. Kris-Etherton, C. Y. Ho, and M. A. Fosmire, J. Nutr. 114, 675 (1984).
147. M. Sugano, T. Ishida, and K. Koba, J. Nutr. 118, 548 (1988).
148. K. Sundram, H. T. Khor, and A. S. H. Ong, Lipids 25, 187 (1990).
149. D. Klurfeld, Proceedings PORIM International Development Conference, PORIM,
Kuala Lumpur, 1989.
150. A. S. H. Ong and co-workers, FASEB 2: Abst. 7228 (1988).
151. S. Lindsey, J. Benattar, A. Pronczuk, and K. C. Hayes, Proc. Soc. Biol. Med. 195, 261
(1990).
152. F. M. Mattson and S. M. Grundy, J. Lipid Res. 26, 194 (1985).
153. C. W. Laine and co-workers, Am. J. Clin. Nutr. 35, 683 (1982).
154. Anon, Nutr. Rev. 45, 205 (1987).
155. K. Sundram, G. Hornstra, A. C. Houwelingen, and A. D. M. Kester, Br. J. Nutr. 68, 677
(1992).
156. A. Maozuki, F. Arshad, T. A. Razak, and K. Jaarin, Am. J. Clin. Nutr. 53, 1010S (1991).
157. K. W. Ng and co-workers, Am. J. Clin. Nutr. 53, 1015S (1991).
158. K. W. Ng and co-workers, J. Am. Coll. Nutr. 11, 383 (1992).
159. R. Wood, K. Kubena, R. Crook, G. Martin, and S. Tseng, J. Nutr. Biochem. 4, 1 (1993).
160. D. Heber and co-workers, Nutr. Res. 12, S53 (1992).
161. A. S. Truswell, N. Choudhury, and D. C. K. Roberts, Nutr. Res. 12, S43 (1992).
162. A. Bonanome and S. M. Grundy, N. Engl. J. Med. 318, 1244 (1988).
163. D. M. Hegsted, R. B. McGandy, M. L. Meyers, and F. J. Stare, Am. J. Clin. Nutr. 27, 281
(1965).
164. K. C. Hayes, A. Pronczuk, S. Lindsey, and D. Diersen Schade, Am. J. Clin. Nutr. 53, 491
(1991).
165. K. C. Hayes and P. R. Khosla, FASEB J. 6, 2600 (1992).
166. K. Sundram, K. C. Hayes, and O. H. Siru, Am. J. Clin. Nutr. 59, (1994).
167. K. Sundram. K. C. Hayes, and O. H. Siru, (1994).
168. A. A. Qureshi, W. C. Burger, D. M. Peterson, and C. E. Elson, J. Biol. Chem. 261, 10544
(1986).
169. A. A. Qureshi and co-workers, Am. J. Clin. Nutr. 53, 1021S (1991).
170. D. T. S. Tan and H. T. Khor, Am. J. Clin. Nutr. 53, 1027S (1991).
171. A. A. Qereshi and co-workers. Proceedings PORIM International Development
172. M. L. Wahlqvist and co-workers, Nutr. Res. 12, S181 (1992).
173. E. Serbinova, V. Kagan, D. Han, and L. Packer, Free Rad. Biol. Med. 10, 263 (1991).
174. K. Kazunori, A. Khoji, K Ikuo, and M. Sugano, Biosci. Biotech. Biochem. 56, 1420
(1992).
175. K. Sundram, H. T. Khor, A. S. H. Ong, and R. Pathmanathan, Cancer Res. 49, 1447
(1989).
REFERENCES 429
176. P. W. Sylvester, M. Russel, M. M. Ip, and C. Ip, Cancer Res. 46, 757 (1986).
177. K. Komiyama and M. Yamaoka, Chem. Pharm. Bull. 37, 1369 (1989).
178. K. Nesaretnam and co-workers, Nutr. Res. 12, 63 (1992).
179. M. Murakoshi and co-workers, Cancer Res. 52, 6583 (1992).
180. J. Podmore, in R. J. Hamilton and A. Bhati, eds., Application of Modification Techniques
in Recent Advances in Chemistry and Technology of Fats and Oils, 1987.
181. H. Kifli and S. Krishnan, Proceedings of the 1987 International Oil Palm/Palm Oil
Conference, Progress and Prospects. 29th June1st July, 1987.
182. E. Bernardini, Oilseeds, Oils and Fats, Vol. II, Oils and Fats Processing, 2nd ed., B.E. Oil
Publishing House, Rome.
9
Peanut Oil
Harold E. Pattee
North Carolina State University
Raleigh, North Carolina
1. PEANUT ORIGIN AND HISTORY
In 1753, Linneaus described the domesticated peanut species as Arachis (derived

from the Greek arachis, meaning a weed) hypogaea (meaning a underground
chamber) or a weed with fruit produced below the soil. The domesticated peanut
(A. hypogaea) is believed to have originated in an area covered by southern Bolivia
and northern Argentina because of the primitive characteristics associated with the
germplasm from this region (1). Subspecies hypogaea var. hypogaea is the pre-
dominant peanut type found in this area, and Krapovickas (2) hypothesized that
the var. hypogaea may represent the most ancient cultivars because they have the
runner habit, branching patterns, similar to related Arachis species, and no floral
compound spikes.
Additional information now suggests that a second origination event in the area
north of Lima on the west coast of Peru could have been involved in the evolution
of A. hypogaea. Archeological excavations near Casma at Pampa de la Llamas-
Moxeke have recovered peanut shells at a level dated to be approximately
1500 B.C. (35), and gold carvings found in ancient tombs just to the north of Pampa
de las Llamas-Moxeke (6, 7) closely resemble the reticulation of the cultivated
types now grown in the Casma area.
431
432 PEANUT OIL
As peanut is native to South America, the early Spanish and Portuguese

explorers found the Indians cultivating the peanut along with other food crops. It
was from the tropical and subtropical areas of this hemisphere that the peanut
was disseminated to Europe, to both the coasts of Africa, to Asia, and to the Pacific
Islands (8). The Incas of Peru, who achieved one of the worlds most highly devel-
oped agricultural civilizations, cultivated the peanut throughout the long coastal
regions of Peru. Garcilaso de la Vega describes the peanut as another vegetable
which is raised under the ground, called by the Indians ynchic. It is very like mar-
row, and has the taste of almonds. Of its food and medicinal uses: If the ynchic is
eaten raw it caused headache, but when toasted it is wholesome, and very good with
treacle; and they make an excellent sweetmeat from it. They also obtain an oil from
the ynchic, which is good for many diseases (8). Just when the peanut was first
purposefully introduced into Europe and into the colonial seaboard of the south-
eastern United States is not documented. However, from this introduction around
the world, the peanut has become a significant agricultural commodity and its oil
a primary ingredient in the culinary process in many countries.
2. GLOBAL
2.1. Peanut Production

The peanut is known by several names throughout the world, such as groundnut and
earth nut, because the seeds develop under the ground. Peanuts are produced on a
significant basis in more than 30 different countries throughout the world. The
worldwide production for 2002 was estimated to be in excess of 31 million metric
tons (MMT) (9). India, China, and the United States were the three largest produ-
cers of peanuts and accounted for over 70% of the world production in 2002. Peanut
production worldwide has undergone significant increases in the last 30 years
(Table 1). In 1972, the average production was 14.4 MMT, 1980 16.0 MMT,
1990 21.6 MMT, and 20002002 32.0 MMT. (9, 10). Some of the production
increase was a result of a 22% increase in the area harvested between 1972 and
2002. However, the major factor was the increase in yield from 0.93 MT/ha in
the 1970s to 1.4 MT/ha in 2001/2002. Among the three major producers, India
had a 42% increase in production between the 1970s and 1990s but decreased
16% between the 1990s and 2002. China increased production 179% between
the 1970s and 1990s and another 136% to 2002. The U.S. production increased
15% in the first 20 years and has remained near 1.9 MMT since the 1990s
(9, 10). Total area harvested and production levels in the next tier of eight countries
has averaged approximately 4.63 mha harvested and 4.46 MMT produced across
the 30 years. From the data given in Table 1, the high total area harvested for
this tier of eight countries was 5.38 mha in 1972 with a low of 3.93 mha in
1990. Highest production occurred in 2000 at 5.50 MMT and the lowest in 1990
at 3.46 MMT (Table 1). World exports of peanuts from the producing countries
have only increased 22%, from 1178 MMT in 1972 to 1518 MMT in 2002.
TABLE 1. Major Countries and World Peanut Production and Utilization (mha or MMT) Across 30 Years.1
Domestic Consumption
Area Total Total
Year Country Harvested Production Supply Exports Crushed Food Feed; Seed; Waste Total Distribution
1972 China 1878 2092 2092 42 1018 768 264 2050 2092
1972 India 6990 4092 4342 33 3511 532 266 4309 4342
1972 United States 601 1485 1663 236 386 768 78 1232 1663
1972 Argentina 370 440 457 2 331 46 21 398 457
1972 Brazil 506 590 590 78 401 59 52 512 590
1972 Burma 633 390 390 0 228 142 20 390 390
1972 Indonesia 407 483 483 29 91 340 23 454 483
1972 Nigeria 1220 772 797 284 383 100 30 513 797
1972 Senegal 1100 540 540 7 385 45 103 533 540
1972 Sudan 690 568 568 156 91 210 111 412 568
1972 Zaire 451 230 230 0 81 137 12 230 230
1972 World 18121 14421 16263 1178 8569 4873 1270 14712 16263
1980 China 2339 3600 3600 305 1667 1257 371 3295 3600
1980 India 6801 5005 5205 71 4059 325 650 5034 5205
1980 United States 566 1045 1512 228 202 663 232 1097 1512
1980 Argentina 197 243 282 74 147 12 11 170 282
1980 Brazil 235 310 312 37 196 37 42 275 312
1980 Burma 514 431 431 0 319 91 21 431 431
1980 Indonesia 508 791 806 2 47 682 75 804 806
1980 Nigeria 650 530 530 0 204 220 106 530 530
1980 Senegal 1064 521 521 3 258 101 159 518 521
1980 Sudan 894 707 707 133 377 151 46 574 707
1980 Zaire 480 320 320 0 107 185 28 320 320
1980 World 17508 16040 17805 1113 8507 5697 1989 16193 17805
Area Total Total
Year Country Harvested Production Supply Exports Crushed Food Feed; Seed; Waste Total Distribution
1990 China 2907 6368 6369 448 3250 2209 462 5921 6369
1990 India 8309 7514 7514 45 5999 490 980 7469 7514
1990 United States 732 1634 1964 296 313 916 129 1358 1964
1990 Argentina 198 311 341 110 123 31 30 184 341
1990 Brazil 95 157 172 2 50 95 20 165 172
1990 Burma 554 472 472 10 320 96 46 462 472
1990 Indonesia 600 860 1011 0 45 850 93 988 1011
1990 Nigeria 500 250 260 0 80 100 60 240 260
1990 Senegal 914 703 758 4 480 146 93 719 758
1990 Sudan 540 325 325 20 145 135 25 305 325
1990 Zaire 530 380 380 0 129 229 22 380 380
1990 World 19089 21656 23498 1304 11705 7791 2157 21653 23498
2000 China 4856 14437 14437 450 6800 6047 1140 13987 14437
2000 India 8100 5700 5700 100 4300 500 800 5600 5700
2000 United States 541 1481 2138 239 248 988 166 1402 2138
2000 Argentina 251 395 409 177 142 21 19 182 409
2000 Brazil 102 196 216 3 60 125 18 203 216
2000 Burma 530 640 640 12 390 162 76 628 640
2000 Indonesia 650 1040 1178 0 64 1030 70 1164 1178
2000 Nigeria 1210 1470 1475 0 510 670 290 1470 1475
2000 Senegal 1030 1003 1028 9 420 395 149 964 1028
2000 Sudan 550 370 370 5 210 135 20 365 370
2000 Zaire 491 382 382 0 120 232 30 382 382
2000 World 22644 31120 33225 1387 14174 13886 3021 31081 33225
2002 China 5000 14500 14500 500 6950 5950 1100 14000 14500
2002 India 8100 6700 6700 105 5060 600 935 6595 6700
2002 United States 551 1702 2395 293 292 1090 167 1549 2395
2002 Argentina 200 315 335 160 125 21 19 165 335
2002 Brazil 100 195 215 3 60 124 18 202 215
2002 Burma 530 640 640 12 390 162 76 628 640
2002 Indonesia 650 1000 1219 0 62 1075 68 1205 1219
2002 Nigeria 1230 1510 1515 0 528 685 297 1510 1515
2002 Senegal 750 500 523 5 160 252 85 497 523
2002 Sudan 550 370 370 4 211 135 20 366 370
2002 Zaire 500 390 390 0 132 233 25 390 390
2002 World 22507 31837 34155 1518 14901 13971 3056 31928 34155
1
Data extracted from http://www.fas.usda.gov/psd/complete_files/OIL-2221000.csv
436 PEANUT OIL
However, China alone has increased exports from 42 MMT to 500 MMT during this
same period (Table 1). This increase accounts for nearly one-third of the total world
peanut exports. On the other hand, the African continent countries of Nigeria, Sene-
gal, and Sudan have had a decrease in the exporting of peanuts from a combined
447 MMT in 1972 to 9 MMT in 2002.
2.2. Peanut Utilization

Peanuts are not a crop that can easily be carried over from one year to the next as
noted by a comparison of the production and total consumption values across years
(Table 1). Utilization of the peanut crop can be classified into the general areas of
crushed, food, feed, seed, and waste. In 1972, the primary utilization in 7 of the 11
listed countries was the crushing of peanuts for oil and utilization of the resultant
meal. The United States food utilization was more than twice that of the next coun-
try, Indonesia. In the United States, food utilization was nearly twice that of crush-
ing utilization. In 2002, the number of countries having food as the primary
utilization factor had increased to six and the United States and Indonesia were
almost equal in food utilization (Table 1).
2.2.1. Oil Hammons (8) review of the origin and early history of the peanut pro-
vides extensive insight into writings of the early Spanish and Portuguese explorers
and the usage of peanuts as an oil source for many purposes. Spanish recognition of
the usefulness of peanut oil is documented by the establishment of an oil mill at
the Mediterranean port of Valencia around 1800 (11). Most authorities credit the
Portuguese with introducing the peanut into African agriculture from Brazil.
West Africa was the primary source of peanut exportation in the nineteenth century.
Brooks (12) provides an overview of the development of the peanut industry in
Africa and peanut exportation from West Africa to other parts of the world. The
first export seems to have been from Gambia to Britain in 1834 involving 213 bas-
kets, but the next year export increased to 47 tons and by the 1840s involved thou-
sands of tons a year. Earliest exports to America were in 1835, and exports to
America dominated the Gambian market from 1837 to 1841. The exportation to
Britain was for crushing, and the dominant reason for American usage was the
pleasing flavor of the roasted peanut.
Development of the European peanut oil industry was stimulated by a worldwide
shortage of fats after the Napoleonic wars, an increase in population, a rise in the
standard of living, and a new working class. As in Britain, French soap and candle-
makers became increasingly dependent on foreign sources of oil supply in the
1830s. Learning of the British peanut imports, French industrialists undertook
experimentation of their own on peanut oil. Jaubert, a Goree trader who had sent
a sample of peanut oil to Marseille in 1833, is credited with initiating the industry
with a shipment of 722 kg of peanuts from West Africa to Marseilles in 1840, when
France reduced the tariff on peanuts (8, 12). Following that shipment, other traders
are reported in 1842 to have brought nearly a 1000 tons of peanuts to Marseilles.
GLOBAL 437
Peanut oil production continued to increase, in Europe, throughout the nineteenth

century. By 1899, 17 factories at Marseilles were crushing about 200,000 tons. An
equal volume was being processed in Britain and other European countries (13).
France continued to be a major peanut importer and oil producer, through the
mid-1970s with 331 MMT crushed in 1972. However, by 1980 and 2000, the
crushed level had dropped to 79 and 8 MMT, respectively (10).
Across the last 30 years, the amount of peanuts crushed for oil worldwide has
increased from 7957 to 14,901 MT. (Table 2). Increases in metric tons crushed in
China and India and the decreases in the South America countries of Argentina and
Brazil account for almost 100% of the changes. The oil produced is virtually all
used within the countries of production. It seems appropriate to note that within
Japan, the industrial use of peanut oil has increased from 4 to 14 MMT between
1990 and 2002. Exporting of peanut oil has decreased nearly 42% from 1972 to
2002. Of the 252 MMT of oil exported worldwide in 2002, four countries, Argen-
tina, Nigeria, Senegal, and Sudan, account for nearly 70%.
2.2.1.1. Oil Extraction Hydraulic pressing, expeller, and/or solvent extraction are
the three general methods for extracting oil from the seed. When hydraulic pressing
is used, it is followed by hot solvent extraction for nearly total recovery of the oil.
Expeller extraction relies on friction and pressure within the expeller, which causes
the meal to heat, thus facilitating the oil extraction process. This process removes
approximately 50% of the peanut oil. The remaining oil is extracted using hexane,
which is later removed through an evaporationcondensation system. Solvent
extraction involves petroleum hydrocarbons or other solvents. Solvent extraction
is accomplished in closed systems where oil is removed and solvent reclaimed
for reuse. The efficiency of extraction with hexane, 95% ethanol, or absolute etha-
nol on peanut grits has been reported (14). Extracted oil is refined by deacidification
with sodium hydroxide to neutralize the free-fatty acids, washing with water at
about 82 C to remove the sodium hydroxide, and then bleaching with bleaching
clay at about 100 C under reduced pressure. The refined oil is then deodorized
by heating under vacuum and blowing superheated steam through the oil. Deacidi-
fication and deodorization of peanut oil and other edible oils by dense carbon diox-
ide extraction has been investigated (15). The purpose of the refining process is to
remove nontriacylglycerol components, including free fatty acids, nonhydratable
phosphoacylglycerols, sterols, pigments, glucosides, waxes, hydrocarbons, and
other compounds that may be detrimental to the flavor or oxidative stability of
the refined oil (16).
2.2.1.2. Alternative Oil Extraction Techniques and Seed Treatment The com-
plete removal of organic solvents used for extracting seed oils is mandatory if
the oil is to be used for human consumption. Supercritical fluid extraction
has emerged as an attractive separation technique because it does not introduce
any residual organic chemicals. Supercritical CO2 is the most commonly used
supercritical fluid (17). CO2 is relatively low cost, nonflammable, nontoxic, and
TABLE 2. Major Countries and World Peanut Oil Production and Utilization (MMT) Across 30 Years.1
Oil On Total Total
Year Country Crushed Production Hand Imports Supply Food Total Exports Distribution
1972 China 1018 254 0 0 254 234 234 20 254

1972 India 3511 1060 0 0 1060 1060 1060 0 1060
1972 United States 386 122 15 0 137 72 72 48 137
1972 Argentina 331 78 2 0 80 0 0 80 80
1972 Brazil 401 112 0 0 112 68 68 44 112
1972 Burma 228 73 0 0 73 73 73 0 73
1972 Indonesia 91 29 0 0 29 29 29 0 29
1972 Nigeria 383 122 0 0 122 11 11 111 122
1972 Senegal 385 128 0 0 128 65 65 63 128
1972 Sudan 91 29 0 0 29 29 29 0 29
1972 Zaire 81 26 0 0 26 22 26 0 26
1972 World 7957 2371 17 401 2789 2325 2337 434 2789
1980 China 1667 417 0 0 417 368 368 49 417

1980 India 4059 1177 0 0 1177 1177 1177 0 1177
1980 United States 202 63 20 0 83 44 44 22 83
1980 Argentina 147 42 0 0 42 0 0 36 42
1980 Brazil 196 62 0 0 62 16 16 46 62
1980 Burma 319 102 0 0 102 102 102 0 102
1980 Indonesia 47 16 0 0 16 16 16 0 16
1980 Nigeria 204 65 0 4 69 69 69 0 69
1980 Senegal 258 83 0 0 83 63 63 20 83
1980 Sudan 377 121 0 0 121 105 105 16 121
1980 Zaire 107 34 0 0 34 33 34 0 34
1980 World 8085 2343 60 319 2722 2410 2411 268 2722
1990 China 3250 813 0 5 818 772 772 46 818

1990 India 5999 1740 0 0 1740 1736 1740 0 1740
1990 United States 313 97 10 5 112 90 90 11 112
1990 Argentina 123 40 0 0 40 5 5 35 40
1990 Brazil 50 14 8 15 37 15 15 18 37
1990 Burma 320 99 0 0 99 99 99 0 99
1990 Indonesia 45 14 5 0 19 13 13 0 19
1990 Nigeria 80 37 0 0 37 37 37 0 37
1990 Senegal 480 153 9 0 162 53 58 99 162
1990 Sudan 145 47 0 0 47 44 44 3 47
1990 Zaire 129 41 0 0 41 40 41 0 41
1990 World 11389 3242 54 302 3598 3274 3292 259 3598
2000 China 6800 2115 0 10 2125 2110 2110 15 2125

2000 India 4300 1245 0 0 1245 1235 1245 0 1245
2000 United States 248 81 14 36 131 111 111 6 131
2000 Argentina 142 42 0 0 42 1 1 41 42
2000 Brazil 60 16 2 0 18 17 17 1 18
2000 Burma 390 123 0 0 123 123 123 0 123
2000 Indonesia 64 20 0 0 20 20 20 0 20
2000 Nigeria 510 230 0 0 230 195 195 35 230
2000 Senegal 420 160 6 0 166 58 58 102 166
2000 Sudan 210 67 0 0 67 22 22 45 67
2000 Zaire 120 38 0 0 38 37 38 0 38
2000 World 14149 4301 32 258 4591 4239 4250 312 4591
2002 China 6950 2175 0 10 2185 2170 2170 15 2185

2002 India 5060 1465 0 0 1465 1451 1465 0 1465
2002 United States 292 93 14 20 127 111 111 5 127
2002 Argentina 125 39 0 0 39 1 1 38 39
2002 Brazil 60 16 0 0 16 15 15 1 16
2002 Burma 390 123 0 0 123 123 123 0 123
2002 Indonesia 62 19 0 0 19 19 19 0 19
Oil On Total Total
Year Country Crushed Production Hand Imports Supply Food Total Exports Distribution
2002 Nigeria 528 238 0 0 238 208 208 30 238

2002 Senegal 160 58 5 10 73 10 10 60 73
2002 Sudan 211 68 0 0 68 24 24 44 68
2002 Zaire 132 41 0 0 41 40 41 0 41
2002 World 14901 4513 30 219 4762 4476 4491 252 4762s
1
GLOBAL 441
easily removed from the oil product by depressurization. However, particle size
does have a significant effect on the extraction rate curves (1820). CO2 is also
U.S. Food and Drug Administration approved and is generally regarded as a safe
compound.
Food-grade butane in a supercritical, low-pressure, liquefied gas extraction pro-
cedure has also been described for oil extraction from peanuts (21). The extraction
process consists of mixing the liquefied butane with the material to form a slurry.
The liquefied gas and oil are moved to a solvent recovery system where the oil is
removed from the butane. The oil is pumped from the solvent recovery system to a
holding tank, and the butane is then transformed into a gas in the solvent recovery
system and transported back to the butane storage tank for reuse.
Aqueous enzymatic oil extraction is another ecofriendly extraction procedure.
It is based on simultaneous isolation of oil and protein from oilseed by dispersing
finely ground seed in water and separating the dispersion by centrifugation into oil,
solid, and aqueous phases. The presence of certain enzymes during extraction
enhances oil recovery by breaking cell walls and oil bodies (22). For peanuts, a
multistep aqueous extraction process has been described with a recovery of
about 98% (23). More recently, the relatively new technique of enzyme-assisted
aqueous extraction has been applied to peanuts with a reported oil recovery of
8692% (24).
Microwave treatment, because of its rapid heating of materials, is being explored
in a multitude of crops for enzyme inactivation (2528), for extraction of natural
products (29), and oil and fat extraction from seeds and food products (3032).
Microwave treatment of peanut seed prior to press extraction increased oil recovery
approximately 10% at an optimum treatment time of 30 seconds (30). However,
free fatty acid content initially increased with exposure time as well as peroxide
value (30). Research on use of microwave treatment in blanching of peanuts indi-
cated an influence on oil stability depending on treatment conditions (33).
2.2.1.3. Oil Extraction By-Product The byproduct of peanut oil production is

peanut meal, and depending on the methods used, the oil content remaining in
the meal range from about 7% to 1%. Human consumption of peanut meal is neg-
ligible except in India and Argentina (Table 3). The primary use of peanut meal is
animal feed. When peanut meal is used for human or animal consumption, careful
consideration should be given to the quality of the meal. Various oilseeds, edible nuts,
grains, and their derived products are subject to mycotoxin contamination (34),
and these mycotoxins may have a detrimental effect on both human and animal
health (35). Worldwide regulations for mycotoxins have been published (36).
Mycotoxins are generally associated with the protein fraction and are not found
in refined oil because of the processing procedures. Unrefined or lightly refined
oil may contain mycotoxins because of the fine residue particles contained therein.
Meal from edible-grade peanuts with low oil content may be processed into flour
for human consumption. When poor-quality grades are used, poor extraction effi-
ciencies or lack of hygienic conditions exist, and the residue should be used as a
fertilizer.
TABLE 3. Major Countries and World Peanut Meal Production and Utilization (MMT) Across 30 Years.1
On Total Total
Year Country Crushed Hand Production Imports Supply Exports Food Feed: Waste Total Distribution
1972 China 1018 0 407 0 407 0 0 366 407 407

1972 India 3511 0 1373 0 1373 869 0 504 504 1373
1972 United States 386 1 163 0 164 0 0 161 161 164
1972 Argentina 331 5 136 0 141 85 0 44 44 141
1972 Brazil 401 0 154 0 154 80 0 74 74 154
1972 Burma 228 0 88 0 88 0 0 88 88 88
1972 Indonesia 91 0 35 0 35 0 0 35 35 35
1972 Nigeria 383 6 147 0 153 137 0 16 16 153
1972 Senegal 385 0 148 0 148 135 0 13 13 148
1972 Sudan 91 20 35 0 55 50 0 5 5 55
1972 Zaire 81 0 31 0 31 0 0 31 31 31
1972 World 8098 33 3158 1030 4221 1431 13 2710 2774 4221
1980 China 1667 0 667 0 667 3 0 598 664 667

1980 India 4059 0 1705 0 1705 394 0 1311 1311 1705
1980 United States 202 4 85 0 89 0 0 85 85 89
1980 Argentina 147 7 57 0 64 42 0 8 8 64
1980 Brazil 196 0 72 0 72 46 0 26 26 72
1980 Burma 319 0 121 0 121 0 0 121 121 121
1980 Indonesia 47 0 18 0 18 0 0 18 18 18
1980 Nigeria 204 0 79 0 79 0 0 79 79 79
1980 Senegal 258 0 95 0 95 49 0 46 46 95
1980 Sudan 377 0 145 0 145 75 0 70 70 145
1980 Zaire 107 0 41 0 41 0 0 41 41 41
1980 World 7813 16 3181 488 3685 549 0 3050 3116 3685
1990 China 3250 0 1300 0 1300 160 0 1028 1140 1300

1990 India 5999 0 2520 0 2520 175 5 2340 2345 2520
1990 United States 313 6 136 0 142 35 0 103 103 142
1990 Argentina 123 0 48 0 48 38 3 7 10 48
1990 Brazil 50 0 20 0 20 3 0 17 17 20
1990 Burma 320 0 105 0 105 10 0 95 95 105
1990 France 0 0 0 253 253 3 0 250 250 253
1990 Indonesia 45 16 17 132 165 0 0 145 145 165
1990 Nigeria 80 0 28 0 28 0 0 28 28 28
1990 Senegal 480 34 183 0 217 166 0 22 22 217
1990 Sudan 145 0 56 0 56 52 0 4 4 56
1990 Zaire 129 0 50 0 50 0 0 50 50 50
1990 World 11350 61 4518 717 5296 626 8 4478 4610 5296
2000 China 6800 0 2660 0 2660 15 0 2645 2645 2660

2000 India 4300 0 1810 0 1810 20 10 1780 1790 1810
2000 United States 248 2 104 0 106 5 0 99 99 106
2000 Argentina 142 5 62 0 67 50 3 5 15 67
2000 Brazil 60 0 24 0 24 1 0 23 23 24
2000 Burma 390 0 123 0 123 10 0 113 113 123
2000 Indonesia 64 8 24 7 39 0 0 33 33 39
2000 Nigeria 510 0 163 0 163 0 0 163 163 163
2000 Senegal 420 3 190 0 193 144 0 44 44 193
2000 Sudan 210 0 81 0 81 76 0 5 5 81
2000 Zaire 120 0 46 0 46 0 0 46 46 46
2000 World 14139 22 5254 255 5531 274 35 5204 5239 5531
2002 China 6950 0 2719 0 2719 10 0 2709 2709 2719

2002 India 5060 0 2128 0 2128 50 10 2068 2078 2128
2002 United States 292 2 127 0 129 5 0 122 122 129
2002 Argentina 125 0 55 0 55 47 3 5 8 55
2002 Brazil 60 0 24 0 24 1 0 23 23 24
2002 Burma 390 0 123 0 123 10 0 113 113 123
On Total Total
Year Country Crushed Hand Production Imports Supply Exports Food Feed: Waste Total Distribution
2002 Indonesia 62 5 23 0 28 0 0 26 26 28
2002 Nigeria 528 0 169 0 169 0 0 169 169 169
2002 Senegal 160 5 68 0 73 63 0 10 10 73
2002 Sudan 211 0 81 0 81 75 0 6 6 81
2002 Zaire 132 0 51 0 51 0 0 51 51 51
2002 World 14901 17 5502 244 5763 222 37 5495 5532 5763
1
MODIFICATION OF OIL CHARACTERISTICS THROUGH BREEDING 445
3. ENVIRONMENTAL AND GENOTYPE EFFECTS

ON THE COMPOSITION PEANUTS
Major factors that influence the oil and other composition components of the peanut
include cultivar and maturity (37) as well as the environmental production condi-
tions of light, temperature, water stress, soil constituents, atmospheric constituents,
herbicides and insecticides, physical damage, and pest attack (38). In the four major
U.S. market-types (runner, virginia, valencia, and spanish), total oil content varies
from 44% to 56% (37, 39, 40). Information on the environmental and genotypic
effects on oil and fatty acid composition in peanuts is available (40, 41). The effects
of production environment on oil composition of varieties grown in Australia (42),
India (43, 44), and the United States (4549) have been reported. In maturity stu-
dies, the total oil (as a percentage of dry weight) increased significantly and then
decreased slightly (50, 51). The most rapid changes in oil percentage occurred in
early maturity stages and corresponded to the time of very rapid increases in seed
dry weight (45, 5154). As the peanut oil content increases across maturity, there is
a concurrent change in fatty acid composition (45). Mature seeds contain more stea-
ric and oleic acids and less arachidic, behenic, and lignoceric acids than immature
seeds. The oleic/linoleic (O/L) ratio also increases with maturity (41, 45). Develop-
ment of new high oleic acid peanut cultivars will be discussed in the next section.
Oil content and fatty acid composition have been studied in aboriginal varieties
of Arachis hypogaea subsp. hypogaea and subsp. fastigiata. These varieties are
important because they contain germplasm that can be used to increase the varia-
bility in the genetic base of the cultivated varieties (55, 56). The A. hypogaea subsp.
hypogaea var. hypogaea cultivars were higher in oleic acid concentration than the
A. hypogaea subsp. fastigiata var. fastigiata, var. aequatoriana, and var. peruviana
cultivars in sources from Peru (57) and Bolivia (58). Similar results were also
obtained from Mexican landrace lines of A. hypogaea subsp. hypogaea var. hirsuta
(59). In contrast, a survey of 16 wild species of Arachis found that the wild species
had higher levels of linoleic acid in comparison with the Arachis hypogaea geno-
types (60)
4. MODIFICATION OF OIL CHARACTERISTICS THROUGH

BREEDING
Modification of fatty acid composition has been a particular goal of breeding pro-
grams because oil quality, fatty acid composition, and protein composition are
highly heritable traits. One of the keys to successful progress in a breeding program
is the availability of rapid, efficient screening systems. Some of the methods for
rapid screening are measurement of the iodine value (IV) by the oil refractive index
(61), estimation of the seed oil content by its specific gravity (62), and estimation of
seed fatty acid composition by use of a small tissue fraction and analysis through
direct transmethylation (63), which improves on the individual seed analysis meth-
od (64). Methods of peanut improvement through breeding programs have been
446 PEANUT OIL
discussed in detail (6568). Most peanut genotypes have 3667% oleic acid (O),
1546% linoleic acid (L), and O/L ratios between 1.19 and 4.46 (6972). While
surveying peanut genotypes for oil quality, it was found that two closely related
experimental lines had 80% oleic acid and 2% linoleic acid (O/L 40) with an
IV of 74 (71). This naturally occurring mutation may have resulted from a mutation
of aspartate at position 150 to asparagine in the cDNA that reduced oleoyl-PC desa-
turase activity (73). Initial oxidative stability studies were done comparing
extracted oil from the experimental high-oleic line with that of an isogenic sister
line with normal fatty acid composition (74). The results indicated that the high-
oleic peanut oil had a greater oxidative stability than the normal-oleic oil. These
experimental lines have been used in breeding programs to develop cultivars
with high O/L ratios (7579). Cultivars having these high O/L ratios do not have
significant differences in oil content (80) nor do they have significant differences in
color, aroma, flavor, or texture (81, 82). It is characteristic of these high oleic acid
lines to have a linoleic acid content of 4% or less. High oleic roasted peanut seed
have a more stable roasted peanut attribute after 6 weeks storage at 22 C, and their
estimated shelf life is approximately two times longer than that of seed from a nor-
mal-oleic variety Florunner (83). Comparison of flavor stability in high-oleic and
normal oleic roasted peanut seed during storage at low relative humidity (84)
or 20 C (85) indicated that the high-oleic sources had better flavor quality and
stability. Use of high oleic oil in roasting of peanuts resulted in slight increases
in shelf life as measured by oxidative stability index (OSI) and peroxide value
(86). The OSI decreased over storage time, but the differential between high-oleic
and normal roasting oils was maintained throughout the storage period. The stabi-
lity of high-oleic peanut, sesame, and soybean blends in comparison with normal-
oleic peanut, sesame, and soybean blends has also been investigated (87), as has the
effect of the high-oleic trait on roasted peanut flavor heritability (79, 88).
5. OIL COLOR
Color is an important quality parameter of edible oil, both in the refining process
and in the marketplace. It is frequently monitored in the product line according to
some commercial standards to maintain a consistent quality. Each oil has its own
characteristic color primarily because of naturally occurring polyphenolic pig-
ments, gossypol, chlorophyll, and carotenoids (89). Therefore, oil color is often
specified according to both market and trade rules established by various associa-
tions. Peanut oil of the first grade for cooking should not exceed 2 Lovibond red
with fixed Lovibond yellow 20 according to Chinese national standard GB5525-
85, and for salad use, it should be no more than 1.5 Lovibond red with fixed Lovi-
bond yellow 15 (90). The Lovibond method, American Oil Chemists Society
(AOCS) Method Cc 13e-92 (91), is practiced primarily outside the United States
and Canada (90), and AOCS Method Cc 13e-45 or Wesson method is used through-
out the Americas (92). Introduction of automated colorimeters made possible the
replacement of the manually operated visual color instrument. An international
PEANUT OIL EVALUATION AND COMPOSITION 447
collaborative study was conducted to establish a broad-scale correlation between an

automated colorimeter (Tintometer Model PFX 990 (The Tintometer Ltd)) and the
official visual colorimeter (Tinometer Model AF710) (93). The automated col-
orimeter was concluded to be an appropriate alternative. Recently, digital image
analysis has been proposed as an alternative method to the visual Lovibond method
(90). The light yellow color of peanut oil is caused by -carotene and lutein (94). As
peanuts mature, a distinct lightening of the oil color can be observed (95). This
lightening of oil color has been suggested as a method to assess maturity (96). How-
ever, because peanut oil color is affected by factors, such as water stress and rate
of curing in addition to maturity (97), this method was replaced by other maturity
evaluation methods (98, 99).
6. PEANUT OIL EVALUATION AND COMPOSITION
Crude peanut oil has a nutlike flavor, which is removed by refining (14). Flavor
quality ballots for oil quality have been described (100) and incorporate separate
ballots for grading and flavor intensity. The flavor quality ballot only describes
the flavor characteristics and does not include the suspected cause or process of
any off-odors (101). Lexicons of roasted peanut flavor terms are available, and
the origins of these flavor terms have been discussed (102). Although there is no
U.S. standard of identity per se, peanut oil must be suitable for human consumption
and conform to the identity characteristics defined by the Codex Alimentarius
Commission (103). The various chemical and physical characteristics for peanut
oil are given in Table 4.
Heat of fusion, or latent heat, is the quantity of heat required to change 1 g of
solid to a liquid with no temperature change. This latent heat increases with increas-
ing molecular weight. Heat of combustion is the amount of heat produced by com-
bustion of 1 kg of oil (104). The heat of combustion increases with the chain length
of the fatty acids for both monoacylglycerols and triacylglycerols (107).
The Hehner value expresses the percentage of water-insoluble fatty acids plus
unsaponifiable matter in an oil or fat (105). This method is of greatest value in test-
ing butterfat purify. Like most vegetable oils, peanut oil has a higher Hehner value
than butterfat (108). Lipids with soluble fatty acids will have lower Hehner values
than those with a greater proportion of high-molecular-weight fatty acids. The IV,
or Wijs iodine number, is the number of grams of iodine absorbed under standard
conditions by 100 g of fat. Peanut oils IV of 82107 indicates it is more saturated
than corn, cottonseed, or linseed oil but is less saturated than coconut, palm, or
butter oil (37). Oil from the high oleic peanut varieties has an IV usually between
73 and 77 (41).
Peroxide value is the measure of reactive oxygen content of a fat in terms
of milliequivalents per 1000-g fat, following AOCS method Cd 8-53 or AOAC
Method 965.33 (109). Elevated peroxide values indicate that lipid oxidation has
taken place (110). Free fatty acids can serve as substrates for lipoxygenase and per-
oxidase (111), both of which are inactivated during heating (112). Once the cell
448 PEANUT OIL
TABLE 4. Characteristics of Peanut Oil.
Characteristic Value Reference
Acetyl value 8.59.5 37

Acid value (maximum)
Refined 0.6 mg KOH/g oil 103
Cold Pressed 4 mg KOH/g oil 103
Calculated gums (phospatides x 32) 0.35% 104
Color (Lovibond, maximum) Yellow 1625; 2.0 red 37
Color (visual) Light yellow 37
Flavor and odor
Refined Bland 14
Cold Pressed Shall be characteristic of the natural product 103
Free from foreign and rancid odor or taste
Heat of fusion (unhydrogenated) 21.7 cal/g 37
Heating value 40.4 mJ/kg 104
Hehner value 9596 105
Insoluble Impurities (% maximum) 0.05 103
Iodine no. (Wijs) 86107 103
Kinematic viscosity (21.1 C) 70.7cSt 104
Melting point 03 C 106
Melting point of the fatty acids 2230 C 106
Moisture and volatiles 0.23% 106
Peroxide value (maximum)
Refined 10 meq peroxides O2 /kg oil 103
Cold Pressed 15 meq peroxides O2 /kg oil 103
Polenske value 0.5 37
Refractive index (nD40 C) 1.461.465 103
Reichert-Meisl value 0.5 37
Saponification number 187196 103
Smoke point (minimum) ~226.4 C 14
Specific gravity (20 C) 0.9120.920 103
Specific heat (Cp, liquid oil) 0.4914 0.004 T ( C) 107
Surface tension 35.6 mN/m 104
Thiocyanogen value 0.5 37
Titer 2632 C 37
Unsaponifiable lipids 0.40% 104
structure is disrupted, lipoxygenase reacts with linoleic, linolenic, or arachidonic

acid [either as the free acid, triacylglycerols, or methyl or ethyl esters (113)] to
form hydroperoxides. Hydroperoxides can undergo further decomposition to
form pentanal and hexanal, both of which are detectable by headspace analysis
(114). These oxidation products are correlated with reduced flavor scores (100)
and cardboard and painty defects (100). Although the peroxide value is used as
an indicator of oil oxidation, the Kreis test was found to be a better predictor of
oxidation than the peroxide value for peanut oil (115).
The Plenske value and ReichertMeissel values are indicators of steam-volatile
water-soluble (butyric, caproic, and caprylic) or water-insoluble (capric and
lauric) fatty acids, respectively (37). These tests were designed for detecting
low-molecular-weight fatty acids in oil and adulteration in butterfat (106). Butterfat

has a ReichertMeissel value of 1734.5 (110).
The thiocyanogen value (TV) is a measure of the amount of the reagent absorbed
by 1 g of fat. GLC methods have largely displaced this method for determining the
content of oleic, linoleic, and linolenic acids when IVs are determined (116).
Methods for calculating fat composition using the IV and TV have been discussed
(110).
For soap making, the melting point of the fatty acids (titer value) is an important
parameter (117). The titer value for peanut oil is lower than that for cottonseed oil
(3037 C), cocoa butter, and animal fats and oils (118) but is higher than that for
corn (1420 C) and/or linseed oil (1921 C) (37).
The unsaponifiable matter is largely sterols and methylsterols (119, 120).
Detailed compositional analysis of the unsaponifiable fraction will be discussed
under the sterol subheading.
Before the development of gas chromatography and high-pressure liquid chro-
matography, the presence of peanut oil (as an olive oil adulterant) could be detected
because peanut oil contains about 5% arachidic acid. Arachidic acid is insoluble in
cold alcohol unlike stearic and palmitic acids (110). Methods for the detection of
arachidic acid include the Bellier, Evers, EversBellier, and Renard tests (110,
121). Arachidic acid is predominant in the lecithin and cephalic fractions of peanut
oil (122). Detection methods for toxic oils as an adulterant in edible oils such as
peanut oil have been reviewed (123, 124).
Advances in instrumentation have brought about proposals of new methods for
oil content and quality measurements. Near-infrared transmittance spectroscopy has
been used as a nondestructive method for the determination of oil content in pea-
nuts (125). Fourier-transform infrared methodology has been applied as a quality
control method in determining peanut oil in high fat products such as peanut butter
(126) and monitoring changes in peanut oil and other oils under oxidative condi-
tions (127). Although Fourier-transformRaman spectroscopy has been applied to
the classification of fats and oils including peanut oil (128), differential scanning
calorimetry has been used to follow changes in the thermal characteristics of frying
oils such as peanut oil (129).
6.1. Fatty Acids

Peanut oil is composed of mixed acylglycerol of approximately 80% unsaturated
and 20% saturated fatty acids (37). In mature peanuts, the oil is 96% triacylglycerol
(130) with the main fatty acids being palmitic, oleic, and linoleic (40). Other fatty
acids found in peanut oil are arachidic, 11-eicosensoic, behemic, and lignoceric
acids. The long-chain fatty acids are usually found at about or slightly less than
2%. The percent of free fatty acids in peanut oil varies between 0.02% and 0.6%
(131). Lipase hydrolysis of triacylglycerols into free fatty acids and glycerol occurs
before germination (132) and during adverse storage (97). Consequently, high free
fatty acid values indicate poor handling, immaturity, mold growth, or other factors
that lead to triacylglycerol hydrolysis (133).
450 PEANUT OIL
TABLE 5. Reported Fatty Acid Composition Ranges of Peanut Oil.
Fatty Acid Percentage
Reference 102 135 141

Palmitic 8.014.0 7.412.5 5.310.4
Stearic 1.04.5 2.74.9 2.24.4
Oleic 35.069 41.367.4 52.882.2
Linoleic 12.043.0 13.935.4 2.927.1
Arachidic 1.02.0 1.21.9 1.11.8
Eicosenoic 0.71.7 0.71.4 0.72.4
Behenic 1.54.5 2.13.6 2.23.9
Lignoceric 0.52.5 0.91.7 1.01.9
With maturation, the percentage of oleic acid increases while linoleic acid per-
centage decreases slightly (41, 45). Oxidative stability of peanut oil is highly cor-
related with the ratio of oleic acid to linoleic acid (134); thus, oil stability is
correlated with maturity. Cooler production climates lower the O/L ratio, resulting
in oil with a shorter shelf life. Other environmental conditions, such as drought
(135), and dry-land farming (45) will also lower the O/L ratio, and selecting soils
with a more basic pH and increasing iron while avoiding overfertilization will
increase the O/L ratio (136). Application of growth regulators has been shown to
reduce the O/L ratio (137, 138), decrease the eicosenoic acid content (137), and
increase oil yield (139). Herbicides have been shown to have a slight effect on
the oleic and linoleic acid content (137, 140). Fatty acid composition of peanut
oil can also be widely influenced by cultivar source (141, 142). Varietal variations
in fatty acid composition are summarized in Table 5 and by Young (143). It is
again important to indicate that in high oleic acid peanut cultivars, the general char-
acteristic is a linoleic acid content of 4% or less (41).
6.2. Triacylglycerol Structure

Interest in the triacylglycerol structure of peanut oil arose from observations that
peanut oil showed atherogenic effects in rabbits and other animals (144147).
This atherogenicity has been attributed to the triacylglycerol structure of peanut
oil (148150) because treatment of peanut oil with a base, to bring about randomi-
zation, reduced the atherogenicity to that of corn oil (151). However, the results of
the Kritchevsky studies (148, 149, 151) have been questioned (40) on the basis that
they did not include other vegetable oils for comparison and a lack of data for
appropriate statistical analysis. More recent studies (152155) have shown that pea-
nut oil and peanut product-based diets produce a reduction in total and LDL cho-
lesterol.
Various studies have identified anywhere from 18 to 84 different triacylglycerol
species in peanut oil (149, 150, 156, 157). Although many different triacylglycerol
species have been identified, the data are conclusive concerning a nonrandom dis-
tribution of fatty acids in the sn-1, -2, -3 positions of the triacylglycerols. As the
composition of the peanut oil changes, so does the spatial arrangement of the tri-
acylglycerols (158). The predominate triacylglycerol species are OOL, OOO, OLL,
POL, and POO (O oleic, L linoleic, P palmitic) (157). Oleic acid is present
in high concentration at all three positions, and linoleic acid is found primarily in
the sn-2 position. The shorter chain length saturated fatty acids, palmitic and stea-
ric, are mainly located in the sn-1 position and less in the sn-3 position. The longer
chain length saturated fatty acids, arachidic, behenic, and lignoceric, are located in
the sn-3 position. Eicosenoic acid is also frequently located in the sn-3 position
(156, 157, 159). Peanuts are grown under many different environmental conditions,
and such environmental differences can also influence the composition of the pea-
nut oil and the triacylglycerol species (160). Because peanut triacylglycerol struc-
ture and composition and total oil composition are affected by environmental
factors and diverse genetic background (158), their atherogenic potency (148)
and oxidative stability (74) may also be affected by these conditions.
6.3. Phospholipids
The phospholipid content of peanut oil can vary from 0.6% to 2% depending on the
maturity of the peanuts from which the oil is extracted (161). The major phospho-
lipids of peanut oil are phosphatidic acid (PA), phosphatidylcholine (PC), phospha-
tidylethanolamine (PE), phosphatidylglycerol (PG), and phosphatidylinositol (PI).
The composition of the phospholipid fraction is influenced by maturity and by
the postharvest stresses to which the peanuts are subjected (162). The concentra-
tions of PA, PE, PC, and PG were higher in immature seed, and PI was lower,
when compared with mature seed. The concentration of all phospholipids except
PG increased when peanuts were subjected to a curing temperature of 40 C.
When the peanuts were frozen before curing, a significant increase was observed
in PA and PG, whereas PC and PE decreased in comparison with the controls. Oxi-
dative stability of peanut oil has been postulated for some time to be caused by con-
stituents in addition to the linoleic acid content and tocopherol content (163). More
recently, it has been reported that phospholipids act in a synergistic manner with
tocopherols in lengthening the onset of the induction period of lipid oxidation
(164, 165). The degree of unsaturation of the acyl fatty acid chain has an added
effect on the length of the induction period (164). PE and PI appeared to be
more effective than PC in increasing oil stability (164). The usually high concen-
tration of PC in raw peanut oil contributes to the efficiency of the degumming pro-
cess during refining (166). A critical concentration of PC is needed to ensure that a
gum is formed for the removal of the phospholipids.
6.4. Tocopherols
Tocopherols are considered a moderate antioxidant in the peanut oil. The Codex
Alimentaris standard for tocopherols in peanut oil (103) indicates a range of
48373 mg/kg for alpha-tocopherol, 0140 mg/kg for beta-tocopherol, 88389
mg/kg for gamma-tocopherol, and 022 mg/kg for delta-tocopherol. Total tocopherol
452 PEANUT OIL
content ranges from 130 to 1300 mg/kg. Tocotrienols should not be detectable in
peanut oil. Tocopherol content in the oil can be affected by variety, production loca-
tion within the United States, maturity, and temperature of seed storage (37). Sto-
rage of peanut seed at 38 C vs. 22 C reduced alpha-tocopherol content by about
25%. A multiyear study on oil composition of peanuts exported from Argentina,
China, and the United States found tocopherol content to be the highest in the
U.S. source and lowest in the China source (167). Alpha- and gamma-tocopherols
were found to be the most abundant forms. Tocopherol form influences the antioxi-
dant capacity. Gamma- and delta-tocopherols were found to be significantly better
antioxidants than alpha-tocopherol, in that either of the first two would protect oil
approximately twice as long as a similar concentration of the latter (168). In unpro-
cessed expeller-pressed peanut oil, the tocopherol content did not affect antioxidant
activity when the oil was stored at 2% relative humidity (RH) vs. 91% RH (169).
Total tocopherol content in oil may be reduced during the degumming and the
bleaching processes by 20% and 60%, respectively (170). Peanut oil tocopherols
are also lost during frying when peanut oil is used as a cooking oil (171). Tocophe-
rols are also known as vitamin E; thus, peanut oil can serve as a good source for this
vitamin particularly when the oil is unrefined. The vitamins found in peanuts are
given in Table 6.
TABLE 6. Vitamin Content of Peanuts (Units per 100 g Dry

Weight) (37).
Constituent Units
Fat soluble
Vitamin A 26 I.U.
Carotene (provitamin A) Trace (<1 ug)
Vitamin D ND
Vitamin E 26.3 59.4 mg/100-g oil
Alpha-tocopherol 11.9 25.3 mg/100-g oil
Beta-tocopherol 10.4 34.2 mg/100-g oil
Delta tocopherol 0.58 2.50 mg/100-g oil
Vitamin K ND
Water soluble
B-Complex
Vitamin B1Thiamine 0.99 mg
Vitamin B2Riboflavin 0.14 mg
Vitamin B6Pyridoxine 0.30 mg
Vitamin B12Cyanocobalamin ND
NiacinNicotinic acid 12.816.7 mg
Choline 165174 mg
Folic acid 0.28 mg
Inositol 180 mg
Biotin 0.034 mg
Pantothenic acid 2.715 mg
Vitamin C 5.8 mg
NDNondetectable.
USES 453
TABLE 7. Codex Alimentarius Standard Levels

of Desmethylsterols in Peanut Oil (102).
Constituent % Total Sterols
Cholesterol ND3.8
Brassicasterol ND0.2
Campesterol 12.019.8
Stigmasterol 5.413.2
Beta-sitosterol 47.469.0
Delta-5-avenasterol 5.018.8
Delta-7-stigmastenol ND5.1
Delta-7-avenasterol ND5.5
Others ND1.4
Total sterols (mg/kg) 9002900
NDNondetectable, defined as 0.05%.
6.5. Sterols
Sterols are a minor constituent of peanut oil, varying from 0.09% to 0.3% (172).
Refining can remove nearly 61% of the sterol content. The Codex Alimentaris stan-
dards for desmethysterols in peanut oil (103) are given in Table 7. Detailed analyzes
of the unsaponifiable lipid fraction from peanut oil can be found in the literature.
Analysis of the unsaponifiable fraction of Nigerian peanut oil indicated the total
fraction to be about 0.4%, and when subdivided by TLC, the fractions were sterols
60%, hydrocarbons 27%, the remainder aliphatic alcohols, and other minor compo-
nents (119, 120). Beta-sitosterol comprised 64% and campesterol 15% of the sterol
fraction. The major triterpene alcohols included 24-methylenecycloartanol at 46%
and cycloartanol at 33%. A more recent report on the separation of the unsaponifi-
able components of Madagascar peanut oil (173) indicated that the sterol fraction
was composed of 72% beta-sitosterol and about 17% campesterol. The 4-alpha-
methylsterol fraction was primarily composed of citrostadienol (20%), obtusifoliol
(17%), gramstisterol (15%), and cycloeucalenol (14%). The triterpene alcohol frac-
tion was composed of 14-methyl-cycloeucalenol (42%), cycloartenol (22%),
cycloartanol (15%), and lupeol (10%). Use of peanut oil as frying oil also results
in the loss of phytosterols (174). The major sterol component, beta-sitosterol, has
recently been shown to inhibit cancer growth (175) and may offer protection from
colon, prostate, and breast cancer.
7. USES
Peanut oil is used mainly for edible purposes in the preparation of shortening, mar-
garines, and mayonnaise, as a cooking and frying oil and as a salad oil. As indicated
previously, the primary use of edible peanuts outside North America is the produc-
tion of peanut oil (Tables 1, 2), and the oil may be hydrogenated into vanaspati, an
Indian analogue to margarine (176). Because of the high smoke point (229.4 C),
454 PEANUT OIL
refined peanut oil is often used in deep-fat frying (14), but hydrolysis of acylglycer-
ols into free fatty acids during frying leads to a decrease in smoke point (107). For
both frying and as a salad oil, peanut oil is considered to be superior to soybean oil
and develops fewer flavor defects with long-term use (177). Peanut oil is considered
to be superior in the manufacture of pourable dressings because of its ability to hold
solids in suspension longer (178). However, because peanut oil solidifies at 03 C,
it does not meet the definition for salad oil, which must remain clear after 5.5 hours
of immersion in an ice bath at 0 C (179). A nonedible use of peanut oil as a diesel
fuel has been investigated (180183), but it is more expensive than conventional
No. 2 diesel fuel and has the added drawbacks of lower heating value, greater sur-
face tension, greater viscosity, and greater density (104).
7.1. Peanut Oil as a Protectant

In developing countries, there is a need for economical and locally available mate-
rials that can be used as a protectant, particularly as a seed protectant. In Nigeria,
peanut oil is recommended for control of rice weevils (Sitophilus oryzae L.) (184)
and as a protectant of maize from damage by the maize weevil (Sitophilus zeamais
Motsch.) (185). Protection can last up to 180 days. Control of Sitophilus granaries
L. with peanut oil was effective for up to 90 days of storage for wheat (186). The
use of peanut oil for the control of Callosobruchus maculates (F) in cowpea grain
has been reported (187) and its mode of action investigated (188, 189). Applications
of the method have been reported from Gambia (190), Senegal (191), Nigeria, and
Colombia, South America (192). In India, peanut oil is used as a protectant against
Callosobruchus chinensis L. in chickpea (Cicer arietinum L.) (193). In Sahel, pea-
nut oil is used for protecting leguminous tree seeds against seed beetles (194). Other
protectant applications are its use as a protectant against infestations of Cryptolestes
pusillus and Rhyzopertha dominica in stored grains, such as maize and sorghum
(195). Application of peanut oil to apples as a postharvest treatment has been shown
to reduce superficial scald (196). Peanut oil has also been evaluated for control of
the parasitic tracheal mite [Acarapis woodi (Rennie)] in colonies of the honeybee
[Apis mellifera (L.)] (197).
8. DIETARY ASPECTS
Dietary aspects of high fat content products such as peanuts and peanut products
and of peanut oil are often in question. One point is the high atherogenic potential
of peanut oil, which has been attributed to its triacylglycerol structure (148150),
because treatment of the oil with a base to bring about randomization reduced the
atherogenicity to that of corn oil (151). Another study has suggested that the lectin
in peanut oil may significantly contribute to its atherogenic properties (198). Con-
tinued human epidemiological studies have shown a 3050% reduction in cardio-
vascular disease in individuals who ate nuts, including peanuts, four to five times a
week (199201). Another human subjects study found that the use of high oleic
REFERENCES 455
acid peanuts as the fat source in a low-fathigh-monounsaturate diet produced sig-

nificant positive changes in blood lipids in postmenpausal women, including reduc-
tion of total cholesterol from 264 to 238 mg/dl (202). Additional evidence for the
benefits of a diet high in monounsaturated and polyunsaturated fats and low in satu-
rated fat on body function is found in a recent study in which the subjects consumed
one of five diets: a low-fat diet, one including olive oil, one including peanuts and
peanut butter, one including peanut oil, and a typical American diet. Results indi-
cated that the diet including peanuts and peanut butter, the one including peanut oil,
and the diet including olive oil (all low in saturated fat and cholesterol, and high in
monounsaturated fat) lowered total cholesterol and LDL cholesterol. Further, each
of these three diets lowered triacylglycerol levels, but they did not lower the
beneficial HDL cholesterol (203, 204). Peanut oil because of its beta-sitosterol
may inhibit cancer growth (175) and may offer protection from colon, prostate,
and breast cancer. Snacking on peanuts or peanut products has a satiety effect
that enables individuals to control hunger without leading to a weight gain (205).
9. ALLERGENICITY
The allergenicity of peanuts is well documented (206). Because peanuts are among
the most potent allergenic foods, based on the prevalence of peanut allergy and the
frequency of reported severe adverse reactions (207209), peanut oil has been the
most thoroughly studied (210). It has been shown that the most peanut-allergic indi-
viduals can safely consume refined peanut oil, whereas unrefined oil can provoke
reactions in some of the same individuals. However, some other studies report cases
of allergic individuals reacting to peanut oil that presumably had been refined (211,
212). This has led to a debate about the safety of refined oils and specifically
whether to label each oil individually because of the potential risk of allergenicity.
It has been suggested that the discrepancy between these observations was caused
by processing differences (210). It was further suggested that there needs to be a
standardized and validated methodology for measuring the protein content and
immunoreactivity of the residual protein in the peanut oil. Such a standard metho-
dology can then be used to maintain process specifications. Thresholds of reactivity
to allergens also need to be established to assess fully the risk from very small
amounts. It has been questioned whether high oleic acid peanuts differ in their aller-
genic properties from normal peanuts. Investigation of this question concluded that
a high content of oleic fatty acid has no effect on peanut allergenicity (213).
REFERENCES
1. H. T. Stalker and C. E. Simpson, in H. E. Pattee and H. T. Stalker, ed., Advances in Peanut

Science, Am. Peanut Res. & Educ. Soc., Stillwater, Oklahoma, 1995, p. 15.
2. A. Krapovickas (English translation), in P. J. Ucko and I. S. Falk, eds., The Domestica-
tion and Exploitation of Plants and Animals, Gerald Duckworth Co. Ltd., London, U.K.,
1969, pp. 424441.
456 PEANUT OIL
3. D. J. Banks, Am. J. Bot., 74, 724 (Abstr) (1987).

4. D. J. Banks, Am. J. Bot., 75, 158 (Abstr) (1988).
5. D. J. Banks, T. Pozorski, S. Pozorski, and C. B. Donnan, Am. Peanut Res. Educ. Soc., 25,
34 (Abstr) (1993).
6. W. Alva, Nat. Geograph., 174, 510 (1988).
7. C. B. Donnan, Nat. Geograph., 174, 550 (1988).
8. R. O. Hammons, in PeanutCulture and Uses, Am. Peanut Res. & Educ. Assoc.,
Stillwater, Oklahoma, 1972, p 17.
9. USDA. FAS, Production Estimates and Crops Assessment Div. (Oct. 2002). World
Agricultural Production. Available: http://www.fas.usda.gov.
10. USDA. (2002). FAS, Production, Supply & Distribution Online, Washington, D.C.
Available: http://www.fas.usda.gov/psd/complete_files/default.asp.
11. F. R. Johnson, The Peanut Story, Johnson Publishing Co., Murfreesboro, North Carolina,
1964, p. 164.
12. G. E. Brooks, J. African History, 16, 29 (1975).
13. Unknown, Scientific Amer., April 21, 243 (1900).
14. J. G. Woodroof, Peanuts: Production, Processing, Products, Avi Publishing Co.,
Westport, Connecticut, 1983.
15. G. R. Ziegler and Y-J. Liaw, J. Amer. Oil Chem. Soc., 70, 947 (1993).
16. G. Hoffman, The Chemistry and Technology of Edible Oils and Fats and Their High Fat
Products, Academic Press Inc., San Diego, California, 1989.
17. M. Matricardi, R. Hesketh, and S. Farrell, Technical Note-20, Supercritical Fluid
Technologies, Inc, Newark, Delaware, 2002.
18. E. Stahl, E. Schutz, and H. K. Mangold, J. Agric. Food Chem., 28, 1153 (1980).
19. J. J. Snyder, J. P. Friedrich, and D. D. Christianson, J. Amer. Oil Chem. Soc., 61, 1851
(1984).
20. J. W. Goodrum and M. B. Kilgo, Trans. ASAE, 30, 1865 (1987).
21. T. H. Sanders, in Proc. World Conference on Oilseed Processing and Utilization,
American Oil Chemists Society, Champaign, Illinois, 2001, p. 141.
22. A. Rosendahl, D. L. Pyle, and K. Niranjan, Enzyme Microb. Technol., 19, 402 (1996).
23. K. C. Rhee, C. M. Cater, and K. F. Mattil, J. Food Sci., 37, 90 (1972).
24. A. Sharma, S. K. Khare, and M. N. Gupta, J. Amer. Oil Chem. Soc., 79, 215 (2002).
25. M. Esaka, K. Suzuki, and K. Kubota, Agric. Biol. Chem., 50, 2395 (1986).
26. M. Esaka, K. Suzuki, and K. Kubota, J. Food Sci., 52, 1738 (1987).
27. P. N. Mahesware, D. W. Stanley, and F. R. Van De Voort, J. Amer. Oil Chem. Soc., 57, 194
(1987).
28. R. Vetrimani, N. Jyothirmayi, P. Haridis Rao, and C. S. Ramadoss, Lebensm.-Wiss.
Technol., 25, 532 (1992).
29. B. Kaufmann and P. Christen, Phytochem. Anal., 13, 105 (2002).
30. M. Ramesh, P. Haridis Rao, and C. S. Ramadoss, Lebensm.-Wiss. Technol., 28, 96
(1995).
31. L. E. Garcia-Ayuso, J. Velasco, M. C. Dobarganes, and M. D. Luque-de-Castro, J. Agric.
Food Chem., 47, 2308 (1999).
REFERENCES 457
32. J. R. J. Pare, G. Matni, J. M. R. Belanger, K. Li, C. Rule, B. Thibert, V. Yaylayan, Z. Liu,

D. Mathe, and P. Jacquault, J. Assoc. Offic. Anal. Chem. Int., 80, 928 (1997).
33. T. D. Rausch, M.S. thesis, Dept. Food Sci., North Carolina State University, Raleigh,
North Carolina, 2002.
34. L. Stoloff, in J. V. Rodicks, C. W. Hesseltine, and M. A. Mehlman, eds., Mycotoxins in
Human and Animal Health, Pathotox Publishers, Park Forest South, Illinois, 1977, p. 9.
35. J. E. Smith and R. S. Henderson, eds., Mycotoxins in Animal Foods, CRC Press, Boca
Raton, Florida, 1991.
36. FAO, Worldwide regulations for mycotoxins 1995: A compendium, FAO Food and
Nutrition Paper 64, Rome, Italy, 1997.
37. W. Y. Cobb and B. R. Johnson, in PeanutCulture and Uses, Am. Peanut Res. Educ.
Assoc., Stillwater, Oklahoma, 1972, p. 210.
38. J. L. Harwood, Inform, 5, 835 (1994).
39. C. E. Holaday and J. L. Person, J. Food Sci., 39, 1206 (1974).
40. E. M. Ahmed and C. T. Young, in H. E. Pattee and C. T. Young, ed., Peanut Science and
Technology, Am. Peanut Res. Educ. Soc., Yoakum, Texas, 1982, p. 660.
41. P. C. Andersen and D. W. Gorbet, J. Agric. Food Chem., 50, 1298 (2002).
42. B. D. Harch, K. E. Basford, I. H. DeLacy, P. K. Lawrence, and A. Cruickshank, Genet.
Resour. Crop Evol., 42, 243 (1995).
43. R. Jambunathan, R. Sridhar, K. Raghunath, S. L. Dwivedi, and S. N. Nigam, J. Sci. Food
Agric., 61, 23 (1993).
44. S. L. Dwivedi et al., Peanut Sci., 20, 84 (1984).
45. C. T. Young, M. E. Mason, R. S. Matlock, and G. R. Waller, J. Amer. Oil Chem. Soc., 49,
314 (1972).
46. R. E. Worthington, R. O. Hammons, and J. R. Allison, J. Agric. Food Chem., 20, 727
(1972).
47. C. E. Holaday and J. L. Pearson, J. Food Sci., 39, 1206 (1974).
48. D. F. Brown, S. M. Cater, K. F. Mattil, and J. G. Darroch, J. Food Sci., 40, 1055 (1975).
49. I. B. Hashim, P. E. Koehler, E. E. Eitenmiller, and C. K. Kvien, Peanut Sci., 20, 21
(1993).
50. H. E. Pattee, A. E. Purcell, and E. B. Johns, J. Amer. Oil Chem. Soc., 46, 629 (1969).
51. H. E. Pattee, E. B. Johns, J. A. Singleton, and T. H. Sanders, Peanut Sci., 1, 57 (1974).
52. T. H. Sanders, J. A. Lansden, R. L. Greene, J. S. Drexler, and E. J. Williams, Peanut Sci.,
9, 20 (1982).
53. T. H. Sanders, A. M. Schubert, and H. E. Pattee, in H. E. Pattee and C. T. Young, eds.,
Peanut Science and Technology, Am. Peanut Res. Educ. Soc., Yoakum, Texas, 1982,
p. 624.
54. W. A. Court, R. C. Roy, and J. G. Hendel, Can. J. Plant Sci., 64, 521 (1984).
55. D. J. Banks, Crop Sci., 16, 499 (1976).
56. A. J. Norden, O. D. Smith, and D. W. Gorbet, in H. E. Pattee and C. T. Young, eds.,
p. 101.
57. N. R. Grosso and C. A. Guzman, J. Agric. Food Chem., 43, 102 (1995).
458 PEANUT OIL
58. N. R. Grosso, J. A. Zygadlo, A. L. Lamarque, D. M. Maestri, and C. A. Guzman, J. Sci.

Food. Agric., 73, 349 (1997).
59. D. T. Grimm, T. H. Sanders, H. E. Pattee, D. E. Williams, and S. Sanchez-Dominguez,
Peanut Sci., 23, 111 (1996).
60. R. E. Worthington and R. O. Hammons, J. Amer. Oil Chem. Soc., 54, 105A (1977).
61. A. R. Khan, D. A. Emery, and J. A. Singleton, Crop Sci., 14, 464 (1974).
62. J. B. Misra, S. K. Yadav, and S. Chauhan, J. Sci. Food Agric., 61, 231 (1993).
63. W. L. Zeile, D. A. Knauft, and C. B. Kelly, Peanut Sci., 20, 9 (1993).
64. C. T. Young and G. R. Waller, J. Agric. Food Chem., 20, 1116 (1972).
65. T. A. Coffelt, in G. Robbelen, R. K. Downey, and A. Ashri, eds., Oil Crops of the World:
Their Breeding and Utilization, McGraw-Hill, New York, 1983, p. 319.
66. A. J. Norden, O. D. Smith, and D. W. Gorbet, in H. E. Pattee and C. T. Young, eds.,
p. 95.
67. H. T. Stalker and J. Moss, Adv. Agron., 41, 1 (1987).
68. J. C. Wynne and T. A. Coffelt, in H. E. Pattee and C. T. Young, eds., Peanut Science and
Technology, Am. Peanut Res. Educ. Soc., Yoakum, Texas, 1982, p. 50.
69. H. T. Stalker, C. T. Young, and T. M. Jones, Oleagineux, 44, 419 (1989).
70. R. Jambunathan, A. K. Singh, S. Gurtu, and K. Raghunath, Oleagineux, 48, 415
(1993).
71. A. J. Norden, D. W. Gorbet, D. A. Knauft, and C. T. Young, Peanut Sci., 14, 7 (1987).
72. K. Treadwell, C. T. Young, and J. C. Wynne, Oleagineux, 38, 381 (1983).
73. A. C. Bruner, S. Jung, A. G. Abbott, and G. L. Powell, Crop Sci., 41, 522 (2001).
74. S. F. OKeefe, V. A. Wiley, and D. A. Knauft, J. Amer. Oil Chem. Soc., 70, 489 (1993).
75. K. M. Moore and D. A. Knauft, J. Hered., 80, 252 (1989).
76. D. W. Gorbet and D. Knauft, Crop Sci., 37, 1392 (1997).
77. D. W. Gorbet and D. Knauft, Crop Sci., 40, 1190 (2000).
78. W. D. Branch, Crop Sci., 40, 1823 (2000).
79. H. E. Pattee, T. G. Isleib, D. W. Gorbet, K. M. Moore, Y. Lopez, M. R. Baring, and C. E.
Simpson, J. Agric. Food Chem., 50, 7362 (2002).
80. L. C. Mercer, J. C. Wynne, and C. T. Young, Peanut Sci., 17, 17 (1990).
81. D. A. Knauft, K. M. Moore, and D. W. Gorbet, Peanut Sci., 20, 4 (1993).
82. H. E. Pattee and D. A. Knauft, Peanut Sci., 22, 26 (1995).
83. J. C. Braddock, C. A. Sims, and S. F. OKeefe, J. Food Sci., 60, 489 (1995).
84. J. B. Mugendi, C. A. Sims, D. W. Gorbet, and S. F. OKeefe, J. Amer. Oil Chem. Soc., 75,
21 (1998).
85. H. E. Pattee, T. G. Isleib, K. Moore, D. W. Gorbet, and F. G. Giesbrecht, J. Agric. Food
Chem., 50, 7366 (2002).
86. G. E. Bolton and T. H. Sanders, J. Amer. Oil Chem. Soc., 79, 129 (2002).
87. G. M. Sumainah, C. A. Sims, R. P. Bates, and S. F. OKeefe, J. Food Sci., 65, 901 (2000).
88. H. E. Pattee, T. G. Isleib, D. W. Gorbet, and F. G. Giesbrecht, and Z. Cui, in
K. Rajasekaran, T. J. Jacks, and J. W. Finley, eds., Crop Biotechnology, ACS Symposium
Series 829, American Chemical Society, Washington, D.C., 2002, p. 217.
REFERENCES 459
89. A. E. Bayley, Cottonseed and Cottonseed Products, Interscience Publishers, New York,
1948, p. 213.
90. S. Fengxia, Z. Dishun, and Z. Zhanming, J. Amer. Oil Chem. Soc., 78, 749 (2001).
4th ed., AOCS Press, Champaign, Illinois, 1993.
92. A. A. Belbin, Inform (AOCS), 4, 648 (1993).
93. P. J. Wan, T. W. Hurley, J. D. Guy, and D. L. Berner, J. Amer. Oil Chem. Soc., 74, 731
(1997).
94. H. E. Pattee and A. E. Purcell, J. Amer. Oil Chem. Soc., 44, 328 (1967).
95. K. T. Holley and C. T. Young, Proc. Peanut Improvement Working Group, 1963, p. 52.
96. D. A. Emery and C. L. Gupton, Oleagineux, 23, 99 (1968).
97. T. H. Sanders, A. M. Schubert, and H. E. Pattee in H. E. Pattee and C. T. Young, ed.,
p. 626.
98. T. H. Sanders, E. J. Williams, A. M. Schubert, and H. E. Pattee, Peanut Sci., 7, 78
(1980).
99. APRES, in E. M. Ahmed, ed., Quality Methods, Am. Peanut Res. Educ. Soc., Stillwater,
Oklahoma, 1985.
100. K. Warner, in D. B. Minn and T. H. Smouse, ed., Flavor Chemistry of Fats and Oils,
AOCS Press, Champaign, Illinois, 1985, p. 279.
101. G. V. Civille and C. A. Dus, in A. J. St. Angelo, ed., Lipid Oxidation in Food, American
Chemical Society, Washington, D.C., 1992, p. 279.
102. T. H. Sanders, H. E. Pattee, J. R. Vercellotti, and K. L. Bett, in H. E. Pattee and H. T.
Stalker, eds., Advances in Peanut Science, Am. Peanut Res. Educ. Soc., Stillwater,
Oklahoma, 1995, p. 528.
103. Codes Alimentarius Commission, Joint FAO/WHO Food Standards Programme, Codex
Alimentarius. Vol. 8 - Fats, Oils and Related Products. 2nd ed. Codex Stan 2101999,
Rome, Italy, Revised 2001.
104. J. W. Goodrum and S. E. Law, Trans. ASAE, 25, 897 (1982).
105. M. B. Jacobs, The Chemical Analysis of Foods and Food Products, 3rd ed., Robert E.
Krieger Publishing Co., Huntington, New York, 1973.
106. C. E. McMichael and A. E. Bailey, in M. J. Jacobs, ed., The Chemistry and Technology of
Food and Food Products, Interscience Publishers, New York, 1951, p. 1134.
107. M. W. Formo, in D. Swern, ed., Baileys Industrial Oil and Fat Products, Vol. 1, 4th ed.,
Wiley, New York, 1979, p. 172.
108. A. G. Woodman, Food Analysis, McGraw-Hill, New York, 1931.
109. AOAC, Official Methods of Analysis of the Association of Official Analytical Chemists,
AOAC, Arlington, Virginia, 1960.
110. G. A. Jacobsen, Inform (AOCS), 4, 811 (1993).
111. J. A. Singleton and H. E. Pattee, in G. Charalambous, ed., Analysis of Foods and
Beverages-Headspace Techniques, Academic Press, New York, 1978, p. 359.
112. A. J. St. Angelo, J. C. Kuck, and R. L. Ory, J. Agric. Food Chem., 27, 229 (1979).
113. A. J. St. Angelo and R. L. Ory, Phytopathology, 73, 315 (1982).
114. J. A. Singleton and H. E. Pattee, J. Amer. Oil Chem. Soc., 57, 405 (1980).
460 PEANUT OIL
115. S. Narasimhan, K. G. Raghuveer, C. Arumughan, K. K. Bhat, and D. P. Sen, Food Sci.

Technol., 23, 273 (1986).
116. N. O. V. Sonntag, in D. Swern, Baileys Industrial Oil and Fat Products, Vol. 2, 4th ed.,
117. F. A. Norris, in D. Swern, ed., Baileys Industrial Oil and Fat Products, Vol. 1, 4th ed.,
118. W. Heimann, Hanbuch der Lebensmittelchemis, Band IV: Fette und Lipide (Lipids),
Springer-Verlag, Berlin, 1969.
119. T. Itoh, T. Tamura, and T. Matsumoto, J. Amer. Oil Chem. Soc., 50, 122 (1974).
121. L. W. Aurand, A. E. Woods, and M. R. Wells, Food Composition and Analysis, Van
Nostrand Reihold, New York, 1987.
122. F. Osman, A. E. Ashour, and A. M. Gad, Fette, Seifen, Anstrichmit., 71, 264 (1969).
123. H. E. Cox and D. Pearson, The Chemical Analysis of Foods, Chemical Publishing Co.,
New York, 1962.
124. Nasirullah and K. V. Nagaraja, Oleagineux, 42, 35 (1987).
125. J. B. Misra, R. S. Mathur, and D. M. Bhatt, J. Sci. Food Agric., 80, 237 (2000).
126. F. R. van de Voort, J. Sedman, and A. A. Ismail, Food Chem., 48, 213 (1993).
127. M. D. Gullen and N. Cabo, J. Sci. Food Agric., 80, 2028 (2000).
128. V. Baeten, P. Hourant, M. T. Morales, and R. Aparicio, J. Agric. Food Chem., 46, 2638
(1998).
129. H. Gloria and J. M. Aguilera, J. Agric. Food Chem., 46, 1363 (1998).
130. T. H. Sanders, J. Amer. Oil Chem. Soc., 57, 8 (1980).
131. J. D. Guthie, C. L. Hoffpuir, M. F. Stansbury, and W. A. Reeves, USDA Agricultural
Circular, 61, USDA, Washington, D.C., March 1949.
132. T. H. Sanders and H. E. Pattee, Lipids, 10, 50 (1975).
133. J. I. Davidson, T. B. Whitaker, and J. W. Dickens, in H. E. Pattee and H. T. Stalker, eds.,
Advances in Peanut Science, Am. Peanut Res. Educ. Soc., Stillwater, Oklahoma, 1995,
p. 574.
134. S. P. Fore, N. J. Morris, C. H. Mack, A. H. Freeman, and W. G. Bickford, J. Amer. Oil
Chem. Soc., 30, 298 (1953).
135. I. B. Hashim, P. R. Koehler, R. R. Eitenmiller, and C. K. Kvien, Peanut Sci., 20, 21
(1993).
136. S. L. Dwivedi, S. N. Nigam, R. Jambunathan, K. L. Sahrawat, G. V. S. Nagabhushanam,
and K. Raghunath, Peanut Sci., 20, 84 (1993).
137. C. P. Malik, P. Singh, and U. Parmar, Phytochemistry, 28, 2651 (1986).
138. I. B. Hashim, P. R. Koehler, and C. K. Kvien, Peanut Sci., 20, 106 (1993).
139. P. Singh, U. Parmar, C. P. Malik, M. K. Grewal, and D. S. Bhatia, Peanut Sci., 18, 3
(1991).
140. E. W. Hauser, G. A. Buchanan, W. J. Ethredge, M. D. Jellum, and S. R. Cecil, Peanut Sci.,
3, 56 (1976).
141. R. E. Worthington, R. O. Hammons, and J. R. Allison, J. Agric. Food Chem., 20, 727
(1972).
REFERENCES 461
142. P. C. Andersen, K. Hill, D. W. Gorbet, and B. V. Brodbeck, J. Food Comp. Anal., 11, 100
(1998).
143. C. T. Young, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, Vol. 2, 5th ed.,
144. G. A. Gresham and A. N. Howard, Br. J. Exp. Pathol., 41, 395 (1960).
145. R. F. Scott, E. W. Morrison, R. J. Thomas, and S. C. Nam, Exp. Mol. Pathol., 3, 421
(1964).
146. H. Imai, K. T. Lee, S. Pastori, E. Ponlilio, R. Florantin, and W. A. Thomas, Exp. Mol.
Pathol., 5, 273 (1966).
147. D. Kritchevsky, S. A. Tepper, H. K. Kim, J. A. Story, D. Vesselinovitch, and R. W.
Wissler, Exp. Mol. Pathol., 24, 375 (1966).
148. D. Kritchevsky, S. A. Tepper, D. Vesselinovitch, and R. W. Wissler, Atheroclerosis, 17,
225 (1973).
149. J. J. Myher, L. Marai, A. Kuksis, and D. Kritchevsky, Lipids, 12, 775 (1977).
150. F. Manganaro, J. J. Myher, A. Kuksis, and D. Kritchevsky, Lipids, 16, 508 (1981).
151. D. Kritchevsky, Am. J. Clin. Nutr., 23, 1105 (1970).
152. L. Cordain, Lipids, 33, 229 (1998).
153. S. Satchithanandam, T. J. Flynn, R. J. Calvert, and D. Kritchevsky, Lipids, 34, 1305
(1999).
154. P. M. Kris-Etherton, G. Zhao, A. E. Binkoske, S. M. Coval, and T. D. Etherton, Nutr. Rev.,
59, 103 (2001).
155. R. L. Hargrove, T. D. Etherton, T. A. Pearson, E. H. Harrison, and P. M. Kris-Etherton,
J. Nutr., 131, 1758 (2001).
156. G. Sempore and J. Bezard, J. Amer. Oil Chem. Soc., 59, 124 (1982).
157. J. A. Singleton and H. E. Pattee, J. Amer. Oil Chem. Soc., 64, 534 (1987).
158. J. C. Hokes and R. E. Worthington, J. Amer. Oil Chem. Soc., 56, 953 (1979).
159. T. H. Sanders, Lipids, 14, 630 (1979).
162. J. A. Singleton and L. F. Stikeleather, J. Amer. Oil Chem. Soc., 72, 485 (1995).
163. S. P. Fore, N. J. Norris, C. H. Mack, A. F. Freeman, and W. G. Bickford, J. Amer. Oil
Chem. Soc., 30, 298 (1953).
164. D. H. Hildebrand, J. Terao, and M. Kito, J. Amer. Oil Chem. Soc., 61, 552 (1984).
165. T. Ohshima, Y. Fujita, and C. Koizumi, J. Amer. Oil Chem. Soc., 70, 269 (1993).
166. J. A. Robertson, W. H. Morrison, and D. Burdick, J. Amer. Oil Chem. Soc., 50, 443
(1973).
167. T. H. Sanders, J. R. Vercellotti, K. L. Crippen, R. T. Hinsch, G. K. Rasmussen, and J. H.
Edwards, J. Amer. Oil Chem. Soc., 69, 1032 (1992).
168. R. M. Parkhurst, W. A. Skinner, and P. A. Sturm, J. Amer. Oil Chem. Soc., 45, 641 (1968).
169. A. G. Gopala Krishna and J. V. Prabhakar, J. Amer. Oil Chem. Soc., 71, 1245 (1994).
170. E. M. Prior, V. S. Vadke, and F. W. Sosulski, J. Amer. Oil Chem. Soc., 68, 401 (1991).
171. K. I. Holownia, M. C. Erickson, M. S. Chinnan, and R. R. Eitnmiller, Food Res. Int., 34,
77 (2001).
462 PEANUT OIL
172. A. Seher, Fett. Wissenschaft Technolgie, 89(1), 27 (1987).

173. E. M. Gaydoa, J. P. Bianchini, and J. V. Ratovohery, J. Agric. Food Chem., 31, 833
(1993).
174. L. L. Oehrl, A. P. Hansen, C. A. Rohrer, G. P. Fenner, and L. C. Boyd, J. Amer. Oil Chem.
Soc., 78, 1073 (2001).
175. A. B. Awad, K. C. Chan, A. C. Downie, and C. S. Funk, Nutr. Cancer, 36, 238 (2000).
176. D. K. Salunke and B. B. Desai, Postharvest Technology of Oilseeds, CRC Press, Boca
Raton, Florida, 1986, p. 35.
177. J. C. Cowen, H. Moser, G. R. List, and C. D. Evans, J. Amer. Oil Chem. Soc., 48, 835
(1971).
178. L. Branscome and C. T. Young, Peanut Oil, Georgia Agricultural Commodity Commis-
sion for Peanuts, Tifton, Georgia, 1972.
179. R. G. Krishnamurthy, in D. Swern, ed., Baileys Industrial Oil and Fat Products, Vol. 1,
4th ed., Wiley, New York, 1982, p. 315.
180. J. W. Goodrum, Trans. ASAE, 27, 1257 (1985).
181. M. A. Mazed, J. D. Summers, and D. G. Batchelder, Trans. ASAE, 27, 1371 (1985).
182. M. A. Mazed, J. D. Summers, and D. G. Batchelder, Trans. ASAE, 27, 1375 (1985).
183. J. W. Goodrum, V. C. Patel, and R. W. McClendon, Trans. ASAE, 39, 817 (1996).
184. M. F. Ivbijaro, C. Ligan, and A. Youdeowei, Agric. Ecosyst. Environ., 14, 237 (1985).
185. M. F. Ivbijaro, Proc. Ecol., 6, 267 (1984).
186. E. Tembo and R. F. A. Murfitt, J. Stored Prod. Res., 31, 77 (1995).
187. S. R. Singh, R. A. Luse, K. Leuschner, and D. Nangju, J. Stored Prod. Res., 14, 77 (1978).
188. K. N. Don-Pedro, Pestic. Sci., 26, 107 (1989).
189. K. N. Don-Pedro, Pestic. Sci., 30, 295 (1990).
190. S. D. Cockfield, Trop. Pest Manage., 38, 268 (1992).
191. G. Pierrard, Trop. Pest Manage., 32, 197 (1986).
192. Anonymous, League Int. Food Educ. Newsl., May, 3-4 (1979).
193. S. K. Singal, Z. Singh, Trop. Pest Manage., 36, 314 (1990).
194. C. Cardet, T. Kandji, A. Delobel, and P. Danthu, Agrofor. Syst., 40, 29 (1998).
195. D. Obeng-Ofori, Entomol. Exp. Appl., 77, 133 (1995).
196. K. J. Scott, C. M. C. Yuen, and G. H. Kim, Postharvest Biol. Technol., 6, 219 (1995).
197. N. W. Calderone and H. Shimanuki, J. Econ. Entomol., 88, 805 (1995).
198. D. Kritchevsky, S. A. Tepper, and D. M. Klurfeld, Lipids, 33, 821 (1998).
199. G. E. Fraser, J. Sabate, W. L. Beeson, and T. M. Strahan, Arch. Inter. Med., 152, 1416
(1992).
200. J. Sabate and G. E. Fraser, Prim. Cardiol., 19, 65 (1993).
201. F. B Hu, M. J. Stampfer, J. E. Manson, E. B. Rimm, G. A. Colditz, B. A. Rosner, F. E.
Speizer, C. H. Hennekens, and W. C. Willett, Brit. Med. J., 317, 1341 (1998).
202. D. J. OByrne, D. A. Knauft, and R. B. Shireman, Lipids, 32, 687 (1997).
203. P. M. Kris-Etherton, T. A. Pearson, Y. Wan, R. Hargrove, K. Moriarty, V. Fishnell, and
T. D. Etherton, Am. J. Clin. Nutr., 70, 1009 (1999).
204. P. M. Kris-Etherton, G. Zhoa, A. E. Binkoski, S. M Coval, and T. D. Etherton, Nutr. Rev.,
59, 103 (2001).
REFERENCES 463
205. S. V. Kirkmeyer and R. D. Mattes, Int. J. Obes. Relat. Metab. Disord., 24, 1167 (2000).
206. W. Burks, H. A. Sampson, and G. A. Bannon, Allergy, 53, 725 (1998).
207. C. Loza and J. Brostoff, Clini. Exper. Allergy, 25, 493 (1995).
208. S. M. Tariq, M. Stevens, S. Matthews, S. Ridout, R. Twiselton, and D. W. Hide, Brit.
Med. J., 313, 514 (1996).
209. J. O. Hourihane, S. A. Kilburn, J. A. Nordlee, S. L. Taylor, and J. O. Warner, Brit. Med. J.,
314, 1084 (1997).
210. R. W. R. Crevel, M. A. T. Kerkhoff, and M. M. G. Koning, Food Chem. Toxicol., 38, 385
(2000).
211. A. Olszewski, L. Pons, F. Moutete, I Aimone-Gastin, G. Kanny, D. A. Moneret-Vautrin,
and J. L. Gueant, Clini. Exper. Allergy, 28, 850 (1998).
212. D. A. Moneret-Vautrin, F. Rance, G. Kanny, A Olszewski, J. L. Gueant, G.Dutau, and L.
Guerin, Clini. Exper. Allergy, 28, 1113 (1998).
213. S-Y. Chung, S. Maleki, E. T. Champagne, K. L Buhr, and D. W. Gorbet, J. Agric. Food
Chem., 50, 878 (2002).
10
Rice Bran Oil
Frank T. Orthoefer
1. INTRODUCTION
Rice oil, also called rice bran oil, has been used extensively in Asian countries such
as Japan, Korea, China, Taiwan, Thailand, and Pakistan (1, 2). It is the preferred oil
in Japan for its subtle flavor and odor. Interest in rice oil in the United States was
initiated after WWII, primarily to provide an additional revenue stream to the rice
miller. More recently, interest in rice oil escalated with its identification as a
healthy oil that reduces serum cholesterol (3, 4).
Three facilities were constructed in the United States to produce rice oil (5). The
first facility began operation in the late 1950s, and a second facility was started in
the 1960s. Both were shut down in the early 1980s because of economics. A third
production facility began operation in the early 1990s and continues producing both
bulk and packaged oils for the domestic and export markets. Attempts at further
development of rice oil production have not been successful because of high capital
requirement to construct an oil extraction plant and refining facility and limited
availability of stabilized rice bran (6).
Rice oil is a minor constituent of rough rice when compared with the carbohy-
drate and protein content. Two major classes of lipids are present: those internal
within the endosperm and those associated with the bran. The internal lipids con-
tribute to the nutritional, functional, and sensory qualities of rice (7).
Rice bran is the main source of rice oil. The majority of available bran continues
to be used for animal feeds without being extracted for the oil. The food industry
465
466 RICE BRAN OIL
uses minor quantities of stabilized rice bran as a source of dietary fiber, protein, and
desirable oil.
This chapter reviews the source and composition of rice bran oil, its nutritional
characteristics, production, and refining of the oil and its applications.
2. COMPOSITION OF RICE AND RICE BRAN LIPIDS
The structure of the rice kernel is given in Figure 1. Lipids are present as sphero-
somes or lipid droplets less than 1.5 mm in diameter in the aleurone layer, less than
1.0 mm in the subaleurone layer, and less than 0.7 mm in the embryo of the rice
grain (7, 9). Most of the lipids in the endosperm are associated with protein bodies
and the starch granules as bound lipids (10). The lipids are broadly classified as
nonstarch and starch lipids (Table 1). The majority of the lipids are the nonstarch
lipids. Starch lipids consist primarily of lysophospholipids, triacylglycerols, and
free fatty acids (13). Major phospholipid species are lysophophatidylethanolamine
and lysophosphatidylcholine. The major fatty acids are palmitic and linoleic acids
along with oleic acid. Minor amounts of monoacylglycerols, diacylglycerols, and
sterols are also found. Glycolipids found are diglycosyl monoacylglycerols and
monoglycosyl monoacylglycerols. The component sugars are galactose and glu-
cose.
The nonstarch lipids in the aleurone, subaleurone, and germ layers were 8691%
neutral lipids, 25% glycolipids, and 79% phospholipids, although these are vari-
able because of different milling degrees (11). The fatty acid composition of
Figure 1. Relative proportion of major rice caryopsis components (8).

COMPOSITION OF RICE AND RICE BRAN LIPIDS 467
TABLE 1. Lipid Composition of Rice and its Fractions (7, 9, 11, 12).a
Nonwaxy
Nonstarch Lipids in Rice Fractions Starch Lipid in

Brown Milled
Property Hull Brown Milled Bran Germ Polish Rice Rice
Lipid content 0.4 2.7 0.8 18.3 30.2 10.8 0.6 .05
Saponification no. 145 181 190 184 189
Iodine no. 69 94 100 99 101
Unsaponifiable matter 26 6 6 6 34
Fatty acid composition Wt % of total
Palmitic 18 23 33 23 24 23 46 45
Oleic 42 35 21 37 36 35 12 11
Linoleic 28 38 40 36 37 38 38 40
Others 12 4 6 4 3 4 4 4
Neutral lipids, % of total lipids 64 86 82 89 91 87 28 26
Triglyceride 71 58 76 79 72 4 2
Free fatty acids 7 15 4 4 5 20 21
Glycolipids, % of total lipids 25 5 8 4 2 5 19 16
Phospholipids, % total lipids 11 9 10 7 7 8 53 58
Phosphotidylcholine 4 9 3 3 3 4 4
Phosphatidylethanolamine 4 4 3 3 3 5 5
Lysophosphatidylcholine <1 2 <1 <1 <1 21 23
Lysophosphatidylethanolamine 1 22 25
a
Based on 6% bran-germ, 4% polish, and 90% milled rice from brown rice.
nonstarch lipids showed 2225% palmitic, 3741% oleic acid, and 3741% linoleic
acid (Table 2). The brown rice non-starch lipids was 1418% in germ, 3941% in
bran, 1521% in polish, and 2533% in milled rice. The composition was 8387%
triacylglycerol together with 79% free fatty acids, diacylglycerols, sterols together
with sterol esters, hydrocarbons, and wax. Oil extracted from rice bran contained
20.1% total lipid, 89.2% neutral lipids, 6.8% glycolipid, and 4.0% phospholipid
(14). A component of rice bran oil that has promise as a nutraceutical compound
is g-oryzanol (15). g-Oryzanol was first isolated from soapstock from rice oil
TABLE 2. Major Lipid Classes of Crude Bran Oil Extracted from Raw Rice Bran
and Their Fatty Acid Composition (14).
Fatty Acid Composition (%)
Lipid classa wt% 14:0 16:0 18:0 18:1 18:2 18:3 20:0 saturated unsaturated
TL 20.1 .40 22.21 2.21 38.85 34.58 1.14 0.61 25.43 74.57
NL 89.2 0.43 23.41 1.88 37.24 35.29 1.07 0.68 26.40 73.60
GL 6.8 0.09 27.34 0.28 36.45 35.76 0.18 27.61 72.39
PL 4.0 0.11 22.13 0.16 38.11 39.32 0.17 22.40 77.60
a
TL total lipids; NL neutral lipids (nonpolar lipid and free fatty acids); GL glycolipids; PL phospholipids.
468 RICE BRAN OIL
CH3O
3 2
7 8
4 1
H CH CH COOR
5 6
(ROH see below)
ROH = campesterol
= O sitosterol
= cycloartenol
= 24 methylene-cycloartenol
= cyclobranol
Figure 2. Major ferulates in oryzanol (9).
refining (16). Although originally thought to be a single compound, it is now known

to be a mixture of steryl and other triterpenyl esters of ferulic acids (cycloartenyl
ferulate, 24 methylenecycloartenyl ferulate, and b sitosterol ferulate and campesteryl
ferulate) (Figure 2). It is present at 1.52.9% of rice bran oil with a m.p. of 138.5 C.
The oryzanol content is dependent on rice grain variety with long grain rice at
6.42 mg/g and medium grain rice at 5.17 mg/g (17).
Tocopherols and tocotrienols (tocols) are present in rice oil (Figure 3). Crude
rice bran oil was found to contain, per 100 g of oil, 1946 mg of a-tocopherol,
13 mg of b-tocopherol, 110 mg of g-tocopherol, and 0.40.9 mg of d-tocopherol,
1433 mg of a-tocotrienol, and 9-69 mg of g-tocotrienol (18, 19) (Table 3). The
mean tocol content was 93 mg/100 g for crude oil and 50 mg/100 g for refined
oil (19). Close to 370 mg/100 g has been reported (20). Rice bran stabilization
and storage (21) and method of extraction (22) affects the concentration of tocols
in the oil. g-Tocotrienol is more stable and persists to a greater extent during storage
than other tocols (21). Other factors influencing toco content are milling and variety
(17, 23). Long-grain varieties have higher levels of tocotrienols than medium grain
rice (17).
R3
R2 O
HO R1 R2 R3
R1 -T(3) CH3 CH3 CH3
Tocopherols (T) -T(3) H3 H CH3
-T(3) H CH3 CH3
R3 -T(3) H H CH3
R2 O
HO
R1
Tocotrienols (T3)
Figure 3. Structure of tocopherol and tocotrienol (9).

COMPOSITION OF RICE AND RICE BRAN LIPIDS 469
TABLE 3. Tocopherol and Tocotrienol Concentrations (mg/100g) in Raw Rice Bran

and Commercially Available Refined Oil (14).
Source a-T b-T g-T d-T a-T3 g-T3 d-T3
Rice bran 6.3 0.9 3.20 0.20 3.8 12.0 0.7

Brown ricea 0.63 0.09 0.32 0.02 0.38 1.2 0.07
Crude oila 31.50 4.50 16.00 1.00 19.0 60.0 3.5
Refined oil 8.2 12.80 1.3 2.1 42.9 3.5
a
Calculated values.
Waxes are present as long-chain fatty acid esters with fatty alcohols, methanol,
and ethanol. Fatty acid analysis showed that behenic (C:22), lignoceric (C:24), and
palmitic acids (C:16) are the major fatty acid for longer alkyl esters and oleic and
palmitic for the shorter alkyl esters (Table 4) (24). The major alcohols found are for
longer alkyl esters. These are as follows:
TABLE 4. Fatty Acids of Sterol and Alkyl Esters, Alcohols of Longer Alkyl Esters,
and Alkanes and Alkenes of Rice Bran Waxy Lipids (24).
Fatty Acids Composition of:
Carbon and Alcohols of

Double Bond Longer Alkyl Shorter Alkyl Longer Alkyl
No. Sterol Esters Esters Esters Alkane Alkenes
14.0 0.6 1.8 2.2

16.0 11.1 23.8 35.5
18:0 1.0 3.8 0.8
18:1 33.1 2.9 60.2
18:2 51.1 0.3 1.5
18:3 2.0
20:0 0.7 3.6 0.1
22:0 32.6 2.0
23:0 31.2 1.3
24:0 11.2 0.2
25:0 2.0
26:0 6.3 0.8
27:0 9.5 7.9
28:0 12.5 3.6 1.3
29:0 46.5 38.8
30:0 19.1 3.0 0.9
31:0 23.7 20.8
32:0 10.5 (5.1) 1.4 0.5
33:0 6.5 18.8
34:0 6.6 (18.3) 0.7 1.4
35:0 0.8 8.4
36:0 3.0 (5.4) 0.1
37:0 1.6
470 RICE BRAN OIL
Major alcohols:
Tetratriacontanol C34:0
Triacontanol C30:0
Dotriacontanol C32:0
Octacosanol C28:0
Tetracosanol C24:0
Straight-chain alkanes, alkenes, and branched-chain alkenes (squalene) are detected

in the hydrocarbon fraction. The squalene content is 120 mg/100 g.
Hard and soft waxes are recovered from crude rice bran oil with m.p. of 79.5 C
and 74 C (25). The hard wax consists of 64.5% fatty alcohols, 33.5% fatty acids,
and 2% hydrocarbons. Soft wax includes 51.8% fatty alcohols, 46.2% fatty acids, and
2% hydrocarbons.
3. MILLING OF RICE
Todays modern rice mills efficiently separate hulls from paddy rice followed by
bran removal (Figure 4) (6). Milling consists of rubber roll dehullers, paddy separa-
tors, abrasive milling (whitening), and possibly friction mills. The bran and polish
consist mainly of the outer layers of rice caryopsis. These include the pericarp, seed
coat, nucellus, aleurone layer, germ, and part of the subaleurone layer of the starchy
endosperm. Rice bran makes up 58% of rough rice, and the polish may account
for an additional 23% (5). Commercial rice bran is a fine, floury material made
up of the outer layers of the brown rice plus pulverized germ, some hull fragments,
and some endosperm (white rice fragments) (8). The particle size distribution of the
bran is shown in Table 5 (26). The particle size of the bran varies significantly with
type of milling and milling condition. The composition of the bran also varies as
a function of milling degree (Table 6) (27). Generally, a low degree of milling is
practiced.
Rice bran is rich in lipids, proteins, minerals, vitamins, phytin, trypsin inhibitor,
lipase, and lectin (hemeagglutinins) (5). Compared with other cereal brans, rice
bran with germ is a little higher in fat content but comparable in protein, fiber,
and ash (Table 7). The high phosphorous content is among the highest of the cereal
grains. Rice bran is also high in silica probably because of the presence of rice hull
fragments. Bran is high in B vitamins and tocopherol, but it contains only a little
Vitamin A and C (28).
Rice bran and germ are used in animal feeds as a low-cost source of protein and
oil (6). Rice mill feed is a combined product produced by huller mills, where
dehulling and milling is a single processing step (5). Raw rice bran, when dehulling
is a separate processing step, has about four times the oil content (1720%) of rice
mill feed (6). Parboiled rice bran produced by cooking of rough rice prior to milling
has a greater oil content, usually above 20%, than raw rice bran. The higher oil con-
tent may be caused by less endosperm contamination, better extractability of the oil
MILLING OF RICE 471
Steps in Rice Milling
Rough Rice
Screening
Destoning
Dehulling
Polishing Bran
Grading and sizing
Milled Rice
Figure 4. Steps in rice milling.
TABLE 5. Particle-Size Distribution (%) of Raw and Heat-Stabilized Brans (26).
Mesh Particle Size (mm) Raw Bran Moist Heat-Stabilized Bran
18 >1000 0 0
1830 1000595 2.4 18.6
3050 595297 30.0 32.7
5080 297177 12.2 18.5
80100 177149 8.5 10.8
<100 <149 46.7 19.4
TABLE 6. Variation in Rice Bran Composition as a Function of the Degree of Milling (27).
Degree of Bran Composition (%)

Milling (%) Protein Fat Fiber Ash NFEa
1st Cone 03 170 17.7 10.5 9.8 45.0

2nd Cone 36 17.6 17.1 10.3 9.4 45.2
3rd Cone 69 17.0 16.5 5.7 8.4 52.5
4th Cone 910 16.7 14.2 5.7 7.5 55.9
a
Nitrogen-free extract.
TABLE 7 Composition (% at 14% Moisture) of Rice Bran and Polish and Other Cereal Brans (26).
Rice

Bran Polish Wheat Corn Barley Rye Oat Bran Sorghum Millet
Constituent Bran Bran Bran Bran Shorts Bran Bran
Crude protein (% N 6.25) 12.05.6 11.813.0 14.515. 7.811.5 11.5 14.6 8.816.2 7.715.0 11.5
Crude fat (%) 15.019.7 10.112.4 2.94.3 4.48.1 2.8 2.6 3.06.8 4.64.7 8.0
Crude fiber (%) 7.011.4 2.33.2 6.810.4 2.69.4 9.6 6.6 20.5 7.49.1
Available carbohydrates (%) 31.152.3 51.155.0 50.759.2 58.962.6 58.4 58.0 61.4 54.364.1 56.0
Crude ash (%) 6.69.9 5.27.3 4.06.5 1.93.4 3.6 4.2 6.3 2.13.0 10.5
Calcium (mg/g) 0.31.2 0.50.7 1.21.3 0.30.4 2.8 0.91.2 0.9 0.8
Magnesium (mg/g) 513 67 5.6 2.5 3.0 4.0
Phosphorus (mg/g) 1125 1022 913 16 58 7.210.5 8.1
Phytin phosphorus (mg/g) 922 1217 10 3.1 6.9
Silica (mg/g) 611 23 2
Zinc (mg/g) 43258 1760 105 21 56
Thiamine (B1) (mg/g) 1224 319 5.47.0 4.2 2.5 4.1 10.6
Riboflavin (B2) (mg/g) 1.84.3 1.72.4 2.48.0 1.5 0.2 3.3
Niacin (mg/g) 267499 224389 181550 22.6 1.5
ENZYMES IN RICE BRAN 473
by solvents, and outward movement of the oil from aleurone and germ cells to the
bran layer (28).
The final physical and chemical nature of bran depends on the following:
1. Rice variety
2. Treatment of the grain before milling
3. Type of milling system
4. Degree of milling
5. Fractionation that occurs during milling (29).
The preferred method for milling of rice that gives hulls, bran, and milled rice is
referred to as multistage or multiple break where shellers (dehullers), pol-
ishers, and whiteners are used. The hull is first removed in shellers, and the dehulled
brown rice undergoes subsequent whitening operations. The amount of contami-
nants in the bran affects the total lipid content. Contaminants are broken rice and
layers from the endosperm. Addition of calcium carbonate, usually at 0.25% of
rough rice as a milling aid during whitening, further reduces the oil content. Other
milling aids such as diatomaceous earth and ground limestone have also been
used.
In developing countries, most rice is milled in a one-stage (huller) mill that
removes hull, bran, and germ as a single mixture. It is estimated that less than
25% of rough rice is fractionated into hull and bran fractions (29).
4. ENZYMES IN RICE BRAN
Rice bran contains active enzymes (30). Germ and the outer layers of the caryopsis
have higher enzyme activities. Some enzymes that are present include a-amylase,
b-amylase, ascorbic acid oxidase, catalase, cytochrome oxidase, dehydrogenase,
deoxyribonuclease, esterase, flavin oxidase, a and b-glycosidase, invertase, lecithi-
nase, lipase, lipoxygenase, pectinase, peroxidase, phosphatase, phytase, proteinase,
and succinate dehydrogenase.
Particularly lipase, but also lipoxygenase and peroxidase, are probably most
important commercially because they affect the keeping quality and shelf life of
rice bran.
Lipase promotes the hydrolysis of the oil in the bran into glycerol and free fatty
acids (FFA) (5). The lipase has been studied extensively. In the intact grain, the
lipases are localized in the testa-cross layer of the rice grains while the oil is in
the aleurone and subaleurone layers and in the germ (26, 26a). The germ, where 60%
of the lipase occurs, is similarly compartmentalized. During milling, the enzyme
and substrate are bought together. The rate of FFA formation is highly dependent
on environmental conditions. Formation of 57% free fatty acids per day has been
reported (29). Up to 70% FFA has been reported for a single month of bran storage.
Production of FFA in a clean U.S.-produced bran is shown in Figure 5 (31). Rice
474 RICE BRAN OIL
90
80
70
60
FFA (%)
50
40
30
20
10
0
20 40 60 80 100 120 140
Time (number of Days)
Figure 5. Free fatty acid (FFA) increase in raw bran during a 135-day storage period (8).
bran oil contains 24% FFA at the time of milling. Less than 5% FFA is desirable
for producing rice bran oil because high FFA results in high refining losses. The
composition of crude rice bran oil produced by hexane extraction of stabilized
bran is shown in Table 8.
Lipase has a molecular weight of about 40,000 Da and an isoelectric point (pI) of
8.56 (32). It is activated by calcium and inhibited by heavy metals. The optimum
pH is 7.58.0, and the optimum temperature is 37 C. It is inactivated by heating at
60 C for 15 minutes. Rice bran lipase preferentially hydrolyzes fatty acids from the
TABLE 8. Crude Rice Bran Oil Composition (8).
Lipid Type Percent
Saponifiable lipids 9096

Neutral lipids 8889
Triacylglycerols 8386
Diacylglycerols 34
Monoacylglycerols 67
Free fatty acids 24
Waxes 67
Glycolipids 67
Phospholipids 45
Unsaponifiable lipids 42
Phytosterols 43
Sterol esters 10
Triterpene alcohols 28
Hydrocarbons 18
Tocopherols 1
STABILIZATION OF RICE BRAN 475
1 and 3 positions in the triacylglycerol molecules. Two subunits are suggested for
lipase, and these are held together by disulfide bonds.
A second rice bran lipase has a pI of 9.1 and an optimum temperature of 27 C
(33). It has a high specificity for triacylglycerols having short-chain fatty acids.
The enzyme, lipoxygenase, is associated with the oxidation of the polyunsatu-
rated fatty acids (PUFA) having a cis, cis-pentadiene structure. The carbonyl pro-
ducts from the degradation, particularly hexanal, have been implicated in the stale
flavor of rice. Lipoxygenase activity is highest in the germ fraction. Three forms of
lipoxygenase have been isolated differing in pH optimum and specificity (34).
5. STABILIZATION OF RICE BRAN
The instability of rice bran has long been associated with lipase activity (35). As
long as the kernel is intact, lipase is physically isolated from the lipids (29).
Even dehulling disturbs the surface structure allowing lipase and oil to mix. Oil
in intact bran contains 24% free fatty acids (2). Once bran is milled from the ker-
nel, a rapid increase in the FFA occurs. In high humidity storage, the rate of hydro-
lysis is 510% per day and about 70% in a month as shown earlier. The objectives
of rice bran stabilization are as follows:
Arrest lipase and lipoxygenase activity.

Improve oil extraction efficiency.
Reduce fines in crude oil.
Sterilize the bran.
Reduce color development.
The lipoxygenase and peroxidase enzymes also have a negative impact on the
oxidative state of the bran (Table 9). Further degradation of the oil occurs as
reflected in an increase in peroxide and thiobarbituric acid value and a decrease
in iodine value. Both lipoxygenase and peroxidase enzymes are inactivated with
lipase inactivation.
TABLE 9. Changes in the Composition of Bran Lipids During Storage of Milyang 23 Rice
Bran at 30 C and 80% RH (28).
Storage Period (weeks)
Oil Property 0 1 2 3 4 5
Free fatty acids (% as oleic acid) 3.6 33.0 40.3 45.8 61.8 68.2
Peroxide value (meq/kg) 32.8 73.2 96.0 109.3 90.6 91.0
Iodine value (%) 96.8 90.2 85.4 83.2 79.0 74.7
TBAa (mg of malonaldehyde) 0.5 0.8 1.1 0.7 0.7 0.6
equivalents/Kg
a
Thiobarbituric acid.
476 RICE BRAN OIL
Lipase activity results in hydrolytic rancidity. There is little or no change in

flavor of the bran with an increase in FFA (5). Lipoxygenase activity, however,
increases with the presence of FFA resulting in oxidative rancidity (36). It is
oxidative deterioration that is responsible for the flavor and odor of rancid rice
bran.
Peroxidase is used as a convenient index of lipase activity. The inactivation tem-
perature for lipases and associated enzymes is dependent on the moisture content.
At 4% moisture, inactivation temperature for lipoxygenase is 40 C, lipase is 55 C,
and peroxidase is 70 C (28).
Methods for stabilization of rice bran have been reviewed (37). These include
dry heating, wet heating, and extrusion. The most practical method has been the
use of extrusion or expansion methods.
In retained heating methods (dry heat), a simple hot air drying reduces the moist-
ure content to 34%. The bran must be kept dry in moisture-proof containers, or the
rehydrated bran will regain its lipase activity. If the bran is heated in the presence of
moisture, the lipase is permanently denatured.
The types of retained-moisture heating methods include extrusion cookers and
sealed rotating drums. Extrusion cooking results in both lipase denaturation and
bran sterilization. When pressure is released, part of the superheated moisture eva-
porates with little or no drying being required. Expanders or expellers are also used
to permit addition of moisture (wet heating) through steam and the formulation of
colletts or pellets from the bran. The colletts aid handling and oil extraction.
Extrusion (dry heat) cookers have been ideal for stabilization because excess
moisture is not added, eliminating the need for drying. The heating of the bran
occurs through conversion of mechanical energy of the screw drive to heat the
bran. Temperatures used for stabilization vary from 100 to 140 C. The bran is
kept hot for 35 minutes after extrusion to ensure lipase inactivation. The hot
bran is then cooled using ambient air.
Extrusion cooking of the bran was pioneered by the Western Regional Research
Laboratory (28, 29, 29a). Dry extrusion was found more suitable for stabilizing
bran to be used as a food ingredient (38). Stabilization within 1 hour after milling
is considered ideal for bran quality.
Wet heating is more effective for bran stabilization for oil extraction than is dry
heating. Lipase is inactivated in 3 minutes at 100 C (37). The equipment that can be
used include steam cookers, blanchers, autoclaves, and screw extruders with
injected steam and water (30). Extrusion with steam injection and up to 10% added
water reduces the temperature required for lipase inactivation. Temperatures are
reduced to 100120 C. Product may be held at 100 C for 1.53.0 minutes before
drying to a stable moisture content. Bran expands as it exits the extruder, and water
flashes to steam (8). Porous pellets assist in solvent percolation during oil extrac-
tion. Fines are agglomerated as well.
Addition of water/steam to bran during wet extrusion requires drying after sta-
bilization. Hot air is simply passed through a bed of pellets. Although this increases
the cost of stabilization, lipase inactivation is permanent with less nutritional
damage to the bran. The recovered oil is lighter in color with lower refining losses.
RICE BRAN TO RICE BRAN OIL 477
The stabilized bran may be stored for extended periods, although extraction should
be completed within 1 month for best quality oil (39).
Parboiling of rice is also an example of wet heat stabilization. The lipase in
rough rice is completely inactivated by either autoclaving for 320 minutes or by
parboiling.
Other stabilization methods that have been investigated are as follows:
1. Refrigeration to reduce the rate of hydrolysis (8)

2. Lowing pH to reduce lipase activity (4)
3. Chemical additions such as sodium metabisulfite (39a)
6. RICE BRAN TO RICE BRAN OIL
Rice bran is the source of rice bran oil (30). Various commercial efforts to extract
the oil have been made over the past 50 years. Initially, use of the oil in traditional
foods was targeted. More recent efforts have emphasized the nutritional benefits of
rice bran oil.
Rice bran oil with a low free fatty acid content can be extracted with hexane
from extrusion stabilized bran. The process flow is shown in Figure 6. Nonstabi-
lized bran, although having a high free fatty acid, can also be used for production
of oil. With nonstabilized bran, the extraction is similar to that of extracting a fine
powder. Preprocessing of the bran through an extruder, expander, or expeller may
be used to form either a flake or pellet that results in improved solvent flow through
an extraction bed (40). Flaked bran with only 712% passing a 25 mesh screen gave
a percolation through a 60-cm bed of 563620 L/m2/min. The oil extraction rate
Rice bran
Stabilization
Pellets
Hexane extraction
Desolventizing
Crude rice bran oil
Figure 6. Process for rice oil production (8).

478 RICE BRAN OIL
was rapid, with 96% of the oil being removed in 5 minutes and only 0.7% residual
oil remaining after 1 hour of extraction.
Earlier methods to recover the oil used hydraulic pressing (28). In a Japanese
system for pressing, the raw bran is cleaned by sifting and air classification to
remove whole and broken grains and hulls, and, in some instances, to recover rice
germ. The bran is then steam cooked, dried, prepressed, and finally expeller pressed.
Hexane extraction may be batch, battery, or continuous type (12). All three sys-
tems were recently operating in Japan. Continuous systems operate in Brazil,
Burma, Egypt, India, Mexico, Taiwan, Thailand, and the United States. The bran
in the most efficient systems is stabilized, pelletized, and, if required, dried. After
the pretreated bran is placed in the extractor, hexane is pumped in and allowed to
percolate through the bran to extract the oil. Countercurrent extraction is used.
The miscella (solvent plus oil) is passed through filters to remove the bran fines
before evaporation for solvent and crude oil recovery. The production of fines
from expander stabilized bran depends on stabilization condition (38). Flake size
is larger if expanded at 120 C, but the flakes are fragile and easily broken. Flakes
with high moisture content were more resistant to breakage. Final bran moisture
was about 6%.
Pelletizing of the bran improves percolation and minimizes fines in the miscella.
Pellets are 68 mm in diameter. Moistening during palletizing reduces the fines pro-
blem. Parboiled bran does not produce the hard pellets found for raw bran possibly
because of protein denaturation during parboiling (33). Binding of the fines in the
pellet is assisted by starch gelatinization during heating of the bran. Parboiled bran
also presents problems with sticking to dryer surfaces resulting in self-ignition in
the dryer. Prior mixing with raw bran alleviates the problem.
The X-M process combines solvent extraction and milling of the rice (41).
Brown rice is pretreated with warm rice oil (0.5%) for 23 hours to soften the
bran. The rice is then milled in the presence of a rice oil miscella. The solvent slurry
is then removed from the rice and the rice oil is recovered. Advantages are that sta-
bilization is not required and the resultant oil had a minimum FFA level. This pro-
cess is no longer used.
Extraction of rice bran oil by supercritical fluid has been investigated (5). Minor
reductions in oil yield may occur. The oil yield with supercritical CO2 is 17.98%,
with CO2ethanol 18.23%, and with hexane 20.21%.
7. REFINING OF THE OIL
The color of crude rice bran oil is dark greenish brown to light yellow depending on
the condition of the bran, extraction method, and composition of the bran. The pig-
ments include carotene, chlorophyll, and Maillard browning products (12, 28). Oil
from parboiled rice bran is generally darker in color than oil from raw rice bran.
The composition of crude rice bran oil has a major effect on refining. The crude
oil typically contains up to 0.5% bran fines and 0.55% wax. Agitated storage tanks
are required. Heated tanks and lines also are necessary to prevent crystallization of
DEGUMMING AND DEACIDIFICATION 479
waxes. Refining losses may be in excess of ten times the FFA when the crude oil has
a relatively low FFA (<10%). Lower refining losses of approximately two times the
FFA have been reported (2, 6, 40).
Refining of crude rice oil involves dewaxing, degumming, neutralization of free
fatty acids, bleaching to improve color, and steam deodorization. Refined rice bran
oil is a light yellow color (Lovibond 3.0 R 30Y) with a mild background odor and
flavor reminiscent of rice. Similar to peanut oil, the flavor and odor are complemen-
tary to the flavor of many fried foods, such as fish, chicken, and chips.
8. DEWAXING
Waxes can increase refining losses (8). The wax content of crude oil depends on the
variety of rice, milling technique, method of oil extraction, and extraction tempera-
ture (2). Extraction temperature affects both the type of wax present and its quantity
(42). For example, extraction at 50 C yields two to three times more wax than
extraction at 20 C.
Initial dewaxing may simply be gravity settling followed by decanting (43). The
oil is gradually cooled to allow for wax crystallization followed by filtration or cen-
trifugation to recover the wax sludge. The foots recovered may be added back to the
defatted bran, sold as an animal feed oil, or further processed for oil recovery and
wax purification. Wax recovery involves acetone washing and fractionation with
isopropanol.
The characteristics of the wax are as follows:
Iodine value 11.117.6

FFA (%) 2.17.3
Phosphorous (%) 0.010.15
M.P ( C) 75.379.9
Attempts have been made to recover the wax using cold and hot extraction (2).
Wax yields of 1.291.82% of the crude oil are obtained. Continuous dewaxing of
rice bran oil by chilling the oil or miscella to less than 20 C followed by filtration
through plate and frame filters is practiced. Kinsey and Hummell (44) reported on
the use of sodium silicate as an aid for dewaxing. The characteristics and physical
properties of a purified rice bran wax are similar to carnauba wax (45).
Additional dewaxing may be used during degumming and alkali refining (8).
Dewaxing of refined, bleached oil by cooling to approximately 5 C followed by
filtration is necessary for production of a high-grade, chill-proof oil.
9. DEGUMMING AND DEACIDIFICATION
The phospholipids in rice oil are similar in composition to other oil sources. These
may be recovered as rice lecithin (5). Production of food-grade lecithin requires
480 RICE BRAN OIL
prior removal of bran fines and waxes. Regular water degumming may be used.
Temperatures above 80 C are required to prevent crystallization and removal
of waxes with the gums. If food-grade lecithin is not being produced, filtration
of bran fines is not required. Pretreatment with phosphoric or organic acid is
necessary to remove nonhydratable phospholipids. Food-grade surfactants may
be added to improve wax removal (46). Degumming at less than 50 C actually
assists in wax removal. Wet gums may be added to defatted bran as a method
for disposal (8).
Both alkali and physical refining have been used for FFA removal (5). With alkali
refining, batch or continuous methods may be used. Oil may be pretreated with
phosphoric or organic acid for phospholipid hydration. The oil is then treated
with 1630 baume (Be0 ) caustic with 20 40% excess. The soaps settle and may
be recovered as soapstock or foots (47).
Continuous refining consists of in-line mixers, heaters, and centrifuges (8). The
combined oils plus alkali are rapidly heated to 5570 C to assist in breaking
the emulsion of hydrated soap in oil. In instances where neutralization is com-
bined with dewaxing, separation is performed at 2832 C. Water washing or
post-neutralization treatment with silicates to remove final traces of soaps and phos-
pholipids is the same as for conventional oils. Miscella refining, or refining while
still in solvent, may also be used (47). Higher refining yields and good-quality
neutralized oil with less color are advantages of miscella refining. Losses were
near the calculated amount (48) based on titrated values. Rice oil miscella is often
variable.
Excessive losses may occur in refining of rice oil. A 5% FFA crude oil has losses
ranging from 12% to 40% by the cup method. The cause of high refining losses is
unknown. It is assumed the losses are caused by the presence of partial esters, oxi-
dized components, and waxes, as well as high FFA acidity (8). Steam refining is
practiced by various refineries in Japan and the United States (2).
In calculating the amount of caustic required for caustic neutralization, the oil is
titrated to a phenolphthalein end point. This titration endpoint includes not only the
FFA, but also the oryzanol compounds. With the higher caustic addition, the ory-
zanol is transferred to the soapstock away from the oil. The nutritional benefit of
these compounds is lost. An alternative indicator for titration uses alkali blue (8).
This indicator reflects the acidity contributed only by the free fatty acids.
10. BLEACHING, HYDROGENATION AND DEODORERIZATION
Standard methods are used for bleaching, hydrogenation, and deodorization of rice
bran oil. Bleaching uses activated carbon or bleaching earth (47). Activated carbon
is seldom used because of high cost and handling difficulties. Bleach clay dosage
depends on the characteristics of the rice oil as well as that of the bleaching earth.
Dosages range from 2% to 10%. Newer silica bleaching earths are more effective in
reaching satisfactory oil colors.
CO-PRODUCTS FROM PROCESSING 481
Deodorization or steam stripping is used to remove objectionable odors resulting

from peroxides, aldehydes, and ketones as well as characteristic rice oil odors
and flavors (12). The oil is heated to 220250 C under 35-mm Hg vacuum.
Semicontinuous deodorizer units are the most common types used. Other designs
have been evaluated (43). After deodorization, the oil is cooled to 60 C and filtered.
Storage of deodorized rice ban oil is the same as for other oils.
Physical refining, also called steam refining, combines deacidification with deo-
dorization. Physical refining is more efficient for high FFA oils giving better yields
of neutralized oil than alkali refining (2).
11. WINTERIZATION
In addition to wax removal, rice bran oil contains sufficient saturated and high melt-
ing glycerides to require winterization to gain a cold test of 5 hours (8, 43). Without
winterization, dewaxed rice oil is frequently cloudy or turbid even at room tempera-
ture or slightly lower.
Winterization consists of cooling the oil under defined rates and to specific tem-
peratures followed by filtration. With rice oil, winterization consists of cooling 30
35 C oil slowly at a uniform rate to 15 C over a 12-hour period with slow agitation,
then further cooling to 45 C without agitation followed by holding over a 24
48-hour period, allowing higher melting components to crystallize. The type of
crystals formed depends on the cooling rate and the temperature differentials.
Large, stable crystals are desired for filterability. Filter aids may be added to assist
separation of the crystals from the viscous oil. Cold tests of the winterized oil of
57 hours are near maximum.
Miscella winterization more effectively separates the high melting solids from
rice oil. Hexane, acetone, and isopropyl acetate are among the solvents used.
The miscella is slowly cooled to 15 C over 12 hours with agitation, then to 4
5 C without agitation, and held for 2448 hours before filtering.
12. CO-PRODUCTS FROM PROCESSING
As with all oils, coproducts of refining represent a significant revenue stream.

Waxes may be concentrated and refined to compete with other organic waxes.
The hard, high melting waxes are preferred for most applications.
Soapstock contains fatty acid soaps and, for oil that is caustically refined, ory-
zanol (510%). The soaps may be acidulated for feed use and the oryzanol isolated
(16). Diethyl ether, alumina chromatography, and crystallization are used for pur-
ification of the oryzanol.
The deodorizer distillate, about 1% of deodorizer feed, contains tocopherols,
tocotrienols, and sterols (Table 10). The tocols are shown in Table 11 and the sterols
in Table 12. Its value is similar to other oil distillates.
482 RICE BRAN OIL
TABLE 10. Rice Oil Deodorizer Distillate Composition (20).
Component Percent (range)
Free fatty acids 2540

Tocopherols 1.53.0
Tocotrienols 4.06.0
Sterols 1525
Squalene 1525
Monoacylglycerols, diacylglycerols, etc 1525
TABLE 11. Approximate Tocol Composition of Rice Oil

Distillate (20).
Tocol (percent)

Tocopherol Tocotrienols
Alpha 67 21
Beta 3 tr.
Gamma 30 77
Delta tr. 2
TABLE 12. Sterol Composition of Rice Oil

Deodorizer Distillate (20).
Sterol Percent
Beta-sitosterol 38
Stigmasterol 18
Campesterol 13
Delta-7-stigmasterol 10
Delta-7-avenasterol 6
Delta-5-avenasterol 5
Others 10
13. COMPOSITION OF REFINED RICE BRAN OIL
A typical specification for finished rice bran oil is shown in Table 13. These are
similar to that for other oils. Rice bran oil has a characteristic nutty, earthly flavor
not unlike peanut oil.
The fatty acid composition of rice bran oil is most similar to peanut or ground
nut oil (Table 14) (8). Palmitic, oleic, and linoleic acids make up more than 90% of
the fatty acids present. The major molecular species of triacylglycerols are palmi-
tic-linolenic-oleic, oleic-linoleic-palmitic, palmitic-linoleic-linoleic, linolenic-lino-
leic-palmitic, and trioleic. As with peanut oil, rice bran oil is most suited for general
frying and cooking applications.
RICE BRAN OIL NUTRITION 483
TABLE 13. Product Specification of Refined, Bleached,

and Deodorized Rice Bran Oil (8).
Characteristic Value
Iodine value (Wijs method, g/100 g sample) 99108

Peroxide value (meq/kg) 1.0 max
Moisture (%) 0.05 max
Color (5.25-in Lovibond red) 5.0 max
Free fatty acid (% as oleic) 0.05 max
Flavor/odor 7 min
Chlorophyll (ppb) 75 max
Saponification value 180190
Unsaponifiable matter 35
Smoke point 213 C
Refractive index 1,4701,473
Specific gravity 0.916
AOMa (hr) 17.5
a
Active oxygen method.
TABLE 14. Chemical Composition of Rice Bran Oil (8).
Physicochemical Parameters Value
Acid value 1.2

Iodine value 100.0
Saponification value 211.8
Unsaponifiable matter 4.2
Fatty acid composition Percent
C14:0 0.6
C16:0 21.5
C18:0 2.9
C18:1 38.4
C18:2 34.4
C18:3 2.2
C20:0
C22:0
14. RICE BRAN OIL NUTRITION
The initial interest in rice bran oil resulted from work with the stabilized rice bran.
Rice bran was shown to be equivalent in serum cholesterol reduction to oat bran in
hamster trials (Table 15) (1). Two clinical studies showed rice bran reduced serum
low-density lipoprotein (LDL) cholesterol in humans (49,50). Defatted bran was
less effective in lowering cholesterol than full fat bran (1). The cholesterol-lowering
activity was concentrated in the unsaponifiable fraction of rice bran oil (Table 16)
(51). Oryzanol was found to contribute to the hypocholesterolemic activity of rice
484 RICE BRAN OIL
TABLE 15. Effect of Rice and Oat Brans on Serum Cholesterol

in Hamsters (1).
Bran in Diet Serum Cholesterol (mg/dL)
Cellulose (10%) 395

Rice bran (47.8%) 270
Defatted rice bran (24.7%) 347
Parboiled rice bran (31.8%) 297
Defatted parboiled rice bran (19.6%) 377
Oat bran (53.7%) 289
TABLE 16. Hypocholesterolemic Activity of Unsaponifiable Matter of Rice Bran Oil

in Rats (51).
Serum Cholesterola (mg/dL)

Diet Total HDL LDL VLDL
Control (peanut oil) (10%) 374 43 331

Rice bran oil (10%) 228 48 240
Control 0.2% unsaponifiables 387 48 339
Control 0.4% unsaponifiables 243 48 195
a
HDL high-density lipoprotein; LDL low-density lipoprotein; VLDL very low-density lipoprotein.
oil in rats (52) and primates (53). A clinical study with 3.1 g/day of rice bran oil
unsaponifiables over a 12-month period resulted in a 14.1% reduction in total
cholesterol and a 20.5% reduction on LDL-cholesterol (Table 17) (54). HDL-
cholesterol rose, and triacylglycerols decreased significantly. Tocotrienols, also pre-
sent in rice bran oil, have been reported to reduce serum cholesterol (55).
The refining method used in rice oil production affects the oryzanol content of
finished oil (3). With alkali refining, most of the oryzanol is removed (Figure 7),
whereas with steam or physical refining, most of the oryzanol (66%) remains in
TABLE 17. Effect of Daily Addition of Rice Bran Unsaponifiables (RBN) on Serum Lipids
(mmol/L) in Hypercholesterolemic Subjects (54).
Serum Lipidsa Start 12 months p
RBN
Cholesterol 6.18 0.33 5.31 0.20 <0.05
LDL cholesterol 4.28 0.37 3.40 0.18 <0.05
HDL cholesterol 0.17 0.02 0.24 0.02 <0.025
Triacylglycerol/HDL 2.16 0.35 1.21 0.21 <0.05
RBN placebo
Cholesterol 5.70 0.21 6.06 0.32 ns
LDL cholesterol 3.95 0.18 4.05 0.31 ns
HDL cholesterol 0.21 0.06 0.22 0.01 ns
Triacylglycerol/HDL 1.54 0.31 1.55 0.20 ns
a
LDL low-density lipoprotein, HDL high-density lipoprotein, ns not significant.
RICE BRAN OIL UTILIZATION 485
Oryzanol (ppm X thousand)

20
10
0
Crude Alkali refining Physical refining
Figure 7. Effect of the refining process on the oryzanol content of rice bran oil (8).
the oil (56). Physically refined rice oil gave a serum lipid response similar to that of
crude rice bran oil. Various refining methods to preserve the oryzanol in the oil have
been attempted. Sodium carbonate instead of sodium hydroxide has been partially
successful in which two-thirds of the original oryzanol in the crude oil is preserved
in the refined oil (57). Adding back unsaponifiables to the oil has been patented
(58). Clinical trials have not been performed with high oryzanol rice bran oil. An
unsaponifiable concentrate was prepared by extracting the soapstock, with hexane
giving a deacidified concentrate with 30% unsaponifiable content.
15. RICE BRAN OIL UTILIZATION
Rice bran oil is used in foods, feed, and industrial applications. Only high-quality
oil is targeted to foods. The use of rice bran oil in Japan, where it is the largest
volume domestically produced vegetable oil, is as a frying oil where its flavor is
preferred over alternative oils. The oxidative stability of rice bran oil is equivalent
to peanut oil and cottonseed oils in deep frying applications (Table 18) (8, 59).
TABLE 18. Frying Evaluation of Rice Oil (15-day results) (8).
Days to Maximum Levelb

Oil typea FFA FOS LY LR TPM
Rice (without additives) 3.91 3.74 6 28.0 31.9

Rice (with additives) 5.62 3.46 7 49.6 34.6
Peanut (with additives) 6.87 3.92 8 21.2 37.5
Cottonseed (with additives) 7.22 4.07 7 28.8 37.2
a
Specifications. 40 lb (18.2 kg) gas fryers, frying temperature 350 F (177 C); hourly rotation: breaded chicken,
fish, onion rings, French fries: 5-ppm dimethyl polysiloxane antifoam, 200-ppm tertiary butyl hydroguinone.
b
FFA free fatty acids, FOS food oil sensor; LY Lovibond yellow; LR Lovibond red; TPM total polar
material.
486 RICE BRAN OIL
TABLE 19. Frying Results Using Blends of Rice

and Soybean Oils (8).
Total Polar Material (%)

Oil Type 10 days 13 days
Rice 21.12 32.78

Peanut 21.07 35.53
Rice/soybean 50:50 24.11 35.80
Rice/soybean 25:75 23.25 40.42
TABLE 20. Days at 145 F (62.8 C) Before Rancid Odor

is Detected (8).
Oil Type Days to Detect Rancid Odor
Rice (without additives) 20

Rice (with additives) 25
Peanut (without additives) 14
Cottonseed (with additives) 31
Blends of rice bran oil with soybean oil reduces the increase in total polar material
(TPM) depending on the amount of rice bran oil in the blend (Table 19). Potato
chips fried in rice bran oil show flavor and odor stability at elevated temperatures
between that of peanut and cottonseed oils (Table 20).
Winterized rice bran oil is an acceptable oil for salad dressing and mayonnaise.
The hard fraction of rice bran oil may be used to replace the plastic fats in margar-
ines and shortening. Hydrogenated rice bran oil is adaptable to specialty shorten-
ings and margarines.
The nonfood uses of rice bran oil are feed formulations, soaps, and glycerin.
Waxes may be used as a carnauba wax replacement in confectionery, cosmetics,
and polishing compounds products.
Use of rice bran oil grows as a specialty ingredient in the cosmetic/personal care
market. The demand is for natural, value-added healthy ingredients (60).
16. RICE OIL PRODUCTION (POTENTIAL)
World rice production is greater than 500 million metric tons. Rice oil production is
estimated at 722.2 thousand metric tons (Table 21). India, China, and Japan are the
leading producers. More than half of rice is processed in small rice mills. This
leaves approximately 2025 million metric tons of bran available for oil production.
The rice bran oil potential is, then, 34 million metric tons.
In the United States, most bran is also produced in small rice mills scattered
in rice production areas with insufficient bran production to justify oil extraction.
REFERENCES 487
TABLE 21. Production of Rice Bran Oil (61).*
Country Thousand Metric Tons
Bangladesh 1.5
Brazil 1.5
Cambodia 4.6
China 90.0
India 472.7
Indonesia 0.15
Japan 65.0
Korea 11.7
Republic of Korea 9.2
Laos 2.6
Burma 17.6
Nepal 7.6
Pakistan 3.7
Sri Lanka 5.5
Thailand 7.8
Vietnam 7.6
Total 722.2
*
Does not include U. S. production, which is 15.918 thousand
metric tons.
Production estimates are for less than 80 thousand metric tons. Only 15.9 to 18
thousand metric tons are produced currently in the United States at a single oil
extraction facility.
17. SUMMARY
Rice bran is an underused coproduct of rice milling. The value is partially captured
through extraction and refining of the rice bran oil. The capital costs have limited
the ability of the U.S. rice milling industry to capture this value. However, rice bran
oil has performance properties competitive to other widely used oils. An additional
advantage of rice bran oil is certainly its nutritional benefits, which include a bal-
ance of fatty acids meeting AHA recommendations. Rice oil contains a mixture of
antioxidants and promotes cholesterol reduction beyond that of more unsaturated
oils. Its taste and performance is complementary to salad, cooking, and frying
applications.
REFERENCES
1. T. Kahlon et al., Cereal Chem., 69(5), 485489 (1992).

2. F. T. Orthoefer, presented at 85th American Oil Chemists Soc. Annual Meeting, Atlanta,
GA, May 9, 1994.
3. R. Cheruvanky, in W. R. Bidlack et al., eds. Phytochemicals as Bioactive Agents,
Technomic Press, Lancaster Press, 2001.
488 RICE BRAN OIL
4. R. Cheruvanky, in L. Johnson and G. Williamson, eds. Phytochemical Functional Foods,

Woodhead Publishing Limited, Cambridge, U.K., 2003.
5. F. T. Orthoefer, in Y. H. Hui, ed. Baileys Industrial Oil and Fat Products, 5 ed., vol. 2, J. W.
Wiley and Sons, Inc. New York, 1996.
6. F. T. Orthoefer and J. Eastman, in E. Champagne, ed. Rice Chemistry and Technology,
American Association of Cereal Chemists, St. Paul, MN, 2003.
7. B. O. Juliano, in P. J. Barnes, ed. Lipids in Cereal Technology, Academic Press, London,
1983.
8. F. T. Orthoefer, in R. F. Wilson, ed. Proceedings of the World Conference on Oilseed
Processing and Utilization, AOCS Press, Champaign, Illinois, 2001.
9. J. S. Godber and B. O. Juliano, in E. Champagne, ed. Rice Chemistry and Technology,
American Association of Cereal Chemists, St. Paul, MN, 2003.
10. W. R. Morrison, in Y. Pomeranz, ed. Cereal Lipids, Vol. 2, American Association of Cereal
Chemists, St. Paul, MN, 1978.
11. N. H. Choudhury and B. O. Juliano, Phytochemistry, 19, 13851389 (1980).
12. B. O. Juliano, in P. J. Barnes, ed. Lipids in Cereal Technology, Academic Press. New York,
1983.
13. W. R. Morrison, J. Cereal Sci., 8, 115 (1988).
14. T. S. Shin and J. S. Godber, J. Agric. Food Chem., 44, 567573 (1996).
15. C. Rukmini and T. C. Raghuram, J. Am. Coll. Nutr., 10, 593601 (1991).
16. G. S. Seetharamaiah and J. V. Prabhakar, J. Food Sci. Tech., 23, 270273 (1986).
17. B. J. Lloyd, T. Siebenmorgen, and K. W. Beers, Cereal Chem., 77, 551555 (2000).
18. K. Tanabe, M. Yamaoka, and A. Kato, Jpn. Oil Chem. Soc., 30, 116118 (1981).
19. K. Tanabe et al., Yuk agaku, 31, 205208 (1982).
20. L. Gingras, presented at Amer. Oil Chem. Soc. Ann. Meeting, May 2002.
21. T. S. Shin, J. S. Godber, D. E. Martin, and J. H. Wells, J. Food Sci., 62, 704728 (1997).
22. W. Hu, J. H. Wells, T. S. Shin, and J. S. Gober, J. Am. Oil Chem. Soc., 73, 16531656
(1996).
23. D. E. Martin, M. S. Thesis, Louisiana State University, Baton Rouge, p. 143.
24. Silto, T. Susuki, and Y. Fujino, Nippon Nogei Kagaku Kaishi, 55, 247253 (1981).
25. S. H. Yoon and J. S. Rhee, J. Am. Oil Chem. Soc., 59, 561563 (1982).
26. B. S. Luh, S. Barber, and C. B. deBarger, in B. S. Luh, ed. Rice Utilization, Vol. II, Avi
Book, Van Nostrand Reinhold, New York, 1991.
26a. B. S. Shastry and R. M. R. Rau, Chereal Chem., 53(2), 190200 (1976).
27. S. Danforth and F. T. Orthoefer, presented at 64th Tri-state Oil Mill Assoc. Mtg., Biloxi,
Mississippi, 1989.
28. B. O. Juliano, in B. O. Juliano, ed. Rice: Chemistry and Technology, Amer. Assoc. of
Cereal Chem. St. Paul, Minnesota, 1985.
29. R. M. Saunders, Food Rev. Int., 1, 465495 (1986).
29a. J. M. Harper, in C. Mercier, P. Linko and J. M. Harper, eds. Extrusion Cooking.
American Assoc. of Cereal Chemists, St. Paul, Minnesota, 1989.
30. S. Barber and C. Benedito de Barber, in B. S. Luh, ed. Rice: Production and Utilization,
Avi, Westport, Connecticut, 1980.
REFERENCES 489
31. F. T. Orthoefer and R. J. Nicolosi, presented at 84th American Oil Chemist Society Annual
Meeting, Anaheim, California, Apr. 29, 1993.
32. M. Funatsu et al., Agr. Biol. Chem., 35, 734774 (1971).
33. Y. Aizono, et al., Agr. Biol. Chem., 40, 317324 (1976).
34. S. Ida, Y. Masaka, and Y. Moriton, Agr. Biol. Chem., 47, 637641 (1983).
35. C. A. Brown, J. Am. Chem. Soc., 25, 248954 (1903).
36. T. Galliard, J. Allen and R. Hamilton, eds. in Rancidity in Cereal Products, Elsevier
Applied Science, New York, 1989.
37. R. N. Sayre et. al., Cereal Foods World, 27, 317322 (1982).
38. J. M. Randall et al., J. Food Sci., 50, 361364, 368 (1985).
39. D. R. McCaskill, S. D. Danforth, and F. T. Orthoefer, presented at 23rd Rice Technical
Working Group, Biloxi, Mississippi, 1990.
39a. R. N. Sayre, D. K. Nagyar, and R. M. Saunders, J. Amer. Oil Chem. Soc., 62, 10401043
(1985).
40. J. V. Prabhakar and K. V. L. Venkatesh, J. Am. Oil Chem. Soc., 63, 644646 (1986).
41. J. W. Hunnel and J. F. Nowlin, in D. F. Houston, ed. Rice: Chemistry and Technology,
1st ed., Amer. Assoc. of Cereal Chemists, St. Paul, Minnesota, 1972.
42. C. Kao and B. S. Luh, in B. S. Luh, ed. Rice Utilization, Van Nostrand Reinhold, New York,
1991.
43. S. C. Chang, R. M. Saunders, and B. S. Luh, in B. S. Luh, ed. Rice Production and
Utilization, Avi Publishing, Westport, Connecticut, 1980.
44. D. N. Kinsey and J. W. Hunnell, U.S. Patent 3,481,960, Dec. 2, 1969.
45. R. Lasztity, United Nations Industrial Development Organization, ID/W. G. 89/10,
March 9, 1971.
46. A. Sah, B. K.D Agrawal, and L. S. Shukla, J. Amer. Oil Chem. Soc., 60(2), 466 (1983).
47. A. C. Bhattacharyya, S. Majumdar, and D. K. Bhattacharyya, J. Amer. Oil Chem. Soc., 63,
11891191 (1986).
48. P. B. V. Reddi, K. S. Murti, and R. O. Feugi, J. Amer. Oil Chem. Soc., 25, 206211 (1948).
49. M. Hegsted, M. Windhauser, and S. Lester, Faseb, 4, 368A (1990).
50. A. Gerhardt and N. B. Gallo, Food Chem. News, 31, (1989).
51. R. Sharma and C. Rukmini, Indian J. Med Rev., 85, 278281 (1987).
52. R. Sharma and C. Rukmini, Lipids, 21, 715720 (1986).
53. R. J. Nicolosi et al., in J. Wadsworth and W. Marshall, eds. Rice Science and Technology,
54. T. R. Watkins et al., Environ. Nutr. Interact., 3, 115122 (1990).
55. A. Qureshi et al., Amer. J. Clin. Nutrition, 53, 1021510265 (1991).
56. A. G. Krishna et al., J. Amer. Oil Chem. Soc., 78(2), 127131 (2001).
57. L. Gingras, Inform II, Nov, 2000.
58. H. Hitasumatsu and Y. Takeshita, U.S. Patent 5, 290, 579 (1994).
59. L. Gingras, D. Hutchinson, and D. McCaskill, presented at Amer. Oil Chem. Soc. Annu.
Meeting, May 2003.
60. D. DeGuzman, Chemical Marketing Reporter, 263, 12 (2003).
61. Food and Agriculture Organization of the United Nations (FAOSTAT 1998).
11
Safflower Oil
Joseph Smith
1. HISTORY AND BOTANICAL DESCRIPTION
Safflower, Carthamus tinctorius L., has a long history of cultivation. Some would
class it as the worlds most ancient crop (1); others feel that olives, dates, and
sesame predate safflower (2). Safflower was produced in Egypt more than 4000
years ago (3), but the most likely area of its origin is in the Euphrates basin
(46). From there it apparently was introduced into Egypt and Ethiopia. Muslim
traders carried safflower seeds across the northern coast of Africa and into present
day Spain, while Arabs introduced it into many parts of east Africa. By the six-
teenth century, safflower was found in several parts of Europe. Turks carried saf-
flower into all parts of the Middle East, from where it spread to Iran,
Afghanistan, and India. From Afghanistan it spread into China more than 2000
years ago (7). It spread to Japan in the third century A.D. (8). Spanish and Portuguese
conquerors brought safflower to the New World, and later emigrants from Portugal
and Russia did the same (9). For much of its history, safflower was used primarily as
a source of dye, a food coloring, a cosmetic, or for medicinal purposes (7, 10, 11).
Dried safflower florets are commonly used as an adulterant or substitute for colorful
saffron, Crocus sativus L., a much more costly spice (1214). Production of saf-
flower oil was carried out in the reign of Ptolemy II (10), and Pliny pointed out
that it could be used as a substitute for castor oil for nonedible purposes (15). While
it had become known as an edible oil during pre-Christian times in Mesopotamia
(16), it was only in more recent times that it began to be used in India as an
491
492 SAFFLOWER OIL
edible oil, and of course, it was not until the middle of this century that it began to
enter world commerce, first as an industrial oil and then as an edible product (17).
Because safflower was introduced to many lands, it is known by a number of
different names, some of which are azafrancillo, bastard saffron, benihana, cartamo,
cnikos, false saffron, ghurtom, hung hua, kafsha, kahil, kajireh, kardi, khardam,
kusumba, onickus, safflor, thistile saffron and ssuff (1, 3).
The safflower plant is a member of the Compositae family. Other members of
this family are the artichoke, chrysanthemum, niger, and sunflower. There are at
least 25 species of the Carthamus genus that grow in the wild (18), but only C. tinc-
torius, which we call safflower, has been domesticated; some quantities of
C. oxyacantha have been gathered and used as oil or food sources in India and
Pakistan (19).
The safflower plant as we know it resembles the Scottish thistle but has yellow,
orange, or red florets rather than the purple bloom of the thistle. However, the com-
mercial species of safflower, C. tinctorius, does not become a weed. The plant
grows to a height of 30150 cm, develops many branches (unless affected by nat-
ural or artificial environmental conditions), and develops a thickened taproot that
can extend down to 4 m.
Each branch terminates in an inflorescence which is a dense capitulum of florets
(individual tubular corollas), commonly called a flower. Each floret flower pro-
trudes from a conical head surrounded by layers of bracts. The leaves, which
develop along the stalk and branches, and the outer layers of bracts usually are
spiny, although the types of safflower grown for the production of dye or food
coloring are spineless, or nearly so. The seeds of the safflower plant develop within
the head in a concentric pattern and are oblate with a flattened top, usually white,
and about the size of a barley kernel (Figure 1) (20).
Safflower is a plant of desert origins, as evidenced by its deep taproot, waxy
leaves, and relatively thick hull. It responds well to moisture and nitrogen. Its
seed has the ability to germinate almost immediately if exposed to moisture at
the proper temperature, unlike a sunflower seed, which must go through a period
of dormancy before germination. The deep root and the many fine laterals that
extend from it have the ability to seek out water and nutrients deep in the soil.
These properties, while they allow safflower to survive in periods of moisture short-
age also limit the areas of the world where safflower can be cropped successfully.
Safflower is normally planted after soil temperatures exceed 4.5 C and does not
begin growing fast until temperatures exceed 15 C. In the interim period, it goes
into a rosette stage after emergence. During this time, it establishes its deep root
system. As temperatures increase, the stem of the plant begins to elongate and
can grow as much as 2.5 cm per day, until maximum height is attained. Branches
and buds form until the plant flowers, after 70 days or more (depending on tempera-
ture at planting time). Flowering can last from 10 days to 3 weeks, and the crop
usually is ready for harvest 45 days after time of full flower.
Flowering normally takes place during the warmest part of the growing season.
If a protracted period of rainfall occurs at the same time, or until harvest time,
unharvested safflower seeds still in the head will germinate and begin to form
HISTORY AND BOTANICAL DESCRIPTION 493
Figure 1. a, Dr. Carl E. Claassen, father of the modern-day safflower, among fully branched
safflower. b, Safflower blossom. c, Safflower seed.
sprouts, which quickly can reduce oil content, increase color and FFA of the
contained oil, and eventually result in total loss of the crop. Therefore, to grow
safflower successfully, it must be planted in regions that have a minimum 120
frost-free growing days, 300500 mm, of annual rainfall or irrigation, and that do
not experience rainfall during the period when safflower is in flower or thereafter.
Most of the farming areas of the world receive some summer rains. If rain
occurs, safflower has a chance of surviving, but this greatly increases the chance
of the plant being attacked by various leaf and head molds, which can limit yield
severely. So safflower production is limited to areas such as Californias central val-
ley and southern Arizona; isolated areas of Mexico and Australia; and the drier
parts of China, India, and the Middle East. Areas where safflower can be grown,
but with greater risk, are U.S. Northern and Great Plains, southern Idaho, and north-
ern Utah; much of northern Mexico; far northern Argentina; and the drier parts of
India and China (Figure 2).
494 SAFFLOWER OIL
Figure 2. Areas in the United States that can support safflower production.
Safflower does not require any specialized equipment to be farmed successfully.

In developed parts of the world, 1570 kg of seed are planted per hectare, with the
lower ranges being planted either because moisture is a limiting factor or because
the crop is to be managed in cultivated rows. In areas with only 300400 mm of
annual rainfall, 15 kg of seed are planted with a grain drill, much as a crop of wheat
would be picked. In an area where plentiful irrigation water is available, a 20 kg of
seed may be planted per hectare. Seed may be planted in three or four drilled
or precision-planted rows on an elevated bed; the groups of rows are spaced
5060 mm apart. In areas where moisture is plentiful, 3570 kg of seed may be planted
per hectare; the higher rate is used to ensure that weeds do not gain a competitive
edge. Normal dates of planting in the United States range from December to March
in Arizona, January to March in Californias San Joaquin Valley, February to early
May in Californias Sacramento Valley and last half of April to the first half of May
in the rest of the country.
Good practice is to incorporate a trifluran herbicide into medium to heavy loam
soil before planting; the seed is placed at a 35100-mm depth, depending on the
moisture level. A well-incorporated herbicide is necessary where weeds can be a
problem, because safflower is not a good weed competitor in its early stages of
growth. Almost 120 units of nitrogen are necessary to attain maximum production.
For the crop to generate maximum yield, it must receive enough rainfall or have
enough moisture in the soil either through preirrigation or subsequent row irrigation
in order to maintain a bright green color and to prevent drying of its lower leaves
until it is past flowering. As the plant approaches maturity, the flowers dry, and the
entire crop attains a golden brown color.
Harvesting is accomplished with a standard grain combine generally set to run
internally at a slightly lower speed than for grain, which prevents cracking of the
seed. The combine should be set to cut only as deep as is necessary to capture all
heads. Harvesting should not begin until the seed has dried in the head to a level of
8% moisture content or lower. Most safflower is harvested at a 45% moisture level.
In parts of India and China, much of the production is done by hand, and gen-
erally red-flowered, lower oil content, spineless varieties are used. Young people
pass through the fields at time of flowering and pluck the florets from the seed
heads, putting them into purses strung around their necks; the seeds are subse-
quently harvested when the crop has matured and dried, to recover oil. The florets
are carefully dried out of the sun and then used for food coloring or (in China and
Sri Lanka) for the production of either red or yellow dye. In India, some green saf-
flower plants are used as a vegetable (21). In Australia, India, and Pakistan the plant
is occasionally used as a grazing crop or fodder for cattle (2224). After harvesting,
the remaining stubble consists of hollow stalks, dried leaves, empty heads, and
some empty hulls or immature seeds. Sheep and cattle were allowed to graze on
this stubble in the United States during the 1950s; today such grazing is confined
to occasional employment by sheep ranchers. Most fields in Mexico are grazed by
cattle after harvest.
Until the twentieth century, safflower tended to be a local crop. No effort was
expended to find species that had better oil content, since most of the interest cen-
tered around the crop as a medicinal or dye stuff source.
In 1925, the U.S. Department of Agriculture obtained samples of safflower seed
from the then U.S.S.R. and India, and over the next 10 years various agricultural
experiment stations and some farmer cooperatives conducted trials. In 1935, the
USDA produced a circular summarizing the trial results; it concluded that safflower
had possibilities as an oil seed crop in the northern Great Plains and far west (25). A
Montana farmer conducted trials with safflower beginning in 1928 and contacted
paint companies, researchers, and others who might have an interest in safflower
oil (26). Several favorable reports ensued (2729). In 1937, a comprehensive report
based on seed obtained from Montana and elsewhere praised safflower oils good
properties (30). Others in Europe had earlier written favorable evaluations of
safflower as a drying oil (31) and as a source of high protein meal (32). In 1947,
496 SAFFLOWER OIL
a comprehensive report on safflower production and drying oil capability was

published (33).
Cargill, Inc. contracted for and processed about 1000 t of safflower seed in the
U.S. northern Great Plains during 1947 and 1948; the company concluded that the
crop was not sound at that time (34). Two men and one company provided the real
impetus for getting safflower established as a crop in the United States: Claassen,
Knowles, and the Pacific Vegetable Oil Corp. (PVO).
Claassen was employed in 1941 as a research agronomist by the University of
Nebraska to assist the newly created Chemurgy Project in evaluating crops that
could become significant contributors to the state of Nebraska (35, 36). After test-
ing many new crops, he settled on safflower and began a breeding program (9).
Claassen found that most safflower introductions were in the 2229% oil content
range but found introductions from Sudan and Egypt that ranged from 33 to 37% oil
content (37). Claassen began to do selection and breeding work, and by 1949, he
had released several lines and described cultural methods for obtaining relatively
consistent yields (3741). The most important line released, N-852, had an oil con-
tent of 3234% and good yielding ability. Several safflower processing companies
were formed in Colorado and Nebraska to commercialize the new releases, but they
quickly failed (34).
Claassens work came to the attention of Knowles at the University of California
at Davis. Claassen had sent portions of his new lines to a number of western coop-
erators for testing, and results that Knowles obtained were quite exciting (42).
Claassen was encouraged by Hoagland, who was by then living in California, to
come out for a visit to see the potential that safflower had in that state. Claassen
visited California in 1949 and traveled to various oil processing companies. Claas-
sen was convinced to resign from the university and join Hoagland in starting up a
safflower planting seed and promotion concern called Western Oilseeds Co. (9).
Initially, Oil Seed Products Co. of Fresno, California, displayed the most interest
in Claassen and Hoaglands work, but it soon became apparent that PVO could offer
much more help because of its strong background in the production and sales of
industrial oils. The primary interest in safflower oil at the time was coming from
paint companies, whereas most oil millers in California were suppliers to the
food industry. The N-852 variety, although it had good yields, was susceptible to
phytophthora root rot under irrigation. Thus, the first tries at growing safflower
in California in 1950 resulted in severe losses, because of the stress of irrigation.
This turned away many growers and millers in Californias cottonseed production
areas, which helped PVO, because its mill was farther north. Claassen was soon
joined by Hoffman, his former assistant at Nebraska; they formed an alliance
with PVO. After Hoagland departed after a dispute, Pacific Oilseeds, Inc. (POI)
was formed, jointly owned by PVO, Claassen, and Hoffman (34).
The combination of PVO and POI formed a near monopoly, dominating the saf-
flower business until 1962. Approximately 95% of the safflower oil sold during that
period went to the paint, varnish, and coatings market in which PVO had the stron-
gest hand. POIs tie with PVO meant that only growers who contracted their saf-
flower crop with PVO got the best seed as new varieties began to be released. This
system was the real key to PVOs dominance, which was probably the first time that
a nonperishable crop was grown under contract. Growers were given contracts that
guaranteed a floor price for the entire production from a given acreage, provided the
best planting seed available for cash or credit, and offered field service and advice
free of charge. In addition to the floor price the grower received a bonus based on
PVOs profits in marketing the crop.
In ensuing years, PVO began to offer long-term contracts under which the
grower was paid 50% of PVOs safflower profit as measured by the firms auditors
and the grower committed his or her entire production of safflower to PVO for 37
years. POI was also offered bonuses, reflecting 50% of the value of any increase in
oil or protein levels in new seeds it developed. In turn, PVO went to great lengths to
keep Knowles and various farm advisers and university officials interested in saf-
flower. Banks, truckers, and warehouses were fully informed about PVOs future
plans and how the safflower business was doing. In addition to making growers,
researchers, and POI feel that they were part of a partnership aimed at improving
safflowers lot, PVO kept oil and meal prices as low as possible to encourage mar-
ket development. Buyers were also offered long-term requirements contracts that
allowed them to purchase oil more than 1 year ahead of delivery date and, in
fact, committed both parties to an assurance that the customer would always be
able to obtain his or her full requirements. To some consumers, PVO also offered
technical assistance agreements wherein a chemist or engineer was sent to the cus-
tomers plant to help solve processing problems and find new, profitable markets for
safflower products. As markets grew, PVO expanded with them, enlarging proces-
sing capability in California, building new oil mills in Montana and Nebraska, and
expanding production to Spain, Mexico, and Australia (34).
Coinciding with this development was the publication of Calories Dont Count
(43), which became an instant best-seller in 1961. The book advocated a diet that
featured daily capsulized doses of safflower oil. Medical research was beginning to
demonstrate the close relationship between diet and heart disease. Eminent
researchers began to show the relationship of cholesterol to heart disease and,
more important, to show that polyunsaturates such as safflower oil would lower
blood serum cholesterol for many people (4446).
A further development enlarging the market for safflower oil was the increasing
importance of the Japanese vegetable oil market. Japan became a consumer of U.S.
safflower seed because of a fluke in the duty and quota structures set up after World
War II. A duty of 20% was established to protect Japanese soybean producers from
cheap imports. To be consistent, this duty was applied to a list of other oilseeds in
addition to soybeans. Imports were also controlled under a very tight quota system
to protect Japanese foreign exchange. But the framers of the duty structure were
unaware of safflower, which was thus omitted from the list. Japanese oil mills
became large clients for duty-free safflower seed, until the duties were gradually
reduced as Japan became a major importer of soybeans and rapeseed (3). In the
United States, safflower oil was being employed primarily in the coatings industry,
but in Japan it was employed primarily as an edible oil. Safflower became quite
popular in tempura oil blends.
498 SAFFLOWER OIL
During the early years of increasing exports of safflower seed to Japan, PVO did
the lions share of the business and was able to tie up a majority of the Japanese
importers through a series of long-term requirements contracts wherein PVO agreed
to supply the Japanese mills increasing needs each year for an agreement to buy
exclusively from PVO. The increasing Japanese demand and the growing popularity
of safflower oil in the United States fueled PVOs expansion in production first to
the western Great Plains and then to Mexico, Spain, and Australia. PVO began to
lose its monopoly position in 1957 with the public release of the Gila variety of
safflower seed (47, 48). This seed had a good yield and oil content and was also
resistant to phytophthora root rot. As the seed became available, practically every
cottonseed milling company in California and Arizona was able to become a pro-
ducer and supplier of edible safflower oil.
Up to that point, safflower seed production had been on a continued upward spir-
al in the United States, which carried through to 1963. Safflower oil had been price
competitive with soybean oil, particularly in the western United States and Japan,
since soybean oil produced in the Midwest was at a freight disadvantage. The intro-
duction into California of new varieties of wheat developed by the Borlaug program
in Mexico allowed California farmers to achieve increasingly better wheat yields.
In the 1950s, safflower was easily able to compete with wheat or barley as a rotation
crop for Californias rice or cotton farmers, but once wheat yields increased and
safflower yields remained constant, safflower seed prices (and consequently oil
prices) were forced to rise to compete for the farmers favor.
Rising prices for safflower and increasingly better water-based paints formulated
from petroleum-based polymers rather than vegetable oils quickly cut industrial
consumption of safflower oil. PVO attempted to stem this tide by introducing pro-
ducts that combined safflower oil with water emulsion technologies, but it was too
late (4951).
The polyunsaturated bubble almost burst when the U.S. Food and Drug Admin-
istration began attacking refiners claims about the ability of these products to
reduce incidence of heart disease and lower cholesterol, but subsequent supportive
statements by the American Medical Association and the American Heart Associa-
tion softened the effect of the attack.
Increasingly competitive supplies of the former U.S.S.R. and U.S. sunflower
seeds and oil helped erode the international market for U.S. safflower producers.
Safflower oil has the highest level of polyunsaturation of the commercial oils,
and the market has recently stabilized in the United States, northern Europe, and
Japan. In Japan in particular, safflower oil has achieved an increasingly larger share
of the gift pack market, wherein fancy tins of safflower oil are exchanged during the
summer and Christmas gift-giving seasons.
Industrial use of safflower oil has declined to 23% of the total market.
Conjugated safflower oil (52) competes with dehydrated castor or tung oils and
very high quality alkyds. Table 1 illustrates the rise and fall of safflower supply
and disappearance in the United States in comparison with major and minor
crops (53).
TABLE 1. Edible Fats and Oils: U.S. Supply and Disappearance, 106 lb.
Item 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a 2003b
Stocks Octobera
Coconut 277 188 251 164 163 84 150 393 152 136 260 227 148
Corn 138 196 150 118 241 116 129 102 135 267 117 104 114
Cottonseed 137 78 81 106 82 94 66 79 76 49 93 39 40
Lard 24 27 26 34 24 23 20 40 21 18 14 10 5
Palm 53 44 33 35 15 31 46 35 48 48 61 70 42
Palm kernel 53 49 88 73 55 22 51 64 73 49 155 128 50
Peanutc 25 51 50 25 40 65 86 41 40 32 31 32 50
Safflower 28 28 18 31 21 44 27 38 48 36 21 17 19
Soybean 1,786 2,239 1,555 1,103 1,137 2,015 1,520 1,382 1,520 1,993 2,767 2,359 1,486
Sunflower 47 100 56 65 82 147 93 60 121 157 136 23 25
Canola 41 71 67 137 54 77 65 112 169 206 110 52 55
Tallow, edible 41 33 41 36 52 34 48 46 43 40 49 24 35
Imports
Coconut 841 1,163 999 1,100 874 1,188 1,438 791 926 1,115 1,093 860 970
Corn 5 7 7 10 11 14 28 42 18 27 61 65 65
Cottonseed 18 38 26 0 0 0 0 48 8 0 0 22 0
Lard 2 3 3 2 2 1 2 2 2 3 6 10 10
Olive 216 253 262 260 227 304 333 355 397 455 455 485 540
Palm 220 267 368 218 236 322 282 284 345 399 490 425 440
Palm kernel 342 302 304 280 262 392 359 401 393 351 330 470 475
Peanutc 1 0 11 4 5 14 10 73 12 79 39 70 70
Canola 815 861 902 938 1,086 1,075 1,088 1,060 1,139 1,193 1,108 929 1,215
Safflower 22 15 16 26 35 30 51 51 33 34 40 43 45
Soybean 1 10 68 17 95 53 60 83 83 73 46 50 85
Sunflower 9 0 7 1 2 22 8 5 4 8 36 60 5
Tallow, edible 6 10 15 18 8 5 2 3 10 32 7 11 10
Production
Corn 1,821 1,878 1,906 2,227 2,139 2,231 2,335 2,374 2,501 2,403 2,461 2,453 2,650
Cottonseed 1,280 1,126 1,119 1,312 1,229 1,216 1,224 832 939 847 876 725 865
Item 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002a 2003b
Lard 1,016 1,011 1,015 1,052 1,013 979 1,065 1,106 1,069 1,050 1,080 1,075 1,100
Peanutc 356 286 212 314 321 221 176 145 229 179 230 286 219
Canola 32 49 406 299 355 342 451 548 617 641 585 541 629
Safflower 69 87 111 115 127 103 115 111 91 88 76 89 91
Soybean 14,345 13,778 13,951 15,613 15,240 15,752 18,143 18,078 17,825 18,420 18,898 18,435 17,020
Sunflower 911 730 580 1,165 860 840 959 1,177 1,046 873 673 320 595
Tallow, edible 1,515 1,414 1,535 1,550 1,559 1,407 1,517 1,677 1,792 1,764 1,932 2,075 2,000
Exports
Coconut 22 0 19 18 12 12 6 11 14 8 7 8 10
Corn 566 712 717 865 977 988 1,118 989 970 951 1,172 890 900
Cottonseed 269 184 248 329 221 232 208 111 141 131 150 110 115
Lard 131 129 119 140 94 103 122 140 189 93 90 105 100
Olive 20 15 11 21 24 21 19 15 12 9 10 12 12
Palm kernel 2 9 4 2 2 2 2 2 2 2 2 2 2
Palm 7 7 7 13 20 9 11 11 11 11 10 11 10
Peanutc 151 52 61 97 108 21 13 10 18 14 8 42 19
Canola 15 16 76 153 147 295 349 272 284 187 255 166 157
Safflower 73 65 75 93 122 83 83 92 51 35 37 37 40
Soybean 1,644 1,461 1,531 2,683 992 2,033 3,079 2,372 1,375 1,401 2,519 2,250 850
Sunflower 471 586 450 978 628 709 815 800 630 545 453 110 200
Tallow, edibled 333 306 316 277 241 181 236 322 224 338 475 485 490
Domestic disappearance
Coconut 910 1,084 1,067 1,083 941 1,111 1,189 1,021 927 983 1,119 930 958
Corn 1,202 1,220 1,228 1,250 1,298 1,244 1,271 1,394 1,417 1,630 1,363 1,618 1,804
Cottonseed 1,088 975 873 1,007 996 1,012 1,004 772 833 672 780 636 750
Lard 885 886 890 924 922 880 925 987 886 964 1,000 985 990
Olive 216 253 262 260 227 304 333 355 397 455 455 473 528
Palm 223 271 359 225 201 298 282 260 335 375 471 425 427
Palm kernel 344 254 315 295 293 362 344 390 414 243 355 511 458
Peanut 179 236 187 206 193 194 217 208 233 244 260 296 275
Canola 801 898 1,162 1,165 1,271 1,134 1,143 1,287 1,435 1,744 1,496 1,301 1,687
Safflower 15 47 40 57 17 67 73 59 86 102 89 93 95
Soybean 12,248 13,012 12,939 12,913 13,465 14,267 15,262 15,652 16,059 16,318 16,833 17,108 16,522
Sunflower 396 188 129 171 168 207 186 320 385 357 370 268 385
Tallow, edible 1,197 1,109 1,239 1,275 1,345 1,218 1,286 1,360 1,581 1,449 1,488 1,590 1,515
a
b
c
August-July year beginning 1982.
d
Disappearance, as defined by the USDA-ERS, means beginning food stocks, production, and imports minus exports, shipments to U.S. territories, and ending stocks.
502 SAFFLOWER OIL
TABLE 2. Typical Fatty Acid Composition of Linoleic

and Oleic Safflower Oils (%).
Fatty Acid Normal Oleic
Palmitic 5.25 4.5

Stearic 1.50 1.5
Oleic 15.00 77.00
Linoleic 77.00 15.00
Others 1.25 2.00
In 1957, scientists in Australia and California independently reported a mutation

that came to be known as oleic safflower (5456). This mutation occurred naturally
and produces a plant and seed that look exactly like linoleic safflower, except for an
oil whose fatty acid distribution is a mirror image of linoleic safflower oil (Table 2).
The initial oleic safflower variety released by Knowles, UC-1 (57), was lower in oil
content and had a poorer yield than conventional varieties available at the time.
This meant that oleic safflower oil was initially sold at a premium. But agronomic
research has since produced varieties that equal or even exceed normal safflower in
yield and that are comparable in oil content.
Oleic safflower oil interested buyers in Japan and the United States when it was
first commercially releasedin Japan as an ingredient for a new mayonnaise and in
the United States as a replacement for peanut oil in most of Frito-Lays western
plants. These markets evaporated, however, when producers were forced to raise
prices because of increasing competition with wheat for western farmland. The
development of markets for oleic safflower oil has been a constant series of steps
forward and then back. The oil has enjoyed good success as an ingredient in arti-
ficial baby milks (because of its excellent stability), in production of premium chips
and snacks (again because of its stability and good frying characteristics), in the
production of cocoa butter substitutes, and as an oil for blending with olive oil
because of its similar fatty acid structure.
In recent years as more research has focused on the role of monounsaturates ver-
sus polyunsaturates and their effects on cholesterol reduction, oleic safflower oil has
begun to receive more attention. In the United States, Saffola Grocery Products has
introduced a grown-without-pesticides salad oil in which linoleic safflower oil has
been replaced by the oleic type. In Japan, several bottlers have begun to feature
oleic safflower oil in their gift-pack campaign both as an individually identified pro-
duct and also in blends with the linoleic type.
The emergence of countries other than the United States as exporters of saf-
flower products became increasingly important. Since 1986, Mexico has been
able to take an increasing percentage of its total supply to world markets, generally
at lower prices than U.S. oils. Earlier, Mexico was limited by government controls
over its exports and poorer varieties of planting seed. These restraints have been
eliminated.
Argentina is able to produce safflower at prices similar to sunflower oil prices

and also has access to good U.S. planting seed varieties. However, Argentina is
able to sell at low prices only by planting safflower in the far north, immediately
after soybeans are harvested. If timing is perfect, a good yield can result. If soybean
harvest is delayed, safflower is also delayed; thus the harvest may be pushed into
the rainy season, causing sprouting and potential loss of the crop.
Recently, Australia has transferred most of its safflower production from
Queensland and New South Wales to Victoria and the south, where rainfall is
more reliable. Australia has not been able to produce a sustained, reliable produc-
tion for world markets. Severe quarantine laws have not allowed more modern vari-
eties of safflower planting seed to be imported; thus most Australian seed is quite
low in oil content.
China has begun to produce some safflower oil. Chinese varieties have tradition-
ally been under 30% oil content, since the bulk of the research there was aimed at
improved seed and floret yields. Safflower produced in the far west and northern
parts of China is high in linoleic levels, and Japan has continually imported small
quantities of Chinese seed because of this. With Chinas emergence into the worlds
economic and political venues, its safflower could become more important as it pro-
duces or imports better varieties of seed and improves its internal transportation
system.
Canada produces limited amounts of safflower, which is aimed primarily at bird
feed rather than oil production. Even more so than in the U.S. northern Great Plains,
Canadas producers face a tough battle trying to squeeze in a long enough growing
season.
The worlds largest acreage of safflower is found in India, and Indian scientists
have undoubtedly published more details about safflower production than
scientists in any other country. India is tying to encourage more safflower
production, but sunflower offers more hope there because of its resistance to
rain. India basically consumes all of the safflower it produces, and this is unlikely
to change.
Spain used to be an important producer of safflower, but sunflower hybrids
have shown much more promise, because of their tolerance of summer rains.
Pseudomonas almost wiped out safflower production in Spain, and current EEC
policy (which offers subsidies to sunflower producers but ignores safflower) does
not leave safflower much chance (58).
The FAOs estimate of world safflower production for the period 19501992 is
shown in Table 3. U.S. safflower acreage fell 13% in 2001 to 188,000 acres. Below
average yields also contributed to a cut in safflowerseed production to 242 million
pounds, making it the smallest crop since 1983. As a result, crush and exports of
safflowerseed in 2001/02 fell to 190 million and 43 million pounds, respectively.
For safflowerseed oil, a recovery in U.S. shipments to Japan boosted 2001/02
exports to 40 million pounds (53). Safflowers future will continue to be limited
by its relatively high cost of production, unless hybrids are developed that can be
produced cheaply.
504 SAFFLOWER OIL
TABLE 3. World Safflower Production by Crop Year (Year of Harvest/Milling) (59, 60).a,b
United World
Argentina Australia Ethiopia India Mexico Spain States Total

Year (ha) (t) (ha) (t) (ha) (t) (ha) (t) (ha) (t) (ha) (t) (ha) (t) (ha) (t)
1950 44 19 370 50 30 16 458 91

1951 45 20 390 53 11 9 460 88
1952 45 21 400 56 23 24 482 107
1953 45 21 406 57 26 25 491 109
1954 45 22 410 57 10 13 479 91
1955 45 22 415 60 23 35 497 122
1956 1 46 23 420 55 36 66 517 150
1957 1 46 23 425 60 37 66 523 155
1958 3 2 48 24 430 60 65 67 560 159
1959 5 2 48 24 433 63 99 113 599 209
1960 2 1 50 25 435 61 26 32 131 152 658 278
1961 4 2 51 26 440 66 33 41 166 160 709 305
1962 2 2 52 26 445 67 37 47 217 334 768 486
1963 8 5 54 27 450 68 36 47 223 358 761 514
1964 19 13 55 27 455 62 36 47 147 255 710 413
1965 24 10 56 28 460 72 59 80 146 262 746 459
1966 38 25 57 29 462 69 165 236 55 30 182 304 955 701
1967 42 16 59 30 478 72 100 149 70 56 181 311 925 643
1968 19 10 60 32 513 78 86 102 55 39 83 169 812 441
1969 11 4 61 34 578 94 145 209 11 5 113 212 939 566
1970 28 9 62 36 580 142 175 288 14 8 102 180 968 673
1971 2 1 34 15 64 39 588 154 265 511 22 13 115 227 1126 891
1972 5 3 11 4 63 24 598 131 199 271 16 8 110 208 1039 676
1973 7 5 12 7 64 24 423 82 198 298 34 20 95 156 874 615
1974 8 8 36 31 64 30 614 191 192 272 34 17 77 150 1054 713
1975 4 3 40 18 64 25 648 212 363 532 34 16 84 175 1269 1001
1976 8 6 13 6 64 30 674 238 185 240 36 20 37 69 1057 632
1977 3 2 39 26 64 30 683 220 404 518 29 13 113 171 1350 989
1978 2 2 75 58 64 30 707 188 429 616 15 15 145 186 1443 1108
1979 2 1 54 30 64 30 703 209 528 635 17 14 152 205 1516 1129
1980 1 1 18 8 65 31 733 279 416 480 20 20 84 111 1340 936
1981 1 1 33 20 65 31 720 335 391 372 12 4 65 101 1289 870
1982 1 1 12 20 66 32 749 421 189 221 20 13 72 117 1114 832
1983 2 2 55 31 66 32 782 396 349 277 19 13 35 92 1320 853
1984 3 2 44 32 66 32 831 501 227 209 20 14 89 124 1290 894
1985 3 2 44 28 66 32 870 497 190 180 19 16 88 110 1288 876
1986 14 10 30 19 67 33 911 348 204 161 15 13 148 147 1385 737
1987 15 11 38 25 67 33 892 353 200 219 12 11 107 155 1331 811
1988 15 11 46 41 68 34 1052 462 200 247 18 13 95 146 1498 960
1989 50 33 33 21 68 34 816 445 150 142 7 6 91 159 1219 850
1990 50 35 19 10 69 34 842 487 157 159 1 1 108 139 1248 873
1991 50 35 37 24 69 35 821 327 94 88 84 115 1160 620
1992 50 35 34 32 69 35 831 350 80 82 136 148 1199 690
a
In units of 1000.
b
Some data from PVO, Agricom International, and Oilseeds International, Ltd.
PHYSICAL AND CHEMICAL PROPERTIES 505
2. PHYSICAL AND CHEMICAL PROPERTIES
2.1. Safflower Seed

Safflower seed (technically an achene) (61) consists of a tough fibrous hull that pro-
tects a kernel comprised of two cotyledons and an embryo. Applewhite (62)
reported that hulls make up 1859% of the seed weight (62), Weiss (63) character-
ized normal hulled seeds as 3849%, and Li et al. (64) noted percentages of
2587.5%. This diversity also shows up in seed weight per 1000 seeds (14105 g),
oil content (11.4847.45%), and fatty acid distribution (linoleic acid, 11.13
85.6%; oleic acid, 6.7481.84%, stearic acid, 0.014.88%, and palmitic acid,
2.129.03%) (57).
Safflower seeds are normally cream to white, but since 1960, breeding has
resulted in great variation in color, ranging from normal hull to thin hull (which
tend to show part of the underlying colored layers) to types with gray, purple, or
brown-striped hulls. Most of this research has been aimed at creating a thinner
hull to increase oil content (Table 4). Although reduction of the hull fraction
TABLE 4. Analyses of U.S. Safflower Seed (65).a
Type Oil Protein Fiber
Analyses of whole seed

Gila 38.1 16.7 22.3
U-5 38.5 17.2 21.1
US-10 36.8 19.4 22.3
Frio 40.1 15.4 20.8
Thick-hull hybrid 37.8 17.3 21.5
Brown striped 47.7 20.3 11.7
Pigmentless brown striped 42.8 22.5 13.6
Thin hull 47.2 21.1 11.2
Analyses of hull
Gila 3.2 4.3 57.1
U-5 2.2 5.0 58.4
US-10 1.4 3.6 60.0
Frio 2.7 4.1 60.4
Brown striped 5.7 8.4 46.9
Thin hull 5.1 10.0 45.3
Analyses of kernel
Gila 60.9 24.9 1.5
U-5 61.8 25.4 1.5
US-10 59.0 29.4 1.5
Frio 64.0 23.0 1.0
Brown-striped 52.7 24.8 0.9
Thin-hull 62.6 25.5 0.9
a
All analyses are percentages on a moisture-free basis.
506 SAFFLOWER OIL
increases valuable oil and protein percentages, too much reduction can produce
other problems. Brown-striped seeds have a distinctly musty odor (66). Thin hull
types can create harvesting, storage, handling, and processing problems (67, 68),
although the more gentle combine harvesters in use for the last decade can probably
deal with the harvesting worry. Brown-striped seeds also were shown to contain
colorless precursors in the hull and kernel that could create dark extracted oil
(68). Ways to remove the precursors and color bodies have been published (68, 69).
Three phosphatides have been identified that are responsible for color formation
from oil extracted from the kernel of safflower seed: phosphatidyl ethanolamine,
phosphatidyl myoinositol, and phosphatidyl choline (7072).
Attempts have been made to produce commercial hybrids of safflower seed by
exploiting heterosis to increase seed or oil content yields (7376). In addition, many
Indian scientists have published on the hybrid theory of safflower. POI marketed a
near hybrid for a short while, which exhibited oil contents in excess of 50%, but it
was not popular with growers and proved difficult to manage in the oil mill. Cargill
marketed hybrids in India for several years. They were initially popular with
growers because of 25% higher yields than standard varieties, but high production
costs and 25% lower oil content caused the program to be phased out. Other hybrid
work (76) was based on white-flowered genetic male-sterile germplasm (7779),
which in turn resulted from colchicine treatment of an introduction from
Afghanistan.
Oleic types of safflower are produced primarily in the United States and to a
minor degree in Mexico. The commonly available types exhibit oleic fatty acid levels
in the 7681% range. Linoleic level decreases proportionally as oleic level increases.
Safflower seeds in the Northern Hemisphere tend to be higher in linoleic acid at
TABLE 5. Characteristics of Commercial Safflower Production.
Protein Linoleic
Country or Oil Moisture (% in Solvent (% in Total
Region Content (%) Extracted Meal) Fatty Acids)
United States
California 39.544 45 25 7578
Arizona 3941.5 45 25 7278
Northern Great Plainsa 2541 59 24 7681
Utah and Idaho 3842 57 25 7678
Canada 3235 59 24 7681
Mexico
San Jose and Quiriego 3038 512 24 6070
Normal types 3537 512 24 7277
U.S. types 3539 612 2324 7277
Argentina 3536 612 2324 7076
India 32 78 2124 7278
China 2832 78 2528 7682
Australia 3538 59 24 7076
a
Wide range caused by loss of oil content in years of early frost. High, basis-S-541 variety; normal range for
local varieties is 3538%.
TABLE 6. Mineral and Trace Element Composition of Some Indian Oilseeds (80).a
Factor Sesame Mustard Groundnut Safflower
Number of varieties 6 8 19 6
Ash (g %) 6.6 0.02 4.7 0.21 3.3 0.09 2.1 0.15
Phosphorus (mg/100 g) 872 35 767 41 500 8 367 10
Calcium (mg/100 g) 1232 28 318 25 77 6 214 28
Magnesium (mg/100 g) 521 42 273 18 239 4 241 18
Iron (mg/100 g) 9.3 0.43 7.9 0.34 2.5 0.23 4.6 0.13
Zinc (mg/g) 122 21 48 3.7 30 1.6 52 3.3
Manganese (mg/g) 13.2 0.26 25.6 4.09 11.0 0.76 11.0 0.78
Copper (mg/g) 22.9 1.70 8.3 0.44 9.0 0.53 15.8 1.54
Molybdenum (mg/g) 2.02 0.068 0.89 0.105 1.66 0.173 0.54 0.08
Chromium (m/g) 0.87 0.052 0.63 0.058 0.48 6.031 0.45 0.060
a
All values are mean SEM for dry weight of the sample.
more northern latitudes because cooler temperatures are usually experienced at and
after the time of flowering. Similarly, oleic levels increase in safflower exposed to
higher temperatures. Oil contents of commercially available oleic types tend to be
slightly lower than linoleic types, but field yields are equal or superior to linoleic
types. Tables 5 to 8 list some of the characteristics of safflower seeds.
2.2. Safflower Oil

Normal safflower oil is pale yellow to golden and has a slightly nutty flavor. Table 9
summarizes the important chemical and physical characteristics of U.S. safflower
oil. Safflower oil exhibits the highest level of linoleic fatty acid of any commer-
cially available oil. This high level, combined with an absence of linolenic
fatty acid, is what has made safflower oil attractive to consumers, initially as a
TABLE 7. Identification and Distribution of Sugars in Safflower Hull and Kernel (81).
Percent Sugars
Percent on Defatted
Safflower Sugars Distribution Percent Basis
Component Present of Sugars Sugars (calculated)
Kernel Uronic sugar glycosides 14.3 0.43

Raffinose 35.8 1.08
Sucrose 46.9 1.42
Galactinol 3.0 0.09
Total 3.02 7.74
Hull Uronic sugar glycosides 38.9 0.37
Raffinose 6.8 0.06
Sucrose 22.6 0.21
Galactinol 2.7 0.025
D-Gluose 14.6 0.14
D-Fructose 14.2 0.13
Total 0.94 0.79
TABLE 8. Amino Acids of Safflower Seed (65).a
Protein Percentage
Protein Factor
Phenylalanine
Glutamic Acid
Aspartic Acid
n 6:25
Tryptophan
Methionine
Threonine
Isoleucine
Ammonia
Histidene
Tyrosine
Arginine
Leucine
Glycine
Cystine
Alanine
Proline
Lysine
Serine
Valine
Defatted, hand scparated
kernels
Normal hull seed
Commercial varieties
Gila 66.4 2.72 3.34 2.42 9.52 9.33 2.96 4.24 20.55 3.80 5.46 4.03 1.86 5.44 1.59 3.84 5.97 2.91 4.20 1.15 5.46
U-5 66.4 2.83 2.43 2.39 9.74 9.40 3.17 4.38 19.69 3.56 5.38 4.06 1.80 5.43 1.75 3.83 6.13 3.03 4.21 0.93 5.44
US-10 71.7 2.63 2.28 2.42 9.69 9.47 2.84 4.16 19.97 3.87 5.09 3.90 1.62 5.24 1.28 3.53 5.87 2.86 4.21 0.90 5.43
Frio 65.6 2.83 2.44 2.40 9.32 9.24 3.04 4.19 19.49 3.78 5.43 4.03 5.49 3.86 5.98 2.91 4.15
Experimental varieties
Normal hull histearic 74.4 2.53 2.50 2.59 9.76 9.23 2.57 3.72 20.14 3.52 4.92 3.67 5.45 3.76 5.85 2.78 3.96
Normal hull hi-oleic 69.0 2.83 2.42 2.42 9.66 9.17 2.86 4.01 19.82 3.58 5.11 3.81 5.25 3.70 5.79 2.74 4.06
Normal hull equal 71.9 2.61 2.42 2.48 10.09 9.44 2.92 4.37 21.24 3.92 5.19 3.95 5.36 3.76 6.12 3.03 4.16
oleic-linoleic
Other normal hull 58.3 2.91 2.60 2.59 10.32 10.29 3.24 4.74 22.39 3.83 5.48 4.51 2.01 5.96 1.69 4.15 6.57 3.16 4.53 1.08 5.17
mutants
Seeds with low hull content
Pigmenticas, striped hull 62.2 2.68 2.51 2.61 10.19 10.25 3.06 4.72 20.45 3.74 5.65 4.39 2.04 5.81 1.61 4.02 6.51 3.09 4.50 1.15 5.42
Brown-striped hull 65.9 2.73 2.41 2.46 9.67 9.69 2.96 4.34 20.50 3.88 5.47 4.10 1.81 5.47 1.61 3.89 6.05 3.00 4.90 1.08 5.45
Thin hull 67.8 2.75 2.39 2.45 9.58 9.40 3.01 4.27 20.12 3.84 5.36 4.06 5.46 3.88 6.07 3.04 4.29
Hulls
Gilla 4.0 2.86 1.22 2.50 2.87 6.38 3.11 4.36 7.82 3.37 4.53 3.29 4.34 3.14 4.62 1.16 3.25
Brown-striped hull 8.1 3.21 1.58 2.15 3.43 7.56 3.39 4.75 8.80 3.78 5.01 3.79 1.65 4.79 1.04 3.47 5.09 1.46 3.70 0.43 5.48
Thin hull 10.2 3.07 1.33 2.30 3.09 7.30 3.03 4.48 7.78 3.26 4.46 3.32 4.31 3.15 4.58 1.52 3.47
Safflower meal
Commercial partially 48.0 2.84 2.32 2.42 8.69 9.13 3.10 4.36 19.34 3.93 5.46 4.17 1.70 5.46 1.62 3.97 6.13 2.48 4.32 5.45
decor-ticated, normal
Experimental undercorti- 38.1 2.63 2.23 2.56 8.34 9.22 2.93 4.17 18.56 3.68 5.27 3.97 1.63 5.32 1.38 3.80 5.86 2.39 4.29 5.41
cated, brown-striped hull
a
In g/16g nitrogen.
TABLE 9. Physical and Chemical Characteristics of U.S. Safflower Oil (82).
Usual Range
Characteristic of California Oil Minimuma Maximuma
Physical
Color (Gardner) 810 11b
Color after heat bleaching, 315.5 C (600 F) 23b 4b
c d
Color, refined, bleached, deodorized 0.51.0 red 15 yellow and 1.5 redd
Specific gravity, 25/25 C 0.9190.924
Refractive index, np 25 C 1.4731.476
Titer, C 1517
Flash point, C ( F) 148.8(300) 121.1(250)
Chemical
Free fatty acids, % as oleic 0.150.6 2
0.030.05d 0.05d
Iodine value (Wijs) 141147 140 155
Unsaponifiable, % 0.30.6 1.5
Peroxide value (at time of shipment) 01.0d 1.0d
Moisture and volatile, %e 0.030.1 0.8
Insoluble impurities, %f 0.010.1 0.3
Moisture and impurities, % 0.050.1b 0.1b
Principal fatty acids, % TFA
Palmitic 46
Stearic 12
Oleic 1612
Linoleic 7579 72
Linolenic Nil
a
Per NIOP trading rules.
b
Nonbreak grade, NIOP.
c
AOCS method Cc 13b-45.
d
Edible grade, NIOP.
e
AOCS method, Ca 2d-25.
f
AOCS method Ca 3-46.
quick-drying oil that could produce films that would not yellow with age and, more
recently, as an edible oil with the highest available level of polyunsaturation. As an
edible oil, the high level of unsaturation also creates problems. Home consumers
using safflower oil for frying must be careful to clean pans quickly after use or a
tough varnish film results, which is difficult to remove. Fresh safflower salad-grade
oil has excellent flavor and odor characteristics, and because it lacks linolenic fatty
acid, it does not display the fishy or beany odors sometimes associated with poorly
refined soybean oil. Unfortunately, it does have a relatively short shelf life (typi-
cally 912 h AOM), which means the oil should be kept cool after the bottle is
opened to maintain freshness.
Oleic safflower oil displays most of the same characteristics as the linoleic type,
except for its fatty acid structure (see Table 2). It has been noted that a blend of
linoleic and oleic edible oils would improve the dietary value of commercial saf-
flower oil (83). Blends of this type began to be marketed in Japan in 1990 and
appear to be achieving good acceptance by the public.
510 SAFFLOWER OIL
A great variation in fatty acid, oil, and protein levels occurs in the world collec-
tion of safflower seeds. Knowless pioneering work in understanding and subse-
quently finding ways to modify these differences inspired many researchers to
publish extensively on this subject (34, 84). Recently, most research on safflower
oil modification has been performed in the United States by private planting seed
companies and by the Sidney Experiment Station of Montana State University; lit-
tle has been published.
A few studies have reported on the location of the fatty acids on the triglyceride
in varying ways. One study used argentation TLC with lipase hydrolysis on a sam-
ple of safflower oil from Kenya that contained 10% total saturated fatty acids. It
was found that 2 mol% was configured with two saturated acyl chains (S) and
one unsaturated acyl chain (U), 26 mol % had a SU2 configuration, and 72 mol
% had a U3 configuration. It was also reported that 3 mol % had two double bonds
attached, 3 mol % had three double bonds, 23 mol % had four double bonds, 19 mol
% had five double bonds, and 47 mol % had six or seven double bonds (85).
Another study measured the position of linoleic acid on the triglyceride in a study
on hydrogenation (86). It was found that 84.6 mol % of linoleic acid was located in
the 2-position in a safflower oil containing 76.4% linoleic acid.
2.3. Safflower Meal

The by-product of the extraction of safflower oil is a grayish tan to brown cake
or meal that exhibits flakes or shreds of whitish safflower hulls. Table 10 presents
typical analysis for safflower meal. Most meal produced in the United States is of a
solvent-extracted type. The amino acid and mineral contents of meal are shown in
Table 11.
Australian data from 1959 indicated up to 17,324 kg/ha of green matter
(2762 kg/ha of dry matter) could be gained by grazing safflower as a green crop
TABLE 10. Typical Analyses for U.S. Safflower Meal (87).
Characteristic Aa Bb Cc Dd
Crude protein, % 21.03 20.00 42.0 25.4

Crude fat, % 6.6 0.5 1.3 1.5
Moisture, % 9.0 10.0 9.2 8.0
Crude fiber, % 32.2 37.0 15.1 32.5
Ash 3.7 5.0 7.8 5.9
Calcium, % 0.23 0.24 0.4 0.37
Total phosphorus, % 0.61 0.24 0.4 0.8
NFE, % 40.0
TDN, % 57.0
a
Expeller pressing of safflower seed without decortication.
b
The low end fraction of meal that resulted from prepresssolvent extraction of safflower seed followed by two
fraction tail-end decortication.
c
The high end fraction of meal that resulted from prepresssolvent extraction of safflower seed followed by
two-fraction tail-end decortication.
d
Prepresssolvent extraction of safflower seed without decortication. Typical California, 1992.
PROCESSING 511
TABLE 11. Amino Acids and Minerals in Safflower Meal (87, 88).a
Factor A B C
Methionine 0.4 0.33 0.69

Cystine 0.5 0.35 0.7
Lysine 0.7 0.7 1.3
Tryptophane 0.3 0.26 0.6
Threonine 0.47 0.5 1.35
Isoleucine 0.28 0.27 1.7
Histidine 0.48 0.5 1.0
Valine 1.0 1.0 2.3
Leucine 1.1 1.2 2.5
Arginine 1.2 1.9 3.7
Phenylalinine 1.0 1.0 1.85
Glycine 1.1 1.1 2.4
Calcium 0.28 0.37 0.44
Phosphorus 0.78 0.80 1.41
Potassium 0.79 0.79 1.33
Magnesium 0.36 0.37 1.33
a
See Table 10 for explanation of A, B, and C. Numbers are percents.
with 1112% protein (22). Indian researchers have presented several papers that
showed promise concerning the production of fodder and ratoon seed in northern
India (8292). In 1993, a U.S. farmer was able to harvest approximately
8000 kg/ha of green hay, which measured 18% protein from a safflower crop that
failed to mature.
3. PROCESSING
3.1. Extraction
Much of the safflower processed in India in the past was crushed by a mortar-and-
pestle-like device called a ghani. Seed was cleaned by hand and then introduced
into a chakki. This machine, which consisted of two horizontal stone wheels, one
of which was turned by a blindfolded bullock, partially dehulled the cleaned seed
passing between the stones. Hand winnowing and sieving next removed the hulls
from the seed kernels. The meats were pressed into balls after the addition of about
6% water. About 15 kg of the balled kernels were introduced into the ghani, an
inverted conical mortar into which a heavy pole was placed. The pole was held to
the side of the mortar by heavy weights and dragged around the perimeter by a team
of oxen. A small amount of heated oil was added, and crushing then proceeded for
45 min, after which the oil was allowed to drain out through a small hole. A ghani
could process about 100120 kg of seed per day.
More recently, ghanis capable of processing 150175 kg per day were some-
times motor driven. Animal-powered ghanis could obtain 1116% residual oil in
the extracted meats, while motor driven models could extract 1012%. Today,
512 SAFFLOWER OIL
some cast-iron ghanis, expellers, and solvent-extraction plants are used, in addition
to the older stone devices (9395). The oil extracted by a ghani is clarified by set-
tling and decanting or by water washing. The oil is placed in tins for local sale.
Most safflower was first processed in the United States by continuous screw
press expellers. Some processors attempted decortication, but the nature of saf-
flower seed acts against successful decortication. To prevent the oil from scorching,
water-cooled shafts were recommended. Oil so treated could easily be heat
bleached to a level below 4 Gardner color. Expellers such as the Anderson Super
Duo Duplex could process about 15 t of safflower seed per day, leaving 78% resi-
dual oil in the remaining cake. However, the principal problem encountered in pro-
cessing safflower seed through expellers was the propensity of expeller-processed
safflower meal to burn in storage (96). The combination of a reactive polyunsatu-
rated residual and the fibrous texture of safflower meal created many fires in the
1950s and early 1960s. Once safflower processing shifted to prepresssolvent
extraction, which brought residual oil contents down below 1.5%, most storage pro-
blems were eliminated.
These same expellers, if employed in a prepressing mode wherein 1517% resi-
dual oil remains in the cake that is sent to the solvent extraction unit, can process
4550 t of seed per day. Prepressing of safflower produced under California
conditions (or the equivalent) requires no cooking, flaking, or cracking of the
seed before extraction and results in oil capable of being heat bleached to 13
Gardner color.
In the early 1960s, PVO produced an air gun device that decorticated safflower
seed satisfactorily, but the idea was abandoned because it required too much energy
and was extremely noisy (34). A PVO researcher developed a method for decorti-
cating safflower meal after extraction, which employed the principle that the fine
particles produced in grinding safflower cake are high in protein and the coarser
particles are more fibrous (97). This method, using vertical hammer mills to grind
the cake and a combination of air classification and screening was employed by
several California mills in the 1960s and 1970s to produce safflower meal of
42% protein, in addition to an 1820% protein middle fraction and a 6% protein
hull fraction. More recently, most mills have returned to only producing as is
meal of approximately 25% protein content, because the high amount of energy
consumed by the tail-end process cost more than the additional return gained
from the high protein fraction.
The high cost of energy encouraged some mills in the 1980s and 1990s to
replace or supplement prepress expellers with caged expander-extruders, which
are capable of removing approximately 66% of the available oil through the caged
portion of the extruder and to produce collets that are ideal for efficient solvent frac-
tion. Extruders require much less horsepower per ton of seed processed than expel-
lers and cost less to maintain (98100).
Horizontal basket or moving bed solvent extractors are preferred over vertical
tower extractors in processing safflower cake. The fibrous nature of safflower pro-
vides a natural channel through which the solvent can move, and the bed acts as
a natural filter medium. Tower extractors generally have problems extracting
PROCESSING 513
safflower seed because the hulls tend to float, sometimes carry over in the top of the
extractor, and cause excessive wear in a towers rotary seal.
3.2. Refining, Bleaching, and Deodorizing

Safflower oil that is extracted from seed in good condition is easy to refine because
it is low in FFA and contains few gums or impurities. Conventional caustic refining
systems work well. This most important factor in handling safflower oil, destined
for edible use, is to limit exposure to air throughout the extraction, refining, bleach-
ing, deodorizing, and packaging cycle. Nitrogen blanketing should be employed if
deodorized oil is to be stored for more than a few hours. Generally speaking, saf-
flower oil processed by expeller processing will contain just enough free fatty acids
and impurities to require refining before deodorization; in most cases, safflower oil
prepressed from California, Arizona, or northern Mexico seed can be introduced
directly to deodorizers. California prepress oil normally will meet a varnish makers
nonbreak grade without further processing.
Safflower seed that is produced in areas with late summer rains or cool weather
cycles that interfere with maturation can produce dark-colored or greenish oils that
are often higher in FFA as well. If the seed has sprouted before or during harvest or
has been attacked by Alternaria, Pseudomonas, or other head-rot diseases, the
resulting oil can be quite difficult to refine and extremely difficult to bleach.
While safflower oil may, on occasion, display minute traces of a fine, lacy wax
(101), most U.S. refiners neither winterize nor dewax safflower oil, feeling that a
brilliant oil can be delivered without it. Japanese refiners generally insist on refining
and bleaching safflower oil to under 1.0 red color, followed by winterization to
avoid problems with minute amounts of wax that may appear in the oil in the winter
months in the north.
3.3. Production of Margarine and Mayonnaise

If proper steps are not taken, physical crystal changes (polymorphism) can take
place in the production of safflower margarines, resulting in a product with a sandy
texture (102). The b-crystalline form that results consists of large crystals instead of
the smooth, uniform mixture desired in a margarine; safflowers uniform triglycer-
ide structure encourages production of b-crystals. This problem can be solved by
incorporation of a small amount of more saturated oil into the margarine mix. PVO
solved the problem in its Saffola margarine by adding 5% cottonseed oil, which
also improved the products mouth feel (34). A 1966 patent described a blending
of liquid safflower with selectively hydrogenated safflower and peanut oils (103).
Soft safflower margarines, wherein a highly hydrogenated safflower lattice was
employed to encapsulate a larger portion of liquid safflower oil, have been success-
fully produced by several companies (34). The methods employed to produce these
types of margarine structures have been reviewed (104, 105). It has been shown that
Cr(CO)3 catalysts can be used to selectively hydrogenate safflower oil and retain a
9095% cis configuration (106108). Several studies have reported on safflower oils
514 SAFFLOWER OIL
good taste, appearance, odor, and texture in mayonnaise and frozen salad dressings,
where it exhibits excellent qualities in repeated freezethaw cycles (34, 102,
109, 110).
3.4. Industrial Processing

Although this Chapter is concerned with oils that are used in edible products, it is
well to remember that safflower oils recognition in modern times occurred because
of interest in its excellent properties as a semidrying oil. Safflower oils light color,
ability to heat bleach to near water whiteness, low level of free fatty acids and
impurities, and lack of linolenic acid make it an ideal vehicle for white house paints
and varnishes and for the production of alkyd resins. It is easy to polymerize via
kettle bodying without the need for vacuum equipment; capable of producing excel-
lent blown, limed, or maleated oils; and acts as a good source for conjugation or
methyl esters (34).
4. ECONOMICS AND MARKETING
As mentioned, safflower is a crop that has been grown for thousands of years, pri-
marily for local use. As people traveled they carried safflower seeds with them, gen-
erally for personal use. It is only in recent times that safflower has entered world
commerce; still much of what is produced remains in the country where it is grown.
The price of wheat has been the dominant factor affecting the price that U.S.
farmers must receive for safflower seed to put safflower into their cropping plans.
In its early years of U.S. production, safflower oil competed directly with soybean
oil for market share and soybean futures on the Chicago Board of Trade, offered as
a reasonable medium for hedging safflower seed and oil prices. But, more recently,
safflower prices have borne little relationship to the market for soybean oil, and saf-
flower oil has become a product that is impossible to hedge.
In 1997, U.S. farm wholesale prices were the following: safflower oil, tanks,
$0.59 lb; soybean oil, tanks, $0.24/lb. In 2002, prices for were safflower oil, tanks,
$0.79/lb, soybean oil, tanks, $0.19/lb (53).
4.1. Safflower Seed

In the United States, most safflower seed is grown by farmers who have agreed to a
contract of sale with an oil mill or grain dealer before planting the crop. Because
there is no daily market for safflower seed posted in the newspaper and there are no
quotations available from the commodity futures markets, most banks or other
financing agencies encourage farmers to contract their crop in advance. There is
no other way for the bank to protect any funds that have been loaned with the
crop as collateral.
The usual safflower production contracts state that the farmer will deliver the
entire yield from a given number of acres or hectares. The buyer assumes the
ECONOMICS AND MARKETING 515
risk of yield. Besides stating the number of hectares to be planted and their location,
contracts usually specify the type of planting seed, the name of a landlord (if any),
what compensation he or she is to receive, and of course, the price and point of
delivery.
The National Institute of Oilseed Products (NIOP) publishes an annual rule book
that covers specifications and standards of trade for many vegetable oils, including
safflower and oleic safflower. Rules 7.1 g and h (formerly 110 g and h) and 7.1 i
(formerly 110 i) are the NIOP rules for safflower seed and oleic seed, respectively.
When combined with the state of Californias official standards for safflower seed,
little room exists for argument as to the meaning of a contract between buyer and
seller.
Safflower seed is usually sold domestically on a dockage-free basis with no limit
on the amount of dockage a shipment of seed might contain. Dockage is defined as
any foreign material plus parts of the safflower plant other than seed, empty or
partly filled seeds and broken parts of the seed small enough to pass through a
screen opening of 1.78 mm. Moisture content is required to be <8%, unless the
buyer is willing to accept a higher percentage in exchange for a penalty. Most
oil mills will accept limited quantities of seed up to 12% moisture content, if their
schedule permits such seed to be processed immediately. Moisture content levels
higher than 12% cause problems in expeller operation. Some mills also operate
grain dryers, which allow them to accept higher moisture content seed.
Normally, buyers require safflower seed that contains more than 5% dockage or
green foreign matter or that is higher than 8% moisture to be cleaned before accept-
ing it. This is to prevent heating in storage. Because most buyers in California pur-
chase safflower seed from the farmer free on board a truck at the edge of the field,
making the buyer responsible for the cost of freight to the elevator or oil mill site,
the farmer is normally also charged for the cost of freight involved on dockage in
excess of 5%. In other states, in Mexico, and most other parts of the world, the
farmer is responsible for the delivery of the seed to the buyers location.
In most parts of the world, except India, safflower seed is handled in bulk. In
California this is accomplished in large aluminum-sided, bottom-dumping, open-
top truck trailers of approximately 1012 t capacity each, two of which are hauled
in tandem to a field by a truck tractor unit. The trailers are left by the field to be
filled by the farmer, and the tractor unit returns and hauls the full trailers directly to
the oil mill or export terminal (in some cases up to 250 km away) or to a closer
grain elevator for intermediate storage. In other parts of the United States, safflower
is delivered in many types of grain trucking equipment and much of it is delivered
to small country elevators where it is stored, cleaned if necessary, and subsequently
loaded onto trucks or railroad hopper cars (which can hold between 50 and 70 t of
safflower seed) for delivery to a buyer.
Correct methods for sampling of safflower seed are specified in the NIOP rules.
Three probes in each truck trailer with a grain probe is the preferred method. Sam-
pling takes place on delivery of the seed to the first place of rest and is conducted if
possible by a third party (a representative of the California State Department of
Agriculture or an employee of the receiving elevator company) to minimize
516 SAFFLOWER OIL
disputes. Moisture content and refractive index (if oleic safflower is delivered) of
the parcel is checked immediately and a carefully split portion of the sample is then
forwarded to the nearest state of California or other independent laboratory for
determination of dockage or, should the sampled seed be defective, other factors.
In other states besides California and Arizona, the purchase contract specifies a
price that is based on a dockage-free sample, a certain level of oil content, and the
moisture content at the time of delivery. This is necessary because of the effect that
cold temperatures, rain, snow, disease, or drought can have on the oil content of an
individual crop. In California and Arizona, little variation in oil content occurs from
year to year in a particular variety of safflower. In the northern Great Plains states,
the oil content of a seed that might be 3840% in a normal year can be as low as
2025% under adverse conditions. In the mountain states, these variations are
usually less extreme. Safflower seed in the Great Plains is normally purchased on
a 38% oil content basis (sometimes 40% is used as a basis) with reciprocal allow-
ance of 2% for each 1% variation in oil content (fractions in proportion) applied to
the agreed on price.
Because the United States does not use the metric system, prices in California
and Arizona are normally quoted in dollars per short ton of (2000 lb, 907.185 kg)
and in the Great Plains and mountain states most transactions are fixed in cents per
pound (453.59 g). In the rest of the world, metric tons or quintals prevail, except in
some parts of India and China. Prices paid to farmers in various parts of the United
States vary because of quality and distance from final markets.
The price offered for safflower seed in California is shaped by several market
elements. The principal factor is the amount of land in the central valley that
will be committed to rice, cotton, and tomatoes, the three primary income-produ-
cing crops in the area. Safflower, sugar beets, grain, and corn compete for the
remaining cultivated land; the competition between wheat and safflower is the
most intense. Experienced farmers favor safflower over wheat if the contracting
price for safflower multiplied by an average yield of 2.5 tons/ha equals or exceeds
the perceived price for wheat multiplied by a yield of 5.06.0 tons/ha. Safflower
buyers usually begin negotiating with farmers in October, since farmers must
make the decision to withhold planting wheat at that time. Wheat is normally
planted in late November through January, and safflower is planted during February
through April. During the 1980s the acreage of safflower planted in California
would decline sharply when prices fell below $275/t. In the 1990s, this value
was about $330/t, because of inflation and the increasing prices for other crops.
The second factor that affects the price of safflower is the condition of the mar-
ket for safflower oil. For example, there may be a surplus of oil from the previous
crop, Mexico may be forecasting a large harvest (which occurs 34 months before
the U.S. harvest), or Japanese buyers may be experiencing a slowdown in their
domestic market.
Because safflower oil is a specialty that serves a market that responds little, if at
all, to price changes, these two factors tend to slow down dealers desire to buy the
seed, and a rationing process takes place. Dealers either delay their opening gambits
to contract for safflower seed from the next crop or offer low prices that do not
compete with other alternatives. Side issues that affect supply (drought, disease, or
floods; longshore or transportation strikes; etc.) and price (changes in government
support for competing crops or in import or export regulations, etc.) also affect
these decisions. Safflower prices are not affected by prices for other commodity
types of oils such as sunflower, soya, and canola, except in periods of wild upward
price movements. If prices for other oils climb above $880/t safflower oil prices
move up accordingly or the safflower oil disappears into export markets as a
replacement.
Of course, the demand for safflower can be changed by longer-term fundamental
changes. In Japan, safflower oil is identified as an eminently healthy oil that is given
as a gift. Should medical research find that polyunsaturates, and particularly saf-
flower oil, cause medical problems that outweigh its benefits, demand for the oil
would crumble. On the other hand, if long-term medical studies show that mono-
unsaturates, including oleic safflower oil, are preferred over the types of oils, even
over linoleic safflower oil, there might be a shift in the ratio of linoleic to oleic
safflower oil consumption. This appears to be happening in Japan.
Safflower seeds produced in California are located close to the ultimate domestic
safflower oil markets as well as near export terminals for ocean shipping to Japan or
Europe. Safflower seeds produced in the Great Plains, however, are generally priced
$50/t below California prices for several reasons. First, Great Plains safflower seed
is generally 35% lower in oil content than western seed, and in years of bad weath-
er it can be much lower. Second, while the oil produced therefrom is generally
24% higher in linoleic fatty acid than California seed, it is normally 0.25% higher
in FFA and 12 Gardner color units higher, with generally a greenish tinge, all of
which necessitates higher refining and bleaching costs. Great Plains prepress oil
normally cannot be deodorized without prior refining. Finally, Great Plains seed
must face a long railroad trip to markets in California, or if delivered locally for
processing, the oil and meal produced from it face long trips to consumer markets.
Safflower growers in the mountain states face similar discounting problems. Moun-
tain-grown seed usually is closer to California seed in quality but has no local
milling or customer base so all seed must be delivered over a long distance.
These factors do not apply to the markets for safflower seed sold for bird feed.
Birdseed buyers specifications emphasize seed color (pure white seed is preferred)
test weight (a weight in excess of 0.4739 kg/L is desired), and purity (less than 1
2% foreign material is preferred). Oil content is not a factor. Seeds that have heavy
white hulls and, accordingly, low oil content are preferred for birdseed use. Conse-
quently, birdseed buyers, whose customers are located predominantly in the eastern
half of the United States or overseas, prefer to contract in the Great Plains and
mountain states, where they compete with $50/ton lower seed prices and enjoy a
$4050/ton freight advantage to eastern markets.
It is hard to judge the exact size of the market for birdseed safflower, but as feed-
ing of wild birds increases in the United States, most dealers believe it has exceeded
20,000 t annually. China generally enjoys the reputation of supplying the best bird-
seed quality, since much of Chinese seed is below 30% in oil content and normally
has white hulls. Weather and transportation factors sometimes increase difficulties
518 SAFFLOWER OIL
for marketers of Chinese-origin seed. Indian seed was a factor in world birdseed
markets until 1989, when the Indian government banned the exportation of seed
to improve local supplies of oil. This undoubtedly contributed to the increase in
the U.S. birdseed market.
Exportation of safflower seed to Japan was the largest factor in the expansion of
U.S. safflower seed production. When high duties on the oilseeds entering Japan
began to be relaxed, U.S. safflower seed exports declined. Once more than 10 oil
mills were engaged in processing safflower seed in Japan, now only 2 mills con-
tinue to crush safflower seed there. The remainder of Japans needs for safflower
oil are covered by imports of safflower oil. Safflower seed exports are governed
by the NIOP Rule 7.1 g (former Rule 110 g Export). Many factors are involved
in the Domestic Rule (Rule 7.1 h), but export terms require measurement of oil con-
tent and payment of a premium or discount much as most seed is purchased in the
Great Plains states. The export safflower seed rule establishes a price basis point of
34% oil content with a premium/discount of 2% for each 1% of oil content variation
with fractions in proportion. The 34% level is used, even though most seed exported
today is in the range of 4143% oil content, because this was the level of oil content
available when safflower first started being exported. Export rules also allow only a
maximum of 3% dockage for an export shipment to be considered correct, although
provisions are made for allowing shipments containing up to 6% maximum in
exchange for a penalty of an additional 0.2% for shipments measuring between 3
and 6% dockage.
Because one oceangoing vessel normally carries 3,0005,000 t, and up to
15,000 t, the sampling methods are different from those used for truckloads. The
NIOP rules call for oil, moisture, and dockage analyses to be performed separately
on samples representing each 1,000 t, or fraction thereof, loaded on a vessel. In the
case of oil content analyses, identical samples of each 1,000-t lot are presented to
five different independent laboratories, each laboratory reports its analyses for the
entire load on a weighted average basis, the results of the laboratory with the high-
est and lowest oil contents are discarded and the results of the remaining three are
averaged and used for payment purposes. In this manner a fair analysis is made,
because safflower oil content is difficult to measure accurately.
Almost all safflower seed exported to Japan from the United States has come
from California and in some years, Arizona. This is because the Japanese prefer
the quality of West Coast production, preferring not to pay a high ocean freight
cost on a seed that is lower in oil content; generally produces higher color and refin-
ing costs; and may contain the fungus Sclerontinia sclerotina, which although
found regularly in the Great Plains has not yet been observed in California. Japan
imports some safflower seed from Australia and a small quantity from China, to
obtain the 80% linoleic safflower oil that Chinese seed can guarantee. Japan
also imported safflower seed from Mexico at one time. The United States has
exported safflower seed to Europe, but recently the high price of safflower seed
and low value of safflower meal in Europe have made the processing of safflower
seed impractical, and imports have been confined to safflower or oleic safflower
oils.
4.2. Safflower Oil

In todays marketplace safflower occupies a unique position. It is the oil with the
highest level of linoleic acid available commercially. It continues to enjoy a favor-
able reputation in the mind of consumers, which is a legacy of the polyunsaturated
boom of the 1960s. Most safflower oil produced today reaches consumers as a
refined, deodorized, and bleached salad oil; as a principal ingredient in margarine;
and in several forms of mayonnaise and salad dressings. A small percentage of the
total oil produced, primarily prepress oil, is bottled and sold to consumers without
any further refining, bleaching, or treatment of any kind other than filtration. In the
United States and Europe, a small segment of the market wants an oil that has not
been exposed to chemicals (in this case hexane). Bottled prepress oil generally has
a short shelf life, perhaps of less than 2 weeks, and once opened the oil needs to be
refrigerated so it does not develop strong odors.
Two sellers dominate the U.S. grocery market for all safflower-edible products,
although a number of other companies produce small quantities for health food
venues. Bottled safflower salad oil generally retails at more than $1 per bottle high-
er than canola, corn, sunflower, or soybean oil. Customers for safflower oil make up
a small but dedicated segment of the market. Safflower salad oil brands have never
achieved over a 7% market share, and without heavy advertising, this level drops
in half.
In Japan, the premium price is almost an advantage, and the companies market-
ing safflower oil enjoy better margins for the product than other oils produce. They
exert strict quality controls, market the oil in beautiful and expensive gift packs, and
engage in heavy advertising to maintain market share. In Japans gift-giving sea-
sons, safflower oil has achieved a premier status among all oils. Some say it has
captured up to 85% of this market.
In Mexico, safflower oil occupied a preferred status for many years in grocery
stores catering to the affluent. When first produced in Mexico, a sizable portion of
the safflower oil produced was used as an adulterant in sesame oil. Over time, saf-
flower became the premier oil in the marketplace and puro cartamo would com-
mand a substantial premium. Safflower oil itself soon began to be adulterated
with sunflower and other oils, and eventually consumers became aware of this
and switched loyalty to branded oils that were cheaper. Little safflower oil is found
in Mexico because it is generally exported to the United States or Europe, and lower
priced sunflower, canola, or soybean oil is imported in its place.
A similar situation has taken place in Australia, where the bulk of safflower
grown is no longer processed for the local market but is exported as seed or oil.
A small amount of Australias safflower total is devoted to producing so-called
organic safflower oil. Because Australia still has virgin farmland, it is possible to
produce a crop of safflower using no herbicides, insecticides, or fertilizer. Some
organic safflower oil is also produced in the United States. In the last 3 years the
most successful program has been operated by Saffola Grocery Products Co., which
markets so-called Grown Without Pesticides safflower oil. Saffola has chosen to use
this method, because it wishes to establish its own definition for the purity of its
520 SAFFLOWER OIL
product, in contrast to organic oils, which are usually defined by government edict,
subject to periodic change. The Grown Without Pesticides (GWP) regime requires
that the safflower is planted on land that has had no chemicals applied to it for at
least 6 months and that shows no residue levels. The farmer is allowed to apply
fertilizer but no planting seed fungicides (allowed in organic farming) or other che-
micals. A thorough auditing scheme that includes inspection of the crop throughout
the growth cycle and inspection of harvesters, trucks, and storage facilities for
cleanliness and lack of chemical sprays is employed. This is more rigorous than
the standards employed by the organic industry, which works primarily on the hon-
or system.
The GWP program for oleic safflower oil has been successful when there is hea-
vy advertising. Saffola has not been able to expand beyond the regional market
because of the cost of advertising. The Japanese gift-pack market, which also
uses heavy advertising, includes some oleic safflower oil. One manufacturer is sell-
ing a blend of linoleic and oleic safflower oils to combine the good attributes of
both oils in a single package; its largest competitor markets the oils separately to
give the customer a choice.
In India much oil is still sold by small mills that simply filter oil from the press
and supply the product in small tins or even in the consumers own vessel. Safflower
production is by and large a neighborhood affair in India. While the government is
encouraging more production of all types of oilseeds, sunflower, which has much
wider adaptation than safflower, enjoyed spectacular increase in production in the
1990s.
The European market consists of three areas. First, safflower oil is an ingredient
in sunflower-based margarines, helping to maintain a guaranteed level of polyunsa-
turation. This market area may be in decline, because some manufacturers lowered
their polyunsaturated guarantee levels in early 1994, opting perhaps to feature low
saturation or higher monounsaturation attributes in the future. Safflower oil also
finds a small but dedicated audience because of its high level of unsaturation. A
portion of this market prefers to use either unrefined prepress oil or a form that
has been gently deodorized. The third market area is for safflower, and particularly
oleic safflower, oil that is used for blending with other oils. When safflower oil
became more expensive than other oils, this market area virtually vanished.
Much more rigorous and sophisticated control measures by government authorities
have also restricted attempts at blending.
Like safflower seed most safflower oil also is traded under rules established by
the National Institute of Oilseed Products, in this case Rule 6.11 and 6.12. Rules for
both domestic and export shipments are in force, with the primary difference being
that the export rules require more analyses to be performed before payment. Of
course, some U.S. buyers, many of whom have never heard of the NIOP, establish
their own specifications for the safflower oil they purchase, but by and large their
standards meet or exceed the NIOP grades.
It is outside the scope of this article to examine the medical literature that fueled
the polyunsaturated boom of the 1960s (34) and that has continued to provide
impetus to the U.S., European, and Japanese safflower oil markets. Whether
TABLE 12. Historic U.S. Oleic Safflower Plantings and Production.
Crop Year Plantings (ha) Production (t)
1967 405 953

1968 5,221 11,031
1969 8,580 23,014
1970 5,868 12,973
1971 13,462 32,922
1972 8,843 20,321
1973 15,480 24,222
1974 11,615 22,801
1975 21,004 43,316
1976 9,632 23,678
1977 9,594 24,540
1978 14,569 21,037
1979 19,010 36,940
1980 13,345 31,351
1981 9,340 24,721
1982 4,856 12,701
1983 4,917 12,610
1984 11,550 31,026
1985 10,958 27,994
1986 11,635 31,425
1987 4,290 11,340
1988 8,094 20,684
1989 10,805 29,393
1990 11,343 29,908
1991 10,891 22,803
1992 33,634 48,680
1993 36,430 73,564
monounsaturation or simply lack of saturation will become the wave of the future
is unknown, but from an historic viewpoint it is interesting to observe how oleic
safflower production slowly increased since the crop was introduced in 1967
(Table 12).
In the United States in 2003, 221 103 acres of safflower were planted and
212 103 acres were harvested. Forecasts for 2004 are for 142 103 acres to be
planted and 133 103 acres to be harvested (111).
4.3. Safflower Meal

Safflower meal, the by-product of the production of safflower oil, contains all of the
hull. The high fiber content of the hull limits its value. In California, safflower meal
is employed primarily as an ingredient in dairy feeds; it is also used in beef cattle
feed and to a limited extent in poultry mixes. In the 1960s and 1970s, when saf-
flower meal was being decorticated by the tail end process, the resultant high
(3842%) protein fraction found good employment in chicken and turkey rations.
PVO produced three meal fractions of 42, 20, and 6% protein. Although PVO and
522 SAFFLOWER OIL
others spent considerable time searching for alternative uses for safflower hulls, the
best bet in that period was to export the 6% fraction to Japan, where safflower hulls
were used as low cost filler in many types of compound feeds. Japan was also a
regular consumer of 20% protein (sometimes purchased basis 20% proteinfat com-
bined analysis), but in todays market, safflower meal from the United States is not
competitive in the Japanese market.
Safflower hulls find their best market when incorporated in safflower meal, and
none has been produced separately for many years in the United States, because
most mills produced only two fractions when decorticating, 20 and 42% portions.
Today, the energy consumed in separating safflower meal fractions exceeds the
premium that can be gained from the high protein fraction, so most mills confine
themselves to offering as is meal of 25% protein.
Numerous studies have shown safflower oil to be a good feed product for beef
cattle (112116), dairy cattle (117119), poultry (120123), and lamb (124127),
and is generally available at price levels that are similar to the lowest prices for
alfalfa hay, grain screenings, almond hulls, and other low protein feeds.
Promising experiments have been done to produce protein flour or protein iso-
lates from safflower meal. The USDA compared safflower protein isolate with
isolate from soy and found the safflower product to be quite useful. The study
also outlined the cost of investment and production for the process envisioned
(128130). Other researchers have written extensively on this subject (131134).
A factory would need considerably more than the total U.S. supply of safflower
meal to produce an economically viable protein isolate. Unless a scientific break-
through can materially reduce the hull portion of a safflower seed while retaining
satisfactory yields, meal will continue to sell for a modest price and to be consid-
ered a second-rank product. NIOP Rules 8.1.18.1.3 established the factors guiding
the trade in safflower meal.
5. QUALITY ASSESSMENT
Although most of the standard tests for measuring physical and chemical character-
istics of a product work well for safflower seed and its products, some unique
problems have arisen over the years.
5.1. Safflower Seed

When safflower was first introduced into the United States, the Fred Stein Co. was
the first to produce a chart, the Steinlite moisture meter, calibrated specifically for
safflower seed, allowing moisture to be rapidly and correctly determined at the
elevator. Most moisture meters available today work well on safflower seed.
During its growth cycle, a safflower head will respond to the amount of moisture
available. If there is plenty of moisture, many of the individual seeds that have
begun to form in the head will fill completely. If moisture is restricted or if a sudden
trauma such as disease or removal of water occurs, some of the seeds that have
QUALITY ASSESSMENT 523
begun to fill will stop filling and others will not even begin. This results in a mixture
in each safflower head of some seeds that are plump and completely filled and
others that appear to be the same but that, on inspection, are empty or only partially
filled. For the laboratories performing dockage tests on the thousands of samples
representing each truckload delivered, it can be a daunting task to find which seeds
are empty or partially filled. Originally, the dockage analysis method adopted by the
state of California employed a series of hand screens, followed by winnowing
through a Bates aspirator and hand picking of the resultant sample to find empty
seeds that escaped the aspirator. This method was too slow, and when used to mea-
sure samples containing high amounts of empty hulls, as is often encountered in
Great Plains safflower, up to 30 passes through an aspirator were required to find
all empties.
During the 1950s and early 1960s, PVO and the California State Department of
Agriculture performed hundreds of experiments together aimed at producing a sim-
pler and more reproducible test for dockage. As a result, modifications to the Carter
Dockage Tester were developed that allow consistent measurement of dockage.
This method was adopted by the California Department of Agriculture and by
the NIOP, incorporating the procedure as part of Rules 7.1 g D and E (135).
Determining the oil content of safflower seed in the laboratory by solvent
extraction is also more difficult than for other oilseeds because of the vast differ-
ence in texture of the hull compared to the kernel within. The hull must be cracked
or all of the oil will not be extracted. But in cracking the seed, the kernel tends
to mash as well and small amounts of oil can be lost in the process, a small amount
is important when the sample contains only 5 g of seed. Since many people
expressed dissatisfaction in safflower oil content analyses, PVOs control labora-
tory worked for a long time to develop a better method than the standard AOCS
procedure (136). This method of analysis is now part of the NIOP rules for
safflower (137).
The NIOP also conducted extensive tests to develop methods for better sampling
of safflower seed. Field run safflower seed is fairly difficult to sample. Although
pure safflower seed is relatively smooth flowing, the seed delivered by a farmer
can contain portions of stalks and stems; parts of the head that held seed; and
leaves, and other foreign material. Safflower seed, which has traveled over bumpy
roads for 50200 km inside a truck, for 3000 km in a railroad hopper car, or for
10000 km in the hold of a heavy grain carrier on its way to Japan, tends to stratify,
and unless the sampling device reaches all levels of the product, the sample is not
representative. Japanese buyers, who were receiving 5,00015,000 tons of safflower
at a time, found that the oil content and dockage analyses performed at time of ship-
ment did not reflect what the oil mills obtained as a final outturn in the milling of
the same seeds. The NIOP adopted standard sampling and dividing procedures
aimed at reducing variation in results, and these procedures now are incorporated
in their rules (138).
It is particularly important to remember that the sample used in analyzing saf-
flower oil contents must be first cleaned of all dockage (including empty hulls),
unlike the common method of measuring sunflower oil contents, which is
524 SAFFLOWER OIL
performed on seed containing admixture. This puts a premium on good sampling,

good cleaning, good division of the sample, and consistent performance of the ana-
lysis itself.
Safflower seed oil content can also be determined by the use of nuclear magnetic
resonance (NMR), and today most plant breeders employ NMR techniques to mea-
sure their new lines. NMR techniques can be performed on only one half of a seed,
so the other half can be planted if the results of the analysis are promising. In its
earlier versions, processors tended to feel that NMR analysis produced oil content
results that were slightly higher than found by standard solvent extraction analysis
or than what was actually obtained at the oil mill. This has been disproven in the
case of safflower seed, and the industry has adopted NMR analyses in large part
to speed up paperwork. Because of the relatively small amount of safflower seed
being measured for oil content annually, no one has taken the time to prove that
present-day NMR procedures should be used to substitute for the standard AOCS
procedure.
The USDA published what may have been the first practical procedure for
quickly determining if a truckload of seed is a linoleic or oleic variety (138). It
involves squeezing a few seeds in a small hand-powered press to obtain a few drops
of oil. A drop of oil is placed on the glass prism cell of a hand-held refractometer.
The refractive index has a straight-line relationship with the iodine value or fatty
acid distribution of the oil, hence it is easy to determine if the seed in question
meets an oleic standard or not, so long as a temperature correction is applied.
Recently, it has become simpler to compare the unknown sample to a known oil
standard, eliminating the need to apply a temperature correction. Temperature cor-
rections are difficult to measure accurately in the field under the time pressure of
harvest.
5.2. Safflower Oil

Measurement of safflower oils various chemical and physical characteristics is
quite straightforward and only minor changes have occurred over 50 years in the
rules governing the safflower trade. In 1990, the requirement for certification that
safflower oil demonstrate a negative halphen test was dropped. The emergence of
better and better GLC technology eliminated the need for a color test of cottonseed
oil adulteration.
The U.S. Department of Agriculture Utilization Laboratories at Philadelphia,
Peoria, and particularly Albany, California, contributed a major body of work
that measured various factors that affect the quality of safflower and its reaction
to various processes. Oxidation reactions of safflower oil and methods for following
heat-generated changes in composition during deep-fat frying were studied in depth
(139143). USDA scientists at Albany (144) and Peoria (145) analyzed the head-
space volatiles of safflower methyl esters and safflower oil, respectively, subjected
to accelerated oxidation and found them to be the most reactive of all oils tested.
One study showed vinyl-n-amylkelone to be the compound responsible for the gen-
eration of metallic off-flavors in oxidized safflower oil (146). USDA researchers
STORAGE AND TRANSPORTATION 525
demonstrated the effects of oxyfatty acids, malonaldehyde and diclorocarbenes,

respectively, on oil flavor and storage reactions (147149).
It has been demonstrated that tocopherols in linoleic safflower oil were more
stable than tocopherols in oleic safflower oil (150). The USDA did room odor stu-
dies that showed that oleic safflower did well compared with all other oils used in
the study (151). A broad study was conducted of the effects of various substances
on the oxidation of safflower oil in deep frying (152); of high temperature reactions
in the presence of amino acids (153); and of the effect of amino acids on emulsions
(154), dried emulsions (155, 156), and chemical and organoleptic properties
(155, 156).
5.3. Safflower Meal

As mentioned, safflower meal tends to stratify in storage so the principal problem in
quality control is making sure a truly representative sample is obtained. The USDA
Regional Utilization Laboratory at Albany, California, produced a body of work
concerning safflower meal that allows a better understanding of its attributes and
deficiencies. A survey of the world collections for seeds high in lysine was under-
taken (157), and this work has been continued for both lysine and methionine at the
Eastern Experiment Station of Montana State University (158). The work included
studies of safflower steroids (159161). Another study demonstrated how to remove
deleterious glucosides from safflower meal and then demonstrated possibilities for
removal of these and other negative factors to make safflower meal a more useful
product (162). Others have isolated three conjugated serotin factors (163) and their
related phenolic factors (164).
6. STORAGE AND TRANSPORTATION
6.1. Safflower Seed

The most important element in the storage of safflower seed is anticipation of pro-
blems. If safflower seed buyers maintain contact with the suppliers and inspect the
fields, most problems can be solved before they escalate. Safflower seed that is
below 8% moisture; is free of green weed, seeds, or trash; and has been brought
to room temperature gradually is quite stable and can be stored indefinitely with
no problem. Arranging for outside cleaning and/or drying before delivery to the
oil mill and possible rejection, if the grower is unable to cope with weeds in
the field, is much better than handling such problems on an emergency basis at
time of delivery.
Because oleic and linoleic safflower seeds are virtually identical in appearance,
extreme care is necessary to prevent inadvertent mixing. If a positive paper trail can
be established for identifying fields of linoleic and oleic safflower from time of
planting until delivery to the oil mill or storage point, much more confidence is pos-
sible when the samples are taken and the seed is checked for refractive index to
526 SAFFLOWER OIL
verify positively the type delivered. Each load can be properly directed to its appro-
priate discharge point. If safflower seed is put into storage free of included green
weeds or other plant material, with moisture level that has been brought to equili-
brium at 8% or under, it can be stored indefinitely.
At the time of the year that safflower is harvested, air temperatures are often in
excess of 38 C. If safflower seed is being harvested from a particularly weedy field,
the farmer must be careful to monitor the temperature of the seed in the truck or
trailer into which the harvest is loaded. If a truck is forced to wait overnight in such
conditions, the seed can begin heating to a dangerous level. Similarly, if seed is
brought to a warehouse on a hot day, it is prudent for the warehouse to pull air
through the seed pile or silo until the temperature of the entire mass reaches equi-
librium. Monitoring the temperature within the seed mass by means of thermocou-
ples is mandatory. If temperatures start to rise, air circulation can be started again
until equilibrium is restored. If a hot spot cannot be controlled, it is prudent to reach
that area as soon as possible by turning part of the seed column (in a tank or silo) or
by digging into the side of the pile (if stored in a flat warehouse). This is not a time
to move slowly; tear off the side of the building if that is what it takes to reach a hot
spot. Safflower seed seems to be quite hygroscopic, absorbing moisture from wet
material adjacent to it, which in turn causes germination to start, and hence heating.
Safflower seed that has heated will char to dark brown to black mass and will devel-
op strong odors that can permeate the storage unit if not separated quickly. This is
not to portray safflower seed as a problem seed. Most seed is received in fine con-
dition and stores without problems.
If a problem load is received, it is usually best to process it through a simple
grain cleaner to remove the problem foreign material or, if that is not possible,
to pile the seed in a thin layer on a concrete slab for a few days until it stabilizes.
Safflower seed can be dried in forced air grain dryers. It is wise, however, to avoid
letting temperatures in the dryer exceed 82 C to prevent safflower oil from scorch-
ing. If necessary, two passes through a dryer are to be preferred to one pass at high
temperatures.
Safflower seed is normally not attacked by grain weevils, but these pests
are often attracted to the foreign material contained in seed stored longer than
5 months. Again, careful monitoring is necessary if seed is destined for export.
Treatment with approved fumigants in a timely fashion will avoid having the
shipment graded weevily, which in turn makes the safflower seed sample grade
and subject to rejection by the buyer.
Safflower seed is quite stable under carriage in ocean vessels. Although natural
separations within the vessels interior are preferred, safflower seed can be success-
fully separated from safflower bulk cargoes in the hold by a temporary separation
built from layers of plywood, plastic, and burlap.
6.2. Safflower Oil

Crude safflower oil can be stored and conveyed in normal black-iron vessels or
pipelines without problems. The most important factor is to avoid exposure to air.
UNIQUE USES 527
Safflower salad oil, because of its reactivity, should be stored and shipped in
stainless equipment. The USDA in Peoria found that safflower was more light
stable than soybean or cottonseed oils but still recommended packaging in brown
bottles (165). It was found that a-tocopherol does little to inhibit photoxidation, and
it has been recommended that brown bottling be used to inhibit photoxidation and
to reduce formation of free radicals, which generate objectional off-flavor (166).
The USDA presented an early study of antioxidants (167), and the super qualities
of 20 , 40 , 50 -trihydroxybutyrophenone (TBHQ) have been demonstrated (168). The
limited effects of tempeh oil have been studied (169), and some researchers have
recommended chromans, particularly Trolox C, as producing superior results in sta-
bilizing safflower oil (170). TBHQ was the preferred antioxidant when these pro-
ducts were fashionable. In todays world of environmental concerns, the only
antioxidant used is the addition of citric acid during deodorization. If safflower sal-
ad oil must be stored, nitrogen sparging and blanketing is recommended.
6.3. Safflower Meal

As noted, expeller safflower meal is dangerous to store under any conditions. Sol-
vent-extracted safflower meal is much more stable and can be stored safely so long
as moisture levels are kept low in storage. Safflower meal can hold 10% moisture
under California weather conditions, but it is preferable to maintain it at a 57%
moisture content if storage is for an extended period. Long-term storage in small
diameter tanks should be avoided, because safflower meal tends to bridge under
such conditions and become difficult to remove.
7. UNIQUE USES
7.1. Bird Feed

Approximately 60 million people in the United States provide feed for wild birds
(171). Safflower seed is employed in many wild bird feeding mixes because of its
high oil content; it is also used to feed caged birds, particularly parakeets and par-
rots. Safflower seed is used in the feeding of tame ducks in China and Taiwan and
for tame pigeons in Europe. Safflower fields are considered excellent venues in the
United States by hunters of white wing dove and pheasants, and safflower is often
planted at duck clubs to attract ducks and pheasants.
7.2. Ornamental
Safflower seeds and flowers were used in Egypt to make ornaments, wreaths, and
jewelry (172). Horticultural use is limited because of safflowers prickly foliage, but
it is employed in some places for this reason as a protection for other plants against
children and dogs. The spineless varieties are often used by Spanish and Portuguese
farmers as garden flowers and as sources of food coloring. More recently, the red or
528 SAFFLOWER OIL
orange flowers of spineless safflower have been harvested and then either dried or
sold as a fresh display by florists. If dried carefully, safflower plants and flowers
retain their color and can be used for years in a static display.
7.3. Food Coloring

Safflower has been employed as a source of color and flavor in cooking in every
country where it has been grown with the exception of the countries that have
started large-scale commercial production in the last 40 years. Except for recent
immigrants, safflower has not been extensively so used in the United States. Refer-
ences on the use of safflower florets in cooking are available (1,3). In the areas of
Spain and Portugal where saffron is grown, fields of safflower exist that the local
farmers do not like to acknowledge. Almost surely it was being used as an adulter-
ant to the much more expensive saffron.
7.4. Dye
Another ancient use for safflower is to make dyes. The principal dye, carthamin, is a
bright red colorant that is extracted from red-flowered plants after the yellow dye
has been leached with water. Carthamin (C43H42O22) imparts a scarlet red color to
silk and cotton (173176). Fine examples of its durability from ancient times can be
found in museums in Egypt, China, and Japan (3). Safflower yellow pigment
(C16H20O11) must be removed to allow the red dye to be extracted; in earlier times
the yellow was discarded. A factory has been established in Xinjiang, China, to
manufacture large quantities of both types of dye (173).
7.5. Medicinal
Safflower seed, pollen, florets, and oil have been used for medicinal purposes
almost since cultivation began. In the first century A.D., Pliny wrote that safflower
oil, called oleum cnicium, was used as a milder substitute for castor oil, and Ped-
anius Dioscorides, in De Materia Medica (the leading Western pharmacological
text for 16 centuries), described the use of safflower to color and flavor various
potions and unguents and to act as mild laxatives and flavoring agents (157).
Several Arab texts dating to 10001500 mention safflower as an antidote for poi-
sons and as an agent to induce sweating as a fever cure (172). Known as Kusumbha
oil, safflower was regarded as a purgative in ancient Africa and India. Charred saf-
flower oil is used in India as a treatment for sores and rheumatism, and safflower
seeds are employed as diuretics and tonics (177). China has a long history of use of
all parts of the safflower plant in medicinals combined with many herbal products.
A tea is manufactured in Beijing (178), and a polyamino acid nutrient is manufac-
tured from safflower protein isolates in inner Mongolia. Many discussions on
ancient and modern Chinese medicine are available (7, 173).
USP safflower oil has been used as a carrier for penicillin. Data on use of saf-
flower oil as a vehicle for injection of androgens (179) have been published, and it
UNIQUE USES 529
has been demonstrated that safflower oil could be used as a dipersant for solutions
of intravenous feedings of dextrose (109).
7.6. Cosmetic
Safflower florets were used to color ceremonial ointments in Egyptian tombs (10).
The Journal of American College of Toxicology published a report on the safety of
safflower oil in 1981. The article concluded that safflower oil was safe as a cosmetic
ingredient in the current practices of use (180). Oleic safflower is considered
equally safe, and safflower oil was found to be nonallergenic (181).
Powdered safflower florets have been employed since ancient times as a rouge in
Egypt, China, and Japan, where it was known as beni (7, 177). The Misehnz of
ancient Hebrew literature speaks of safflowers use as a rouge (182). It was mixed
with French chalk and used as a rouge in old England (9). Soot from charred saf-
flower plants was used until recently as the source of kohl, a cosmetic used to dar-
ken the eyelids of Egyptian women (183). In India, safflower oil is used as an
ingredient in soap and the preparation of Macassar hair oil (180).
TABLE 13. Supply of Potential Biodiesel Feedstocks.
Oil Type Total Oil Productiona
Pounds Gallonsb
Millions
Crops
Total 20,030 2,601.3
Soybean 14,935 1,939.6
Corn 2,076 269.6
Cottonseed 1,220 158.4
Sunflowerseed 868 112.7
Canola 353 45.8
Peanut 282 36.6
Flaxseed/linseed 175 22.7
Safflower 118 15.3
Rapeseed 3 0.4
Animal fat
Total 8,772 1,139.2
Lard 1,026 133.2
Edible tallow 1,490 193.5
Inedible tallow 3,623 470.5
Yellow grease 2,633 341.9
Total supply 28,802 3,740.5
a
Pounds of oil production are a 3-year average (19931995) from Oil Crops
Yearbook, October 1997, USDA, ERS with the following exceptions: rapeseed
was calculated by multiplying oil per acre times the 199395 average number
of acres harvested. Number of harvested acres comes from USDA, NASS,
January 1996. Inedible tallow and yellow grease supply comes from U.S.
Department of Commerce, Bureau of the Census, Fats and Oils, Production,
Consumption and Stocks, Annual Summaries 19931995.
b
Pounds are converted to gallons of oil using a 7.7 pounds-to-gallon
conversion rate.
530 SAFFLOWER OIL
7.7. Biodiesel Fuel

Safflower has been examined as a possible biodiesel feedstock. Table 13 is an
example of the amount of oil produced annually in the United States by type of
feedstock. Amounts are reported by weight (i.e., millions of pounds), which are
converted to a liquid volume basis to provide a gallon estimate for each feedstock.
Soybean oil is the largest potential feedstock source for biodiesel. Corn oil, cotton-
seed and sunflowerseed are also relatively large contributors to U.S. vegetable oil
supplies. The production of oil from the other feedstocks is minor, ranging from
353 million pounds for canola to around 3 million pounds for industrial rapeseed.
The total production of oil from crops is about 20 billion pounds per year. Animal
fats and yellow grease add about another 8.8 billion pounds, resulting in about 29
billion pounds of total oil.
On a liquid fuel basis, these feedstocks would equal about 3.7 billion gallons
of diesel fuel, about 13% of the 28 billion gallons of diesel fuel consumed in the
United States for transportation in 1996. If biodiesel was blended with petroleum
diesel fuel, e.g., 20% biodiesel and 80% petroleum (B20), the total supply of this
blended fuel would be about 18.7 billion gallons, or 67% of U.S. annual diesel
consumption. This example uses the total average supply of all crop oils, animal
fats, and yellow grease as the available feedstock supply (182).
REFERENCES
1. D. K. Salunkhe, J. K. Charan, R. N. Adjule, and S. S. Kadam, World Oilseeds, Van

Nostrand Reinhold, Co., Inc., New York, 1992, p. 326.
2. N. W. Simmons, ed., Evolution of Crop Plants, Longman, New York, 1976, pp. 219, 230,
231233.
3. E. A. Weiss, Castor, Sesame and Safflower, Barnes & Noble, New York, 1971, pp. 530
554.
4. A. Ashri, Divergence and Evolution in the Safflower Genus, Carthamus, Final Research
Report USDA PL 480, Project No. A-10-CR-18, Hebrew University, Rehovot, Israel,
1973.
5. A. Ashri and P. F. Knowles, Agron. J., 52(1), 1117 (1960).
6. A. Ashri and P. F. Knowles, Abst. Am. Soc. Agron., 50 (1977).
7. D. Li, ed., Proceedings of the Third International Safflower Conference, Chinese
Academy of Sciences, Bejing, 1993, pp. 3546.
8. Ref. 3, p. 534.
9. C. E. Claassen, ed., Proceedings of the First International Safflower Conference,
University of California, Davis, 1987, pp. 2835.
10. L. Keimer, Die Gartendflanzen im Alten Agypten, Hamburg, Germany, 1924.
11. G. Schweinfurth, Bot. Jahr., 68, 16 (1887).
12. J. S. Ingram, Trop. Sci., 11(3), 177184 (1969).
13. C. L. Madan, B. M. Kapur, and U. S. Gupta, Econ. Bot., 20(4), 377385 (1966).
14. D. Basker and M. Negbi, Econ. Bot., 37(2), 228236 (1983).
REFERENCES 531
15. Ref. 3, p. 531.

16. Ref. 3, p. 529.
17. J. R. Smith in D. R. Erickson, ed., Edible Fats and Oils Processing, American Oil
Chemists Society, Champaign, Illinois, 1990, p. 324.
18. G. Yuan and co-workers, Safflower Genetic Resources and Their Utilization, Science and
Technology Publishing House, Beijing, 1989.
19. Ref. 3, pp. 586590.
20. Ref. 3, p. 742.
21. W. R. Aykroyd, Health Bulletin, No. 23, 4th ed., Ministry of Health, New Delhi, India,
1951.
22. P. J. C. Brauns, Experiences in 1957 and 1959 of the Grazing Value of Safflower, Biloela
Regional Experiment Station, Australia, 1960.
23. S. K. Ranjhan, Ind. Vet. J., 36(6), 267276 (1959).
24. Ref. 3, p. 677.
25. F. Rabak, USDA Circ., 366 (1935).
26. A. Rehbein Sr., Mont. Farmer-Stockman, 35(5), 6 (1948).
27. L. L. Carrick and H. K. Nielsen, Amer. Paint J., 22(22), 79, 1816; 22(45), 1321, 44
46; 22(47), 1214, 20, 4348; 22(48), 2029; 22(49) 5260 (1938).
28. F. Schofield, Natl. Paint, Varn. Lacquer Assoc. Sci. Sec. Cir., 519 522 (1936).
29. J. S. Remington, Paint Manufac., 6, 5052 (1936).
30. J. Van Loon, Verfkronief, 10(4), 280, 282 (1937).
31. H. H. Mann and N. V. Kanitker, J. Soc. Chem. Ind., 38, 3638 (1919).
32. F. Honcamp et al., Tierernahr, 1, 329 (1929).
33. A. T. Pugsley and G. Winter, Aust. Munitions Supply Labs Rep., 171, (1947).
34. J. R. Smith, Safflower, AOCS Publications, Champaign, Illinois, in press.
35. J. Muir, Farm Q, 88, 168172 (1960).
36. R. E. Woodward, Industrial Survey of Safflower, University of Nebraska, Lincoln, 1949.
37. C. E. Claassen and T. A. Kiesselbach, Experiments with Safflower in Western Nebraska,
Bulletin 376, Nebraska Agriculture Experiment Station, Lincoln, Nebraska, 1945.
38. C. E. Claassen, Chem. Dig., 7(3), 1117 (1948).
39. C. E. Claassen, Nebraska Sta. Circ., 87 (1949).
40. C. E. Claassen, M. L. Shuster, and W. W. Ray, Plant Dis. Rep., 33(2), 7394 (1949).
41. C. E. Claassen and A. Hoffman, Crops Soils, 1(8), 57 (1949).
42. P. F. Knowles, Safflower Production in California, Davis, University of California, Davis,
Agricultural Experiment Station, 1949, pp. 17.
43. H. Taller, Calories Dont Count, Simon & Schuster, New York, 1967.
44. L. W. Kinsell et al., J. Clin. Endocrinol. Metab., 12, 909 (1952).
45. E. H. Ahrens et al., Lancet., 1, 6976 (1957).
46. B. B. Stewart et al., Lancet., 2, 521 (1956).
47. D. D. Rubis and D. S. Black, Safflower Recommendations for Arizona, Report 170,
Agriculture Experiment Station, University of Arizona, Tucson, Oct. 1958.
48. D. D. Rubis and D. S. Black, GilaA New Safflower Variety, Bulletin 301, Agriculture
Experiment Station, University of Arizona, Tucson, Nov. 1958.
532 SAFFLOWER OIL
49. L. O. Cummings and I. A. Kneeland, U.S. Patent 3,266,977, Aug. 16, 1966.
50. L. O. Cummings, U.S. Patent 3,669,900, June 13, 1972.
51. L. O. Cummings, J. W. Sjoquist, and J. A. Kneeland, U.S. Patent 3,332,899, July 25,
1967.
52. W. E. Rathjen, L. O. Cummings, and J. A. Kneeland, U.S. Patent 3,278,567, Oct. 11,
1966.
53. Oil Crop Situation and Outlook Yearbook, Economic Research Service, USDA,
Oct. 2002, updated, accessed Aug. 2004.
54. B. Horowitz and G. Winter, Nature, 179, 582 (1957).
55. P. F. Knowles and A. Mitwaki, Econ. Bot., 17, 139 (1963).
56. P. F. Knowles and A. B. Hill, Crop Sci., 4, 406 (1964).
57. P. F. Knowles, A. B. Hill, and J. E. Ruckman, Calif. Agr., 19(12), 15 (1965).
58. J. Dominguez Gimenez, Sesame Safflower Newslett., 3, 63 (1987).
59. FAO, World Crop and Livestock Statistics, 19481955, Process Statistic Series 1, FAO,
Rome, 1987, pp. 332349.
60. FAO, Yearbook Production, Statistical Series 104, Vols. 4046, FAO, Rome, 19871993,
pp. 116, 119, 126, 165.
61. Ref. 3, p. 570.
62. T. H. Applewhite, J. Amer. Oil Chem. Soc., 43(6), 406408 (1966).
63. E. A. Weiss, Oilseed Crops, Longman, London, 1983.
64. D. Li, M. Zhou, and V. R. Rao, eds., Characterizations and Evaluations of Safflower
Germplasm, Geological Publishing House, Beijing, China, 1993, pp. 96233.
65. J. Guggolz et al., J. Amer. Oil Chem. Soc., 45(10), 689693 (1968).
66. A. L. Urie and D. E. Zimmer, Crop Sci., 10, 371372 (1970).
67. W. W. Elbert and P. F. Knowles, Am. J. Bot., 55(4), 421430 (1968).
68. A. L. Urie, Abst. Amer. Soc. Agron., 65 (1976).
69. H. J. Burkhardt and G. Fuller, J. Amer. Oil Chem. Soc., 47, 219221 (1970).
70. H. J. Burkhardt, J. Amer. Oil Chem. Soc., 45(1), 9699 (1968).
71. H. J. Burkhardt, J. Amer. Oil Chem. Soc., 47(2), 6972 (1970).
72. H. J. Burkhardt, J. Amer. Oil Chem. Soc., 48(11) 697699 (1971).
73. A. B. Hill, Oilseeds and New Crops Workshop, University of California, Davis, Feb. 27,
1986.
74. D. L. Smith and C. E. Claassen, Abst. Am. Soc. Agron., 90 (1963).
75. D. D. Rubis, ed., Proceedings of the Third Safflower Research Conference, Davis,
California, 1969, pp. 2732.
76. A. R. Deshmukh, R. M. Patil, and N. Nimbkar, eds., Abstracts Second International
Safflower Conference, Hyderabad, India, 1989, pp. 3233.
77. T. C. Heaton, P. F. Knowles, and B. D. Webster, Abst. Am. Soc. Agron., 163 (1979).
78. T. C. Heaton and P. F. Knowles, Crop Sci., 20(4), 554 (1980).
79. T. C. Heaton and P. F. Knowles, Crop Sci., 47(7), 520522 (1982).
80. Y. C. Doesthale, J. Amer. Oil Chem. Soc., 58(11), 988990 (1981).
81. R. M. Saunders, J. Amer. Oil Chem. Soc., 47(7), 254255 (1970).
82. NIOP, Trading Rules, 19931994 NIOP, Washington, D.C., pp. 7779.
REFERENCES 533
83. Ref. 1, p. 340.

84. G. Robbelen, R. K. Downey, and A. Ashri, Oil Crops of the World, McGraw-Hill, New
York, 1989, pp. vviii.
85. F. D. Gunstone and M. I. Quresh J. Amer. Oil Chem. Soc., 42(11), 961965 (1965).
86. T. N. B. Kaimal and G. Lakshminarayana, J. Amer. Oil Chem. Soc., 56(5), 578580
(1979).
87. R. B. Allen, Feedstuffs, 62(31), 2430 (1990).
88. D. Bath et al., Feedstuffs, 62(31), 3127 (1990).
89. N. Mehrotra and G. P. Shukla, Haryana Agr. Univ. J. Res., 3, 195196 (1973).
90. N. Mehrotra, S. K. Arora, and P. P. Jain in Symposium on the Strategy and Experimental
Approaches for Genetic Improvement of Forages for Increased Animal Productivity,
Indian Society of Genetics, Haryana Agricultural University, Hissar, 1975.
91. I. J. Singh and N. Mehrotra, in Ref. 9, pp. 225227.
92. N. Mehrotra and B. D. Chaudhary, in Ref. 9, pp. 155159.
93. Ref. 3, pp. 761765.
94. P. F. Knowles, Econ. Bot., 21(2), 156162 (1967).
95. Ref. 1, p. 334.
96. Ref. 3, p. 328.
97. G. A. Kopas and J. A. Kneeland, U.S. Patent 3,271,160, Sept. 6, 1966.
98. M. A. Williams, Oil Mill Gazet., 95(3) (1989).
99. M. A. Williams and J. F. Crawford, in press.
100. M. A. Williams, U.S. Patent 4,901,635, Feb. 20, 1990.
101. G. M. Neumunz, J. Amer. Oil Chem. Soc., 55(5), 396A398A (1978).
102. J. E. Blum, J. Amer. Oil Chem. Soc., 43(6), 416417 (1966).
103. W. H. Schmidt, J. P. McNaught, and K. S. Baker, U.S. Patent 3,240,608, Mar. 15,
1966.
104. T. H. Applewhite, ed., Baileys Industrial Oil and Fat Products, vol. 3, Wiley, New York,
1985, p. 6086.
105. S. Zalewski and F. A. Kummerow, J. Amer. Oil Chem. Soc., 45(2), 8792 (1968).
106. G. Ben-et, A. Doley, M. Schimmel, and R. Stern, J. Amer. Oil Chem. Soc., 49(4), 205207
(1972).
107. E. N. Frankel, J. Amer. Oil Chem. Soc., 47(1), 1114 (1970).
108. E. N. Frankel, F. L. Thomas, and J. C. Cowan, J. Amer. Oil Chem. Soc., 47(12), 497500
(1970).
109. R. S. Singleton, R. R. Benerito, and J. L. White, J. Amer. Oil Chem. Soc., 37(2), 8892
(1960).
110. H. L. Hanson and L. R. Fletcher, Food Technol., 15, 256261 (1961).
111. Acreage, National Agricultural Statistics Service, USDA, Washington, D.C., released
June 30, 2004.
112. M. L. Baker and co-workers, Feeding Safflower Meal., Bulletin 402, Nebraska Agri-
culture Experiment Station, Jan. 1951.
113. M. L. Baker, G. N. Baker, and J. K. Matsushima, Safflower Meal as a Supplement for
Cattle, Bulletin 447, Nebraska Agriculture Experiment Station, Aug. 1959.
534 SAFFLOWER OIL
114. G. N. Baker, T. Kosku, and M. L. Baker, Undecorticated Safflower Meal as a Protein

Supplement for Wintering Calves, Bulletin 458, Nebraska Agriculture Experiment
Station, Sept. 1960.
115. N. W. Hilston, C. B. Boubicik and L. Paules, Comparative Value of Soybean Meal,
Safflower Meal and Urea for Fattening Steers, Wyoming Agriculture Experiment Station,
Aug. 1951.
116. N. W. Hilston, C. P. Stroble, and L. Paules, Comparative Value of Soybean Meal,
Safflower Meal and Urea for Fattening Steers, Wyoming Agriculture Experiment Station,
June 1952.
117. Anon., Idaho Agr. Sci., 38(3), 2 (1953).
118. R. H. Ross, G. W. Cleveland, and D. L. Fourt, Am. Dairy Sci. Soc., West. Div. Proc., 34,
7273 (1953).
119. R. H. Ross, G. W. Cleveland, and D. L. Fourt, J. Dairy Sci., 37(6), 671672 (1954).
120. D. W. Pittman and C. I. Draper, Safflower, Its Possibilities and Culture in Utah,
Agriculture Experiment Station, Logan, Utah, 1955.
121. F. M. Kratzer and D. Williams, Poultry Sci., 26(6), 623625 (1947).
122. F. M. Kratzer and D. Williams, Poultry Sci., 30(3), 417421 (1951).
123. H. R. Halloran, Feedstuffs, 33(45), 7071 (1961).
124. E. K. Faulkener and L. Paules, Saffmeal for Fattening Lambs, Wyoming Agriculture
Experiment Station, Aug. 1951.
125. E. K. Faulkener and L. Paules, Saffmeal for Fattening Lambs2nd Years Trials,
Wyoming Agriculture Experiment Station, Mar. 1952.
126. H. Goss and K. K. Otagaki, Calif. Agr., 8(5), 15 (1954).
127. R. M. Dixon et al., in D. J. Farrell, ed., Recent Advances in Animal Nutrition in Australia,
1989, p. 1524.
128. A. A. Betschart, J. Amer. Oil Chem. Soc., 56(3), 454457 (1979).
129. A. A. Betschart, and R. M. Saunders, J. Food Sci., 43, 964 (1978).
130. A. A. Betschart, R. Y. Fogn, and M. M. Hanamoto, J. Food Sci., 44(4), 10221026, 1035
(1979).
131. O. Paredes Lopez and C. Ordorica Falomir, J. Sci. Food Agr., 37(11), 10971103 (1986).
132. O. Paredes Lopez and C. Ordorica Falomir, J. Sci. Food Agr., 37(11), 11041108 (1986).
133. O. Paredes Lopez, C. Ordorica Falomir, and T. A. Carabez, Leb.-Wissensch. Technol.,
21(6), 328333 (1988).
134. D. N. Kulkarni, K. D. Kulkarni, and D. R. Vencatasamy, in the Third International
Safflower Conference, Beijing, China, 1993.
135. Ref. 82, pp. 99101.
136. E. Jacobsen in First Research Conference on Utilization of Safflower, USDA ARS 7474,
Albany, California, 1967, pp. 2831.
137. Ref. 82, pp. 9799.
138. M. H. Neustadt, Rapid Testing of Oilseeds for Oil Quality and Iodine Number of Oil,
USDA Technical Bulletin 1171, Washington, D.C., Oct. 1977.
139. E. N. Frankel, C. D. Evans, and J. C. Cowan, J. Amer. Oil Chem. Soc., 37(9), 418424
(1960).
140. C. D. Evans et al., J. Amer. Oil Chem. Soc., 37(12), 452 (1960).
REFERENCES 535
141. E. N. Frankel et al., J. Amer. Oil Chem. Soc., 37(12), 418 (1960).
144. R. J. Horvat et al., J. Amer. Oil Chem. Soc., 42(12), 11121115 (1965).
145. J. M. Snyder, E. N. Frankel, and E. Selke, J. Amer. Oil Chem. Soc., 62(12), 16751679
(1985).
146. D. A. Forss, E. H. Ramshaw, and W. Stark, J. Amer. Oil Chem. Soc., 39(6), 308 (1962).
147. D. P. Schwartz and A. H. Rady, J. Amer. Oil Chem. Soc., 67(10), 635641 (1960).
148. P. Fitton, E. H. Pryde, and J. C. Cowan, J. Amer. Oil Chem. Soc., 42(1), 1419 (1965).
149. H. G. Kenney, D. Kamanowski, and A. N. Wrigley, J. Amer. Oil Chem. Soc., 42(1), 1922
(1965).
150. G. Yuki, and Y. Ishikawa, J. Amer. Oil Chem. Soc., 53(11), 673676 (1976).
151. C. D. Evans et al., J. Amer. Oil Chem. Soc., 49(10), 578582 (1972).
152. R. J. Sims, J. A. Fiorti, and M. J. Kanuk, J. Amer. Oil Chem. Soc., 49(5), 298301 (1972).
153. R. J. Sims and J. A. Fiorti, J. Amer. Oil Chem. Soc., 52(5), 144147 (1975).
154. T. Riisom, R. J. Sims, and J. A. Fiorti, J. Amer. Oil Chem. Soc., 57(10), 354358 (1980).
155. J. A. Firoti, H. S. Stahl, J. Cseri, and R. J. Sims, J. Amer. Oil Chem. Soc., 52(10), 395399
(1975).
156. J. A. Fiorti, M. J. Kanuk, and R. J. Sims, J. Amer. Oil Chem. Soc., 51(6), 219 (1974).
157. R. Palter, G. O. Kohler, and P. F. Knowles, J. Agr. Food Chem., 17(6), 12981300
(1969).
158. Ref. 9, pp. 214217.
159. R. Palter, W. F. Haddon, and R. E. Luddin, Phytochemistry, 11(7), 23272328 (1972).
160. R. Palter, R. E. Luddin, and G. Fuller, Phytochemistry, 11(2), 819822 (1972).
161. R. Palter, R. E. Luddin, and W. F. Haddon, Phytochemistry, 11(9), 28712874 (1972).
162. C. K. Lyon et al., J. Amer. Oil Chem. Soc., 56(5), 560564 (1979).
163. S. Sakamura, Y. Terayama, and S. Kawakatsu, Agr. Biol. Chem., 42(9), 18051806
(1978).
164. S. Sakamura et al., Agr. Biol. Chem., 44(12), 29512954 (1980).
165. H. A. Moser, C. D. Evans, J. C. Cowan, and W. F. Kwolek, J. Amer. Oil Chem. Soc., 42(1),
3033 (1965).
166. J. A. F. Faria and M. K. Mukai, J. Amer. Oil Chem. Soc., 60(1), 7781 (1983).
167. R. E. Beal, H. A. Moser, and O. L. Brekke, J. Amer. Oil Chem. Soc., 35(2), 9799
(1958).
168. J. W. Thompson and E. R. Sherwin, J. Amer. Oil Chem. Soc., 43(12), 683696 (1966).
169. P. Gyorgy, K. Murata, and Y. Sugimoto, J. Amer. Oil Chem. Soc., 51(8), 377379 (1974).
170. W. M. Cort, A. J. W. Scott, and M. Araujoetal, J. Amer. Oil Chem. Soc., 56(6), 174178
(1975).
171. D. Lilleboe, Sunflower, 20(1), 1214, 1994.
172. Ref. 3, p. 530.
173. D. Li, ed., Abstracts on Safflower, Chinese Academy of Sciences, Beijing, 1993.
174. S. Buelaverie et al., eds., The Merck Index, 11th ed., Merck & Co., Inc., Rahway,
New Jersey, 1989, p. 285.
536 SAFFLOWER OIL
175. F. Mayer, The Chemistry of Natural Colouring Matters, Reinhold Publishing, New York,
1943.
176. R. S. Lewis, Sr., ed., The Condensed Chemical Dictionary, 11th ed., Van Nostrand
Reinhold, New York, 1993, p. 227.
177. Ref. 3, pp. 533534.
178. D. Li and Y. Han, in Ref. 7, pp. 837843.
179. S. L. Shapiro, L. Freeman and S. Kobrin, Arch. Int. de Pharmacodyn., 3(1), 3035 (1957).
180. L. E. Cathers, ed., J. Am. Coll. Toxicol., 4(5), 171197 (1985).
181. D. H. Goodman, J. Allergy., 35, 3842 (1964).
182. I. Low, Die Flora der Juden, Vol. 1, Vienna, 1926.
183. G. Brunton, Qau and Badri, Vol. 1, London, p. 63.
184. U.S. Biodiesel Development: New Markets, Economic Research Service, USDA, 1997.
12
Sesame Oil
Lucy Sun Hwang
National Taiwan University
Taipei, Taiwan
1. INTRODUCTION
Sesame (Sesamum indicum L.) is believed to be one of the most ancient crops culti-
vated by humans (1). It was first recorded as a crop in Babylon and Assyria over
4000 years ago. The seeds of the crop are used both as condiment and oil source.
The Babylonians made wine and cakes with sesame seeds, whereas sesame oil
was used for cooking, medicinal, and cosmetic purposes. Ancient Indians used
sesame oil as lighting oil, and sesame seeds were commonly used in the religious
rites of Hindus. The Chinese believed that sesame seeds could promote health and
longevity.
Sesame seed has higher oil content (around 50%) than most of the known oil-
seeds although its production is far less than the major oilseeds such as soybean or
rapeseed due to labor-intensive harvesting of the seeds. Sesame oil is generally
regarded as a high-priced and high-quality oil. It is one of the most stable edible
oil despite its high degree of unsaturation. The presence of lignan type of natural
antioxidants accounts for both the superior stability of sesame oil and the beneficial
physiological effects of sesame.
In Asia, sesame oil is obtained by pressing the roasted oilseeds and consumed
as a naturally flavored oil without refining. In the western world, sesame oil is
extracted by a multiple-step mechanical expeller and either the virgin oil or the
537
538 SESAME OIL
refined oil is used for salad dressing. After pressing out oil, the remaining sesame
meal contains high-quality protein suitable for human consumption as well as ani-
mal feed. It is also a good source of water-soluble antioxidants.
In this chapter, the properties and processing of sesame oil will be presented, and
the antioxidative components and their effects on oil stability and health will be
summarized.
2. BOTANY OF SESAME
Sesame (Sesamun indicum. L., synonymous with Sesamun orientale L.), also
known as benniseed (Africa), benne (Southern United States), gingelly (India), gen-
gelin (Brazil), sim-sim, semsem (Hebrew), and tila (Sanskrit), is the worlds oldest
oil crop. It belongs to the Tubiflorae order, Pedaliaceae family, which comprises
of 16 genera and some 60 species (2). There are 37 species under the Sesamum
genus (3). Among the 37 species, only Sesamum indicum is widely cultivated.
The wild species such as S. angustifolium, S. calycium, S. baumii, S. auriculatum,
S. brasiliense, S. malabaricum, S. prostratum, S. indicatum, S. radiatum, S. occiden-
tale, and S. radiatum are cultivated in Africa, India, or Sri Lanka in small areas.
The wild species, although low in oil contents, may contribute to favorable agro-
nomic characters (such as resistance to disease, pests, and drought) when used in
plant breeding.
As most of the wild species of sesame were found in Africa, it is generally
believed that sesame originated in Africa. India may also be the origin of some
species (S. capense, S. prostratum, and S. schenckii) of sesame (2, 4). The sesame
species in the Middle East are similar to Africa; they are believed to be spread from
Africa via Egypt (2). Sesame seeds were brought to India and Burma from Africa
and the Middle East (4). Cross-fertilization of the species from Africa and India
results in a large variety of sesame species. India, therefore, became the secondary
center of diversity. Both China and Japan are the major consumers of sesame
seeds; their sesame seeds were introduced from the Middle East as early as in
500 to 700 B.C. Sesame was brought to the United States by slaves from Africa
in the late seventeenth century. The sesame seeds are still known as benne in
the southern parts of United States, a term similar to the African name of sesame
(benniseed).
Sesame grows in tropical and subtropical areas about 40 N latitude to 40 S
latitude (5). Sesame indicum L. is the commonly cultivated species of sesame. It
has 26 somatic chromosomes (2n 26). Sesame is an annual, erect herb that
may grow between 50 cm and 250 cm in height, depending on the variety and grow-
ing conditions. The stems (Figure 1) may have branches and are obtusely quadran-
gular, longitudinally furrowed, and densely hairy. The extent of hairiness on the
stem can be classified as smooth, slightly, and very hairy; it is related to the variety
of sesame. The degree and type of branching of the stem are also important varietal
characters (6).
BOTANY OF SESAME 539
Figure 1. The plant of sesame.
Sesame leaves are hairy on both sides and are highly variable in shape and size
not only among different varieties but also on the same plant. The lower leaves are
opposite, ovate, sometimes palmately lobed or palmately compound, dull green
in color, 317.5 cm long and 17 cm wide, and coarsely serrate, and the petiole
is 5 cm in length. The upper leaves are alternate or subopposite, lanceolate, and
entire or with a few coarse teeth, and the petiole is 12 cm long. The arrangement
of leaves influences the number of flowers born in the axils and thus the seed yield
per plant.
Sesame has large, white, bell-shaped flowers. The flowers are zygomorphic, in
axils of upper leaves, born singly or 23 together, short-pedicelled, and geniculate.
The calyx is small and five parted, and the segments are ovate-lanceolate and 0.5
0.6 cm long. The corolla is tubular-campanulate, 34 cm long, widened upward,
two-lipped, five-lobed with middle lower lobe longest, pubescent outside, white,
pink, or purplish in color with yellow or purple blotches, spots, and stripes
on inner surface. The stamens are four in number, didynamous, and inserted on
the base of the corolla; the anthers are sagittate. The ovary is superior and two-
celled (7).
The fruit of sesame is a capsule (25 cm long and 0.52 cm in diameter), and it
is erect, oblong, brown or purple in color, rectangular in section, deeply grooved
with a short, triangular beak (Figure 2). The capsules may have four, six or eight
rows of seeds in each capsule (Figure 2). Most of the sesame capsules have four
rows of seeds, with a total of 70 seeds per capsule. The capsules with a wider
540 SESAME OIL
Figure 2. Sesame fruits with four (A) or eight (B) rows of seeds in each capsule.
diameter will usually have higher rows of seeds and the total number of seeds per
capsule can be as high as 100200. When the fruit is ripened, it dehisces by split-
ting along the septa from top to bottom (so called open sesame).
Sesame seeds are small (34 mm long and 1.52 mm wide), flat, ovate (slightly
thinner at the hilum than at the opposite end), smooth, or reticulate. The color varies
from white, yellow, gray, red, brown, to black. The weight of 1000 seeds is around
2.5 to 3.5 g. Sesame seeds consists of testa (exo and endo), endosperm, and coty-
ledon (Figure 3). The oil drops are located in the cotyledon. It is generally believed
that the light-colored seeds with thin coats are higher in quality and oil content than
the dark-colored seeds.
Although sesame seeds are higher in oil contents than most other oilseeds and
sesame oil has good flavor and oxidation stability, sesame seeds have never been a
major oil source. The low yield (400500 kg/ha) of sesame seeds and the labor-
intensive harvesting procedure are the limiting factors. When sesame capsules
Figure 3. Structure of the sesame seed (A) and the oil drops in cotyledon (B).
WORLD PRODUCTION 541
are mature, they are fragile and will burst open easily, scattering the seeds on the
ground and thus difficult to collect. Harvesting of sesame seeds is usually per-
formed by cutting the plant stalks and stacking them vertically under the sun
with the cut-ends downward in the threshing yard. Each dried stalk is then shaken
or beaten over a cloth to catch the seeds that fly out from the dried capsules. The
plant breeders have been trying to develop sesame varieties that do not dehisce
when the capsules are mature and thus can be adapted to mechanical harvest
(810). In the middle of the twentieth century, horticulturists developed sesame
with papershell capsules, which is indehiscence allowing mechanical harvesting
and is easier to thresh than the normal type (11). Until today, however, more than
99% of the sesame produced in world is still harvested manually. Numerous efforts
have been made to move sesame from a labor-intensive harvest crop to a mecha-
nically harvest crop for the past 60 years. Considerable progress was made between
1940 and 1965, but there was still a limited amount of manual labor necessary in
the harvest. The first completely mechanized cultivars were developed in the early
1980s, and there has been continuing progress. Progress in mechanizing sesame has
been slow because of the need to combine many characters in order to compromise
between machine-harvesting and plant characteristics such as seed yield and qua-
lity, disease resistance, insect resistance, hail resistance, and drought resistance.
Sesame can become a major oilseed only with lower price achieved by increasing
yields and reducing production costs (12).
3. WORLD PRODUCTION
3.1. Sesame Seed

Sesame ranks eighth in the world production of edible oil seeds. The total annual
production of sesame seeds is around 3 million metric tons (MT) worldwide from
2000 to 2002. This number has increased 33% since 1990. Figure 4 shows the total
tonnage together with the total area of world sesame production from 1990 to 2003.
It is evident that there is a steady increase of both the seed production and the area
of harvest. The highest sesame seed production reached 3.2 million MT in 2001,
with a total harvesting area of 7.5 million hectares (ha) worldwide. The average
yield of sesame seed is around 400 kg/ha worldwide (Table 1). Among the five
continents, Asia has the highest area of harvest (4.6 million ha), which produces
2 million MT of sesame seed annually. Europe has the lowest quantity of seed pro-
duction (only 0.057% of the world total) but the highest yield (4968.5 kg/ha)
of sesame seed. This yield is ten times that of Asia where more than 70% of worlds
sesame seeds are produced. Africa, the origin of sesame seed, is the second
largest sesame-producing continent. It has, however, the lowest yield (only
328 kg/ha) of sesame seed.
China, India, Sudan, Myanmar, and Uganda are the worlds major sesame seed
producing countries. In 2003, China produced 825 thousand MT of sesame seed and
was the worlds largest sesame-producing country followed by India (620,000 MT),
542 SESAME OIL
3.5 8
Tonnage ( 106 Mt )
3 7
6
2.5
Area (106 ha)

5
2 Area of harvest
Seed production 4
1.5
3
1
2
0.5 1
0 0
91
92
96
97
98
99
03
90
93
94
95
00
01
02
19
19
19
19
19
19
19
20
19
19
19
20
20
20
Years
(Data source: FAOSTAT database)
Figure 4. World production of sesame seed (19902003). (This figure is available in full color at
http://www.mrw.interscience.wiley.com/biofp.)
Myanmar (225,000 MT), Sudan (122,000 MT), and Uganda (106,000 MT). These
five countries together supply nearly 70% of the worlds total sesame seed
(Figure 5). Figure 6 shows the fluctuation in annual seed production by these coun-
tries from 1990 to 2003. As the crop yield is very dependent on moisture, the seed
production can vary up or down in any given year due to rainfall. According to
FAO statistics (13), the yield of sesame seed in China grew rapidly from around
700 kg/ha in 1990 to 1099 kg/ha in 2003, whereas India remained around 300 kg/
ha for the past 15 years. Sudan is the lowest among the five major sesame producing
countries in per hectare yield (150220 kg/ha) followed by Mayamar (170
380 kg/ha). Uganda has a relatively high yield (500 kg/ha) of sesame seed, but
the area of harvest is the lowest among the five countries.
TABLE 1. Production of Sesame Seed in the Five Continents in 2003.1
Continent Seed Production (1000 Mt) Area of Harvest (1000 ha) Yield (kg/ ha)
2
Africa 603.827 (21.835%) 1840.382 (27.547%) 328.099
Asia 2014.492 (72.846%) 4602.432 (68.889%) 437.702
Europe 1.575 (0.057%) 0.317 (0.005%) 4968.454
North and Central
America 65.870 (2.382%) 127.254 (1.905%) 517.626
South America 79.655 (2.880%) 110.485 (1.654%) 720.958

World Total 2765.419 (100%) 6680.870 (100%) 413.931
1
Based on FAOSTAT database (2003).
2
Data in parenthesis are the percentage of total.
WORLD PRODUCTION 543
Others China
(866,888 Mt) (825,531 Mt)
32% 30%
Uganda
(106,000 Mt) India
Myanmar (620,000 Mt)
4% Sudan
(225,000 Mt) 22%
(122,000 Mt)
8%
4%
(Data source: FAOSTAT database)

Figure 5. Major sesame seed-producing countries and their percentage shares of the world
production in 2003.
In 2000, the world trade of sesame seed was 620,000 MT, which was 21.5%
of the total production. Japan imported 165,000 MT (26% of the world imports)
and was the largest importer of sesame seed. South Korea was the second
largest importer (70,000 MT) followed by United States (49,000 MT), Taiwan
(35,000 MT) and Egypt (34,000 MT). Although China and India are the top two
sesame seed producers, most of the seeds are consumed locally. Only 1215%
of the sesame seeds produced in India were exported in the past ten years. China
was the world number one sesame seeds exporting country, which exported
0.9
0.8 China
India
0.7
Sudan
Tonnage (106 Mt)
0.6 Myanmar
0.5 Uganda
0.4
0.3
0.2
0.1
0
91
93
94
95
96
97
98
99
00
01
02
03
90
92
19
19
19
19
19
19
19
19
20
20
20
20
19
19
Years (Data source: FAOSTAT database)

Figure 6. Major sesame seed-producing countries (19902003). (This figure is available in full
color at http://www.mrw.interscience.wiley.com/biofp.)
544 SESAME OIL
1725% of its sesame production before 1996. Because of the fast economic
growth in China, domestic demand of sesame seed increased tremendously
after 1996. Although China became the worlds biggest sesame seed producer since
1997 (Figure 6), the export of sesame seed from China dropped from 119,000 MT
(in 1996) to 41,000 MT (in 1997). Starting from 1996, Sudan became the worlds
top sesame exporting country followed by India and China.
3.2. Sesame Oil

Each year, the world consumes close to 120 million MT of edible fats and oils (14).
Soybean oil is the leading oil that accounts for 30% of the world production of
edible fats and oils. In 2003, it is closely followed by palm oil, whereas rapeseed
oil ranked third has only one-third of the production tonnage of soybean oil.
Sesame oil, with an annual production of 760,000 MT in 2003, is the twelfth largest
vegetable oil produced in the world, higher in quantity than olive oil and safflower
oil (13). The production of sesame oil increased 20% in the recent 10 years, it was
632,000 MT in 1992. China has almost doubled the production of sesame oil
(from 142,000 to 210,000 MT), whereas India has decreased the production by
44% (from 236,000 to 131,000 MT) in the above period. Both China and India
are the largest producers of sesame oil, together they account for nearly half of
the total world production of sesame oil. Besides China and India, Myanmar,
Sudan, and Japan are the top five sesame oil producers.
Sesame seed contains high levels of fat and protein. The chemical composition of
sesame seed varies with the variety, origin, color, and size of the seed. The fat con-
tent of sesame seed is around 50% whereas the protein content is around 25%.
Table 2 lists the proximate composition of sesame seeds from different sources.
Sesame seed contains about 5% of ash, whereas the fiber and carbohydrate contents
show large variation. Crude fiber from one variety of Nigerian black sesame
was reported to have 19.6% of crude fiber (15), whereas one variety of Taiwanese
TABLE 2. Proximate Composition of Sesame Seed (%).

Crude Crude
Sesame Crude Fat Protein Carbohydrate Fiber Ash Moisture Reference
Black sesame 35.8 17.2 9.19 19.6 4.01 4.73 15

White sesame 34.6 20.8 9.19 14.2 10.1 4.14 15
Brown sesame 41.3 20.2 10.3 18.6 5.19 4.12 15
Yellow sesame 53.8 22.0 6.85 13.0 6.09 4.28 15
Black sesame 48.456.7 22.830.3 3.410.8 2.87.2 4.45.5 4.66.4 16
White sesame 50.151.7 22.624.1 7.913.2 5.37.5 4.24.5 4.44.7 16
Brown sesame 46.353.1 21.827.6 4.713.6 3.77.3 3.95.4 5.08.2 16
Nigerian sesame 51.5 20.0 12.5 6.0 5.0 5.0 17
whole seed
Dehulled seed 55.0 24.3 10.4 2.0 3.0 5.3 17
black seed contained only 2.81% of crude fiber (16). The carbohydrate content
ranged from 3% to 14% (1517).
Sesame seed has about 17% seed weight as hull, which is high in oxalic acid
(23%), calcium, and crude fiber. Oxalic acid could complex with calcium and
reduce its bioavailability; indigestible fiber would reduce the digestibility of pro-
tein. Sesame seed hull is therefore recommended to be removed if sesame meal
is used for human food (18). When sesame seed is properly dehulled, the oxalic
acid content can be decreased to less than 0.25% of the seed weight (19). After
dehulling, the fat and protein contents are raised, whereas the fiber, ash, and carbo-
hydrate contents are lowered (Table 2).
4.1. Content of Oil

Sesame seed is a rich source of edible oil. It contains more oil than the major oil-
seeds, such as soybean, rapeseed-canola, sunflower seed, and cotton seed. The oil
content of sesame seed varies with the variety of sesame; it may range from 28%
to 59% (2022). The wild seeds contain less oil (around 30%) than the cultivated
seeds because the oil content is an important criterion for seed selection in agri-
culture practice. In general, the cultivated seed has around 50% oil, whereas the
color of the seed coat exhibits slight influence on the oil content. Black seeds
appear to contain slightly less oil than the white and brown seeds in the Japanese
strains (Table 3). The oil content was found to be influenced also by the growing
condition, daily mean temperature, and the cumulative degrees of daily temperatures
during reproductive stage, which showed negative correlation with the oil content
(23).
TABLE 3. Oil Content of Sesame Seed.
Sesame Color of Seed Oil Content

Species Coat (% Seed) Reference
Sesamum indicum L.
Sudan strainsa Black 50.7 20
Sudan strains Brown 52.3 20
Sudan strains White 47.4 55.5 20
Japanese strainsb Black 43.4 51.1 21
Japanese strains Brown 50.5 56.5 21
Japanese strains White 51.8 58.8 21
Turkish strainsc Black 43.3 48.2 22
Turkish strains Brown 42.8 46.9 22
Turkish strains White 43.1 46.3 22
Sesamum alatum T.d Brown 28.1 29.8 20
Sesamum radiatum S. and T.d Black 30.3 33.4 20
Sesamum angustifolium E.d Black 29.229.7 20
a
The cultivated species of sesame grown in Sudan.
b
Forty-two species of sesame grown in Japan.
c
The cultivated species of sesame grown in Turkey.
d
The wild species of sesame grown in Sudan.
546 SESAME OIL
Table 4 lists the chemical and physical properties of sesame oil (24).
TABLE 4. Chemical and Physical Characteris-

tics of Sesame Oil (24).
Properties Range
Relative density 0.915 0.924

(20 C/water at 20 C)
Refractive index 1.465 1.469
(ND 40 C)
Saponification value 186 195
(mg KOH/g oil)
Unsaponifiable matter 20
(g/kg)
4.2. Fatty Acid Composition

Sesame oil belongs to the oleic-linoleic acid group. It has less than 20% saturated
fatty acid, mainly palmitic (7.912%) and stearic (4.86.1%) acids. Oleic acid and
linoleic acid constitute more than 80% of the total fatty acids in sesame oil. Unlike
other vegetable oils in this group, the percentages of oleic acid (35.942.3%) and
linoleic acid (41.547.9%) in the total fatty acids of sesame oil are close (Table 5).
Table 5 lists the first FAO/WHO Codex Alimentarius Standard of the sesame
oil fatty acid composition as reported by OConnor and Herb (25) and the most
recent Codex Standard (24). Besides the four major fatty acids, there are low
TABLE 5. Fatty Acid Composition of Sesame Oil (% Total Fatty Acids).
Kamal-Eldin and Appelqvist(20)

Fatty Acid Codex(24) OConnor(25) Cultivateda Wildb
Myristic (C14:0) NDc-0.1 <0.5

Palmitic (C16:0) 7.912.0 7.012 9.09.6 8.212.7
Palmitoleic (C16:1) 0.10.2 <0.5 0.10.2 0.20.3
Heptadecanoic (C17:0) ND-0.2
Heptadecenoic (C17:1) ND-0.2
Stearic (C18:0) 4.86.1 3.56.0 5.66.4 5.69.1
Oleic (C18:1) 35.942.3 3550 41.945.2 34.348.1
Linoleic (C18:2) 41.547.9 3550 38.041.6 33.248.4
Linolenic (C18:3) 0.30.4 <1.0 0.50.6 0.60.9
Arachidic (C20:0) 0.30.6 <1.0 0.3 0.20.8
Eicosenoic (C20:1) ND-0.3 <0.5 0.1 0.1
Behenic (C22:0) ND-0.3 <1.0 0.1 0.1
Lignoceric (C24:0) ND-0.3 trace trace
a
Sesamum indicum L.
b
Sesamum alatum T, Sesamum radiatum S. and T., Sesamum angustifolium E.
c
ND: Not detected.
percentages (less than 1%) of other fatty acidsmyristic (ND-0.1), palmitoleic

(0.10.2), heptadecanoic (ND-0.2), heptadecenoic (ND-0.2), linolenic (0.30.4),
arachidic (0.30.6), eicosenoic (ND-0.3), behenic (ND-0.3), and lignoceric acid
(ND-0.3). Fatty acid composition varies with the species of sesame seed (20, 22).
Species with high oleic acid and linoleic acid contents are often selected for planta-
tion (22). Sesame oils from the wild seeds, therefore, are higher in saturated fatty
acids than oils from the cultivated sesame seeds (Table 5).
Fatty acid compositions of different lipid classes in sesame oil also show varia-
tion. The major sesame seed lipid is triacylglycerol, which represents nearly 90% of
the total lipid (20). It has a lower percentage of saturated fatty acids and a higher
percentage of unsaturated fatty acids than the other lipid classes, namely, diacylgly-
cerol, free fatty acid, polar lipid, and steryl ester. Slightly higher percentages of
long-chain fatty acids (20:0, 20:1, 22:0, and 24:0) were found in lipid classes other
than triacylglycerol (20, 26).
4.3. Sterols
Sesame oil is relatively high in unsaponifiable matter (2%) compared with other
vegetable oils. The unsaponifiable matter includes sterols, triterpenes and triterpene
alcohols, tocopherols, and sesame lignans. Sterols are present in vegetable oils in
free form or as sterol esters, sterol glucosides, or esterified steryl glucosides, but free
sterols and sterol esters are often the dominant forms. Among the three classes of
sterols, desmethylated sterol is the major one (8589% of total sterols) followed by
monomethylated (911%) and dimethylated (24%) sterols in sesame oil (27).
According to the Codex Standard, sesame oil may contain as high as 1.9% of total
sterols; it is one of the richest oil source of phytosterols (24). Table 6 lists the levels
of desmethylsterols composition in sesame oil. b-sitosterol is the most abundant
sterol in sesame oil. There are also campesterol, stigmasterol, 5 -avenasterol, 7 -ave-
TABLE 6. Levels of Desmethylsterols in Sesame Oil.a
Kamal-Eldin and Appelqvist (27)

-
Desmethyl Sterol Codex (24) Cultivated Sesameb Wild Sesamec
Cholesterol 0.10.5 0.10.2 0.20.3

Brassicasterol 0.10.2
Campesterol 10.120.0 12.516.9 10.320.5
Stigmasterol 3.412.0 6.08.7 4.414.2
b-sitosterol 57.761.9 57.562.0 33.960.2
5 -avenasterol 6.27.8 8.111.5 12.423.5
7 -stigmasterol 0.57.6 0.43.1 0.13.0
7 -avenasterol 1.25.6 0.31.3 0.93.7
Others 0.79.2 3.66.1 4.67.3
Total sterols (mg/kg) 450019000 43356764 342010005

a
Expressed as a percentage of total sterols.
b
c
548 SESAME OIL
nasterol, and 7 -stigmasterol present in descending abundance. Only a trace

amount (<0.5%) of cholesterol was found in sesame oil. Oils from the wild species
of sesame contain higher levels of sterols, especially 5 -and 7 -avenasterols.
These two sterols having the 24;28 ethylidene side chain showed antipolymeriza-
tion effects that could protect vegetable oils from high-temperature oxidation (28).
Phytosterols and cholesterol have similar structures; phytosterols are therefore
competitors of cholesterol absorption. Consumption of phytosterol may lower
blood cholesterol and thus protect from cardiovascular diseases (29). Phytosterol,
especially, b-sitosterol, inhibits the growth of human colon cancer cell (30), pros-
tate cancer cell (31), and breast cancer cell (32).
4.4. Tocopherols
Sesame oil is well known for its oxidative stability; one of the reasons for this
extra-stability is attributed to its tocopherol content. The total tocopherol content
of sesame oil ranges from 330-mg/kg to 1010-mg/kg oil according to the Codex
Standard. Sesame oil from black sesame seeds contains less tocopherols than oils
from brown or white sesame seeds (Table 7). The wild species of sesame, Sesamum
angustifolium E. and Sesamum radiatum S. and T., have higher levels of total
tocopherol (760 mg/kg and 810 mg/kg, respectively) in the oil than the cultivated
species (486680 mg/kg) although they have a black seed coat. Regardless of the
species and the color of seed coat, g-tocopherol is the predominant tocopherol
in sesame oil, whereas d-tocopherol accounted for less than 5% of the total toco-
pherols. a-Tocopherol is present in sesame oil in trace amount only. Among the
different tocopherol isomers, g-tocopherol is a more potent antioxidant in oils
(33), but it has lower Vitamin E value in biological systems than a-tocopherol (34).
TABLE 7. Levels of Tocopherols in Sesame Oil.
Tocopherol (mg/kg Oil)

Sesame Color of
Species Seed Coat a g d Total Reference
Sesamum indicum L.
Japanese strainsa Black 5.2 468.5 12.2 485.9 26
Brown 6.2 517.9 13.6 537.7 26
White 3.8 497.8 20.5 522.1 26
Sudan strainsb Black NDd 527.0 12.6 540 27
Brown 4.8 663.7 11.6 680 27
White 3.1 603.9 13.0 620 27
Sesamum alatum T.c Brown 2.9 310.1 7.0 320 27
Sesamum radiatum S. and T.c Black 6.5 800.3 3.2 810 27
Sesamum angustifolium E.c Black ND 754.7 5.3 760 27
Codex standard ND3.3 521983 421 3301010 24
a
The cultivated species of sesame grown in Japan.
b
c
d
ND: Not detected.
SESAME LIGNANS AND LIGNAN GLYCOSIDES 549
4.5. Protein
The protein content of sesame seed is approximately 25% with a range of 1731%
depending on the source of the seed. Sesame protein is low in lysine (3.1% protein),
but it is rich in sulfur-containing amino acids methionine and cystine (6.1%), which
are often the limiting amino acids in legumes. Comparing with the standard
values recommended by FAO and WHO for children, sesame protein is borderline
deficient in other essential amino acids such as valine, threonine, and isoleucine.
Sesame seed protein, however, contains an adequate amount of tryptophanm, which
is limiting in many oilseed proteins. Because of its characteristic amino acid
composition, sesame seed protein is regarded as an excellent protein source for
supplementing many vegetable proteins such as soybean and peanut to increase
their nutritional value.
The protein efficiency ratio (PER) of sesame seed protein is 1.86 (35). The PER
value can be raised to 2.9 when sesame seed protein is supplemented with lysine
(36). El-Adawy (37) added sesame products including sesame meal, sesame protein
isolate, and protein concentrate to red wheat flour to produce flour blends. It was
found that water absorption, development time, and dough weakening were increas-
ed as the protein level increased in all blends; however, dough stability decreased.
Sesame products could be added to wheat flour up to 16% protein without any detri-
mental effect on bread sensory properties. The addition of sesame products to red
wheat flour increased the contents of protein, minerals, and total essential amino
acids; the in vitro protein digestibility also increased significantly.
Inyans and Nwadimkpa (17) investigated the protein functionality of dehulled
sesame seed flour. They reported that the emulsification capacity was higher at
alkaline condition and ranged from 25-ml oil/g at pH 4 to 66-ml oil/g at pH 10.
The highest foaming capacity (315%) was observed at pH 2. Protein solubility
ranged from 7.9% at pH 2 to 14.2% at pH 10. The viscosity of the flour dispersion
ranged from 2.5 cps at 1% concentration to 7.0 cps at 10% concentration. The se-
same flour could impart desirable characteristics when incorporated into products
such as ice cream, frozen dessert, sausage, baked food, and confectionary.
When sesame seeds were boiled or allowed to sprout, in order to reduce bitter
taste, there was a slight increase in protein content of sprouted seeds and the foam-
ing capacity of flour from boiled seeds was increased (38). The emulsion stability
was improved after sprouting or boiling, whereas the emulsion capacity was low-
ered after boiling. The bitter taste was not detected in flour from boiled seed but
still persisted in that from sprouted seed.
5. SESAME LIGNANS AND LIGNAN GLYCOSIDES
5.1. Lignans
Sesame oil contains high levels of unsaturated fatty acids (more than 80% of total
fatty acids); however, it is highly resistant to oxidative deterioration as compared
with other edible vegetable oils (39, 40). The superior oxidative stability is not
550 SESAME OIL
only attributed to the presence of tocopherols, but it is mainly associated with the
unique group of compounds-lignans (41). Lignans are compounds formed by oxi-
dative coupling of r-hydroxyphenylpropane. They are widely distributed in all parts
of plants. Oilseeds such as sesame and flaxseed are well known to contain abundant
lignans (42). Two types of lignan compounds existed in sesame seeds, the oil solu-
ble lignans and the water soluble lignan glycosides. In raw sesame seed, sesamin
and sesamolin are the two major lignans. Sesamin has been found in other plants,
whereas sesamolin is characteristic of sesame and has not been found in plants
other than Sesamum. Fukuda et al. (43) determined the lignan contents of 14 vari-
eties of commercial sesame seeds grown in Japan and noticed that sesamin content
was always higher than sesamolin content and that the average ratio of sesamolin to
sesamin in the black varieties (0.61.0) was greater than the white varieties
(0.20.5). Other types of lignans such as sesamol, P1, sesamolinol, and sesaminol
were only present in minor quantity as shown in Table 8 (43). The structures of the
sesame lignans are illustrated in Figure 7.
Tashiro et al. (21) further investigated the oil and lignan (sesamin and sesamolin)
contents in 42 strains of Sesamum indicum L. originated from different parts of the
world. The strains included white-, brown-, black-, and yellow-colored seed types.
The results of this study indicated that the sesamin content in the oil ranging from
0.07% to 0.61% with an average of 0.36%, whereas the sesamolin content was low-
er (ranging from 0.02% to 0.48% with an average of 0.27%). There was a signifi-
cant positive correlation between the oil content of the seed and the sesamin content
of the oil, whereas no correlation existed between the oil and the sesamolin
TABLE 8. Lignan Contents in Different Strains of Sesame Seeds.a,b

Strain Color of Sesamolin/
no. Seed Coat Sesamin Sesamolin Sesamin Sesamol P1 Sesamolinol Sesaminol
48 White 821.3 441.2 0.537 2.0 1.6 1.0 1.4

611 White 410.6 441.2 0.537 2.5 1.3 1.0 1.0
630 White 522.7 123.5 0.236 2.5 2.3 0.9 0.3
638 White 885.2 476.5 0.538 NDc 2.9 1.1 1.0
643 White 464.0 229.4 0.494 5.0 2.0 1.1 1.0
785 Yellow 453.3 247.0 0.545 Trace 2.0 0.9 0.3
673 Violet 464.0 317.6 0.684 2.5 1.8 1.5 1.1
675 Brown 528.0 264.6 0.501 Trace 3.8 0.6 0.7
126 Brown 682.7 458.8 0.672 4.0 2.9 1.2 1.0
201 Black 502.5 441.2 0.878 3.6 2.5 1.2 1.1
601 Black 314.3 235.3 0.749 10.8 1.6 1.9 1.1
631 Black 362.7 229.4 0.632 2.5 1.5 0.8 0.5
792 Black 154.7 152.9 0.988 4.9 1.5 0.9 0.9
801 Black 293.3 294.0 1.002 6.5 1.6 1.1 1.2
Mean 490.6 300.4 3.4 2.1 1.1 0.9

SD 198.6 113.6 2.9 0.7 0.3 0.3
a
Data adapted from (43).
b
Unit: mg/100-g oil.
c
ND: Not detected.
O
O O
O O
O
O
O
O
O
O
O
O
O O
O O O
O
Sesamin Episesamin Sesamolin
O OCH3
O H3CO
O O
H3CO
OCH3 OH
OCH3
O O
O O O
O O O
Sesangolin 2-episesalatin Sesamol
O
O
OH OH
O
OH
O O O
O O
O
Sesaminol O Sesamol dimer
(direct-linked)
Figure 7. Structures of lignans.
contents. It was also noticed that the black seed types contained significantly less
oil and a high ratio of sesamolin to sesamin. In the wild species of sesame seeds,
Fukuda et al. (43) found that an Indian variety had an extreme low sesamolin con-
tent (only 14% of its sesamin content), whereas one variety from Borneo contained
552 SESAME OIL
OH O
OH OH O O
CH2
O
O O O O
O O O O O
O
Sesamol dimer Sesamol dimer
(Methylene-bridged) quinone Samin
O
O OCH3 O
OH O
O
O O
O O
O O
OH OH OH
OCH3 OCH3 OCH3
Sesamolinol Pinoresinol P-1

Figure 7. (Continued)
several times more sesamin (1152.3 mg/100 g oil) and sesamolin (1360.7 mg/100 g
oil) than in other species. Kamal-Eldin and Appelqvist (27) determined the contents
of sesamin and sesamolin in three wild species of Sesamum. They reported that S.
radiatum was extremely high in sesamin (2.40% in oil) but contained only a minor
amount of sesamolin (0.02%), whereas S. alatum contained minor amounts of sesa-
min and sesamolin (both were 0.01%); S. angustifolium possessed reasonable
amounts of both sesamin (0.32%) and sesamolin (0.16%).
Other types of lignans were found in wild species of Sesamum. Sesangolin was
present in S. angolense (44) and was the major lignan in S. angustifolium, which
contained 3.15% sesangolin in its oil (27). 2-Episesalatin occurred in S. alatum
(45) and was its most abundant lignan present at 1.37% in its oil (27). The struc-
tures of sesangolin and 2-episesalatin are shown in Figure 7. The contents of
different lignans present in sesame oil are listed in Table 9.
5.2. Lignan Glycosides

Lignan glycosides are the glycosilated forms of lignans; they are water soluble.
Although most lignans are found in the oil-soluble part of sesame seed, lignan
glycosides are present in sesame meal. Sesaminol, sesamolinol, and pinoresinol
TABLE 9. Levels of Lignans in Sesame Oil.

Lignan Contents (% Oil)
Color of Reference
Sesame Species Seed Coat Sesamin Sesamolin Sesamol Sesangolin 2-Epsesalatin
Sesamum indicum L.
Eleven strains Black 0.24 0.27 21
(0.070.40) (0.130.40)
Twelve strains Brown 0.36 0.30 21
(0.110.61) (0.130.42)
Fifteen strains White 0.44 0.25 21
(0.120.61) (0.020.48)
Japanese strains Black 0.45 0.54 NDa 26
Japanese strains Brown 0.46 0.66 ND 26
Japanese strains White 0.66 0.42 ND 26
Sudan strains Black 0.45 0.54 ND ND 27
Sudan strains Brown 0.46 0.66 ND ND 27
Sudan strains White 0.230.72 0.390.41 ND ND 27
Sesamum alatum T.b Brown 0.01 0.01 ND 1.37 27
Sesamum radiatum
S. and T.b Black 2.4 0.02 ND ND 27
Sesamum angustifolium E.b Black 0.32 0.16 3.15 ND 27
a
ND: Not detected.
b
glucosides (Figure 8) are the major lignan glycosides in sesame. Acetone extract of
sesame seed contained sesamolinol and sesaminol (46, 47), and it was revealed that
they were released after treating defatted sesame seed flour with b-glucosidase (48).
Later, three pinoresinol diglucosides (KP1, KP2, and KP4) and one pinoresinol tri-
glucoside (KP3) were isolated from the ethanol extract of sesame seed (49, 50).
Kuriyama et al. (51) analyzed the lignan glycosides composition of white sesame
seed with high-performance liquid chromatography (HPLC) and found eight lignan
glycosides. There were two pinoresinol glucosides with two or three glucose
units, three sesaminol glucosides with one to three glucose units, two sesamolinol
glucosides with one or two glucose units, and one P-1 glucoside with two glucose
units. The total contents of lignan glycosides in white sesame seed were around
100170-mg/100-g seed, with sesaminol triglucoside the most predominant one.
In black sesame seed, the lignan glycosides content varied greatly with the
species of the sesame (from 6.4 to 361.3-mg/100-g seed), whereas sesaminol tri-
glucoside was still the major lignan glycoside (52). This effect of sesame variety
on the lignan glycoside contents was also noticed by Ryu et al. (53). They reported
that a significant difference existed between the black and white sesame seeds in
their sesaminol contents, which were analyzed after hydrolysis of the sesaminol
glucosides. White sesame seeds contained an average of 84.5-mg sesaminol in
100-g seed (ranging from 32.5 to 98.5 mg/100 g), and black sesame seeds contained
113.2 mg/100 g of sesaminol in average with a range of 41.5 to 134.5-mg/
100-g seed. Table 10 lists the contents of sesaminol glucosides in various sesame
seeds.
554 SESAME OIL
O O
O O
O O
OR
O O
O
R = Glc O R = Glc
= Glc-Glc O = Glc-Glc OR
= Glc-Glc-Glc OCH3
Sesaminol glucosides Sesamolinol glucosides
OCH3
R1O
O
(16)
KP1 : R1 = H, R2 = Glc Glc
(12)
KP2 : R1 = H, R2 = Glc Glc OR2
KP3 : R1 = H, R2 = Glc-Glc-Glc OCH3
KP4 : R1 = Glc, R2 = Glc
Pinoresinol glucosides
Figure 8. Structures of lignan glycosides.
TABLE 10. Contents of Sesaminol Glucosides in Different

Sesame Seeds.a
Sesaminol Glucosides (mg/100g Seed)

Color of Seed Coat Meanb Range CV (%)c
Black n 10d 113.2 41.5134.5 23.5

Brown n 5 78.5 39.491.4 8.9
White n 10 84.5 32.598.5 11.8
a
b
Mean values bearing different superscripts are different significantly at 1% level.
c
CV: coefficient of variance.
d
n: represents the number of samples analyzed.
PROCESSING 555
6. PROCESSING
Sesame oil has a long history of human consumption. The processing of sesame
seed to yield sesame oil varies from region to region. The major differences
are (1) whether the seed coat is removed and (2) whether the seed is roasted.
Figure 9 shows the flow diagrams of the processing of three major types of sesame
oils produced worldwide, namely (1) refined sesame oil, which is produced from
unroasted sesame seed either with seed coat or without seed coat; (2) roasted
sesame oil, which is produced from roasted sesame seed; and (3) small mill sesame
oil, which is produced from roasted dehulled sesame seed.
Refined sesame oil is the salad oil grade of sesame oil. It is the most common
type of sesame oil consumed worldwide except in the Orient. Sesame seeds are
cleaned and cooked before oil extraction with expeller. Crude sesame oil is refined
by alkali-refining, bleaching, and deodorization to obtain the refined sesame oil
(Figure 9). Sesame cake from oil extraction with expeller may still contain
1822% of residual oil (54). It is often extracted with solvent or pressed again
to obtain more oil. The desolventized sesame cake can then be processed into
food grade sesame flour if the dehulled sesame seed is used. If the seed coats are
not removed, the sesame cake can only be used as feedstuff because it contains
undesirable constituents. The dehulling process will be discussed later.
Roasted sesame oil has a strong characteristic flavor of roasted sesame seed. It is
the most popular sesame oil consumed in China, Japan, and Korea. It is also
believed to be beneficial to health (40). As shown in Figure 9, sesame seeds are
roasted at 140200 C prior to oil extraction. The conditions of the roasting process
Sesame seed
Cleaning
Dehulling Cooking Soaking Roasting
Oil extraction Roasting Grinding

Oil extraction
with expeller
with expeller
Cooling by
Cooking
spraying water
Sesame Crude Crude Sesame Oil extraction
Dehulling
cake Sesame oil sesame oil cake by expeller
Solvent extraction Milling
Solvent extraction Roasted sesame
Alkali-refining
crude oil
Sesame paste
Sesame flour Bleaching Sesame meal
(Food grade) (Feed grade)
Filtration
Stirring in hot water
Deodorization
Separation of oil by gravitation
Roasted sesame
or centrifugation
Refined sesame oil
oil
Small mill
sesame oil
Figure 9. Flow diagram showing the processing of different sesame oils.

556 SESAME OIL
is of prime importance to the quality of the roasted sesame oil. The effect of roast-
ing on sesame seed and oil will be discussed later. After roasting, sesame seeds are
ground, cooked, and pressed to obtain the crude roasted sesame oil. The crude oil is
simply filtered without further purification to produce roasted sesame oil. The color
of roasted sesame oil ranges from light yellow to dark brown depending on the
roasting conditions.
Small mill sesame oil, also known as Shiang-you, is a unique sesame oil product
of Northern China. It has a light roasted sesame flavor and is light brownish in col-
or. Shiang-you is often used as seasoning oil for cold dishes; it is seldom used for
cooking purpose. Roasted sesame oil, however, is mainly used as cooking oil. The
processing scheme of small mill sesame oil is shown in Figure 9. Sesame seeds are
cleaned and soaked in water for about an hour in order for the sesame seed to reach
a water content of 35%, which can facilitate protein denaturation, assure even heat-
ing, and avoid burning during the subsequent roasting process. Roasting process is
recommended to conduct at 200 C for 30 min. The roasted sesame seeds are cooled
to 140150 C by spraying water. Before milling the roasted sesame seeds with
stone mill, the seed coats are removed by blowing air or sieving through screen.
The milling process is important for the separation of oil from sesame paste.
Successful milling will result in sesame paste with fine particle size, which will
give rise to a higher oil yield (55). After milling, hot water is added to the sesame
paste and stirred slowly (around 30 rpm). The addition of hot water (temperature
above 90 C) is usually conducted three to four times with decreasing amount of
added water. Sesame oil will slowly rise to the top by gravitational force when
the addition of water is completed and the paste is allowed to stand for 1 hour
(56). The processing of Shiang-you is labor-intensive, and the oil yield (around
40%) is low. Many efforts were made to increase the yield of Shiang-you. Yen
and Tsai (57) tried to include soybean oil in hot water to separate oil from sesame
paste. They reported that the highest yield of Shiang-you was obtained with the
combination ratio of sesame paste-soybean oil-boiling water (10 : 9 : 7, w/w).
6.1. Dehulling
Generally, sesame seeds are processed without removal of seed coat. Seed coat con-
tains undesirable oxalic acid and indigestible fiber that may lower the nutritional
value of the meal. The presence of seed coat will also impart a dark color and bitter
taste to the meal. In India, where sesame meal is an important food, dehulling is an
indispensable step of sesame oil processing. Sesame meal prepared with dehulled
sesame seed is non-bitter, light-colored, low in fiber, and rich in protein. Dehulling
is performed either manually at village level or mechanically in conventional oil
mills in India (58, 59). Manual dehulling involved soaking sesame seeds in water
and removal of hulls from the swell and burst seeds by light pounding or rubbing on
stone or wooden block. It is tedious, labor intensive, and inefficient; therefore, it
limits the production of sesame oil and its meal.
Mechanical dehulling can be processed either by soaking sesame seeds in water
followed by removal of seed coat mechanically (58) or by alkali treatment (6062).
PROCESSING 557
In alkali treatment or lye peeling method, sesame seeds are treated with hot lye for
a short time. Either hot 0.6% NaOH for 1 minute (60) or 6% NaOH at 60 C for
10 seconds (61) have been used to decorticate sesame seed. There was no appreci-
able loss in protein and oil contents after alkali treatment. Nag et al. (63) reported
that dehulling not only increased oil content but also produced oil of better color
quality compared with the whole seed. Sesame oil extracted from dehulled sesame
seeds, however, was oxidatively less stable (measured by the Rancimat test) than
that extracted from whole seeds (64). The presence of natural antioxidants such
as g-tocopherol, sesamin, and sesamolin in the seed coat may contribute to the oxi-
dative stability of whole sesame seed oil. In addition, Chang et al. (65) recently
reported that the sesame seed coat also contained phenolic compounds and tetra-
nortriterpenoids, which had good antioxidative activity. Dehulled sesame seeds
are not suitable for roasting process either. Abou-Gharbia et al. (66) demonstrated
that sesame oil prepared from coated seeds had better oxidative stability than from
dehulled seeds either roasted at 200 C for 20 min or without roasting as evaluated
by peroxide value, conjugated diene formation, and TBA value.
6.2. Roasting
In Oriental countries such as China, Japan, and Korea, sesame seeds are often
roasted prior to oil extraction. Roasting is important for the development of desir-
able color and flavor for sesame oil, and it will enhance the oxidative stability of
sesame oil (67). The conditions of roasting may influence the sensory quality and
composition of the roasted sesame oil. When sesame seeds were roasted between
180 C and 260 C for 30 min, Yen (68) reported that the red color unit of the roasted
sesame oil increased with temperature up to 220 C and then decreased while the
flavor score showed an optimum at 200 C. There was almost no change in fatty
acid composition until the roasting temperature was above 220 C (68, 69). The
antioxidant, sesamol, content also increased with roasting temperature up to
220 C and then decreased with higher roasting temperature. The roasting tempera-
ture of 200 C was therefore recommended (68, 70).
Yoshida and Takagi (69) compared the quality of sesame oils prepared at
roasting temperatures between 160 C and 250 C. They found that the typical
dark-brown color was apparent after 15 min, and the roasted sesame seeds had a
burnt smell when the roasting temperature was above 220 C. Roasted sesame oil
obtained from seeds roasted at temperature above 220 C had burnt and bitter
tastes; the peroxide, anisidine, carbonyl, and TBARS values were also higher indi-
cating poor oil quality. They suggested that a high-quality roasted sesame oil would
be obtained by roasting for 25 min at 160 C and 180 C, 15 min at 200 C, and 5 min
at 220 C.
6.2.1. Effect of Roasting on Antioxidative Activity Roasted sesame oil was

reported to be much more antioxidative than unroasted purified sesame oil (71).
Yen and Shyu (67) found that roasted sesame oils prepared from sesame seeds
with different roasting temperatures (between 180 C and 210 C) exhibited
558 SESAME OIL
differences in their oxidative stability. The oxidative stability appeared to increase

with roasting temperature; sesame oil prepared from 200 C roasted seed was found
to exhibit the best stability.
Koizumi et al. (72) also noticed the relationship between the roasting conditions
of sesame seeds and the development of antioxidative activity. The antioxidative
activity of oil obtained from sesame seed roasted at 200 C for as short as 5 min
was higher than at 180 C for 30 min. This observation indicated that the develop-
ment of antioxidant activity in roasted sesame oil depends primarily on tempera-
ture. In an attempt to investigate the contributing antioxidants in roasted sesame
oil, Fukuda et al. (73) reported that either sesamol alone or g-tocopherol alone at
the concentrations present in roasted sesame oil showed weak antioxidant activity.
Even the combination of both was not enough to explain the strong antioxidative
activity of roasted sesame oil. As roasting of sesame seed caused significant brown-
ing (74), the browning products from roasted sesame seed were isolated and found
to show weak antioxidative activity (73). The combination of g-tocopherol, sesame
lignans (sesamol and sesamin), and the browning products was shown to be respon-
sible for the superior oxidative stability of roasted sesame oil (73).
6.2.2. Effect of Roasting on Different Classes of Lipids Roasting of sesame

seed not only affects the antioxidative activity and the lignans of sesame oil,
the lipid composition will also be affected. The lipids in sesame seeds consist of
neutral lipids, phospholipids, and glycolipids. The major lipid fraction is neutral
lipids, which constitute about 9196% of the total lipids. Phospholipids and
glycolipids represent around 3% and 0.36% of the total lipids, respectively (69,
70, 7578).
Roasting will cause a significant reduction in the phospholipids content in
sesame seed because of browning reaction (75). Phospholipids fraction in sesame
seeds decreased with roasting temperature and time (69, 70, 7577). There was
no appreciable change in phospholipids content when sesame seeds were roasted
at 160180 C for 10 min (69); the reduction in phospholipeds content was
6973% at 220 C for 25 min (70, 77) and 96% at 250 C for 25 min (69). Even
with microwave roasting, phospholipids in sesame seed decreased appreciably;
more than half of the original phospholipids were lost after microwaving at
2450 MHz for 15 min (76), and less than 14% were left after 30 min (78). The
highest rate of phospholipid loss was observed in the phosphatidyl ethanolamine
(PE) fraction followed by the phosphatidyl choline (PC) and phosphatidyl inositol
(PI) fractions. This trend became more pronounced with longer roasting time and
higher roasting temperature. After roasting at 220 C for 25 min, PE was completely
destroyed while there were still 22% of PC and 42% of PI left in the roasted sesame
seeds (77). The amino group of PE or PC was suggested to be involved in browning
reaction and donation of hydrogen or electron to tocopherol or sesamol (79).
Glycolipids content of sesame seed, on the other hand, increased with roasting
temperature and time (75, 78). When sesame seeds were roasted in an electric oven
from 120 C to 250 C for 30 min, the glycolipids content was found to increase
from 6.9-mg/1000-g seeds (0.5% of total lipids) to 262.9-mg/1000-g seeds (17.2%
PROCESSING 559
of total lipids) as reported by Yoshida (75). Glycolipid components of microwaved

sesame seeds increased slowly in the first 25 min of heating and rapidly thereafter; a
ninefold increase in glycolipids content after 30 min of microwaving at 2450 MHz
was observed (78). The color of sesame seeds become brownish after heating, indi-
cating that browning reaction has taken place. The browning substances are gener-
ally very polar, and the increase in browning substances may be attributed to the
increase of glycolipids (78).
The dominant component of sesame lipids, neutral lipids, did not change in its
content when sesame seeds were roasted at temperature below 200 C for 30 min.
As the roasting temperature increased to 220 C and 250 C, a significant decrease in
the neutral lipids content was noticed (75). This decrease became more severe when
the roasting time was increased (69, 70, 77). Yoshida et al. (70) examined the effect
of roasting on the molecular species of triacylglycerols. They reported that roasting
for 10 min at 220 C caused a significant decrease not only in molecular species
containing more than four double bonds, but also in the amount of diene and
triene species present in triacylylycerols. They also confirmed that no significant
changes in molecular species or fatty acid distribution of triacylglycerols would
occur within 25 min of roasting at 180 C.
6.3. Extraction of Oil

The tradition way of extracting sesame oil from unroasted sesame seeds in India is
done by ghani, which is basically a large pestle and mortar. The ghani is driven by
bullocks (79). Sesame seeds are cleaned and dehulled before used in the ghani. In
many parts of India, water or brown sugar is added to sesame seed in the ghani to
facilitate oil extraction (80). Sesame oil is removed from the ghani after milling
and allowed to settle, skimmed, and sometimes strained through a cloth before
sale. The bullock-driven ghani is replaced by power-driven mills in most of the
Indian villages in order to improve the efficiency of oil production (58, 80).
The modern commercial methods of oil extraction from oilseeds include (1)
batch hydraulic pressing: Oil seeds are expressed by hydraulic pressure to yield
oil; (2) continuous mechanical pressing: Oil seeds are squeezed through a taper-
ing outlet and oil is expressed by the increasing pressure; and (3) solvent
extraction: Oil seeds are extracted with solvent followed by removal of solvent
to yield oil. These methods are also employed in the extraction of sesame seeds
with some modification.
For unroasted sesame seeds, the commercial extraction of oil is carried out using
a continuous screw-press or hydraulic press. Sesame seeds are small; they are
usually cooked prior to oil extraction. Sesame oil is generally extracted in three
stages (60, 79). The first stage is cold press; the cold-pressed oil obtained after fil-
tration is ready to use and has very good quality. It is light in color and agreeable in
taste and odor. The second stage pressing is conducted with sesame residue under
high pressure; it yields a highly colored oil that needs refining before used for
edible purpose. The residue left after the second stage pressing is extracted for
the third time under similar conditions as the second stage. Sesame oil obtained
560 SESAME OIL
from the third stage pressing has very low quality and is used for nonedible pur-
poses.
Alternatively, unroasted sesame seeds are pressed once followed by solvent
extraction to recover the oil from residue. The oxidative stability of sesame oil
was found to be dependent on the extraction method and seed pretreatment (64).
Extraction of the sesame seeds after effective seed crushing with polar solvent,
heptane-isopropanol (3 : 1, v/v), would yield a more stable oil from whole sesame
seeds because more antioxidative substances and phospholipids could be extracted.
Phospholipids may act as synergists to antioxidants (81).
The extraction of sesame oil from roasted sesame seed is generally performed
with pressing. Solvent extraction is not used because the desirable roasted flavor
may be removed during evaporation of solvent. In commercial production, contin-
uous screw-press or hydraulic press is employed (42). The hydraulic press can be
vertical or horizontal. The continuous screw may be operated twice in order to
increase the oil yield (82). Proper cooking (100 C, 7 min) and addition of water
(12.5%) after roasting can also raise the oil yield (83).
6.4. Refining
Sesame oil from roasted sesame seed has the characteristic flavor and color of the
roasted sesame oil; the filtered crude oil is used without further refining. Sesame oil
from cold-pressed unroasted sesame seed is also used directly after filtration as a
flavored oil. Crude sesame oil from unroasted sesame seeds after screw-press or
hydraulic press or solvent extraction, which varies in color from yellow to dark
amber, may need further refining. Refined sesame oil is usually pale yellow in color.
Crude sesame oil does not require extensive purification and refining. The
suspended meal particles in crude oil can be removed by settling, screening, and
filtering. The filtered crude oil can be used directly or be further refined to remove
impurities such as phospholipids, resins, free fatty acids, and coloring substances.
The refining steps include removal of free fatty acids, gums, and some water-
miscible substances by alkaline treatment, removal of pigments by bleaching,
and removal of odorous substances by deodorization. Degumming is not necessary
because sesame oil contains a limited amount (<3%) of phospholipids (69).
Alkali-refining of sesame oil can use sodium carbonate as the neutralizing
agent in order to reduce the refining loss, because sodium carbonate does not attack
the neutral triacylglycerols. The free fatty acids are first neutralized by sodium
carbonate, and then a weak sodium hydroxide (NaOH) wash is given to improve
color. Liberation of carbon dioxide, which makes the separation of soapstock
difficult, has limited the practice of using sodium carbonate in alkali-refining.
Mukhopadhyay et al. (84) reported an easy way of refining sesame oil with alkali-
enriched dry sodium metasilicate (SMS). This method precluded emulsion forma-
tion, and thus the separation of soapstock could be easily achieved. It is superior to
the sodium carbonate process because no liberation of carbon dioxide is involved.
The reduction in free fatty acids by this dry refining process was dependent on the
PROCESSING 561
alkalinity of SMS-NaOH mixture. Color of the oil was not markedly improved by
this process. The process appears to be useful for the refining of crude oils with low
free fatty acids and medium color. Therefore, it is specially useful for the alkali-
refining of sesame oil.
After alkali-refining, the neutralized sesame oil is bleached with a relatively
lower quantity of bleaching earth as compared with that required for most other
vegetable oils. Bleaching conditions and the bleaching agent employed may
influence the bleaching efficiency. Increase in bleaching temperature was found
to increase bleaching efficiency until a maximum was reached and then decreased
(85). Agitation speed also affect the result of bleaching; 40 rpm (86) or 50100 rpm
(85) was found to be the optimal condition. The higher the ratio of adsorbent/edible
oil, the higher is the bleaching ability of adsorbent (87). Recently, activated rice
hull ash was investigated as the bleaching agent of sesame oil (88, 89). Rice hull
ashed at 500 C for 30 min followed by activation with 6N H2SO4 at 30 C for 60
min was found to possess the maximum bleaching efficiency (88). Using this acid-
activated rice hull ash as bleaching agent, sesame oil could be successfully bleached
at 120 C with an agitation speed of 80 rpm employing 25 mg of rice hull ash per
gram of sesame oil (89).
Bleaching removes most pigments, and the bleached oil is light in color. In order
to produce a bland oil suitable for salad dressing, the bleached sesame oil is further
deodorized. Deodorization is conducted in vacuum with steam at 200250 C as
most other vegetable oils.
6.5. Changes of Lignans During Processing

The two major lignans, sesamin and sesamolin, present in sesame seed are reported
to be responsible for many unique chemical and physiological properties of sesame
seed oil (39). Sesamin and sesamolin, however, do not have antioxidative activity in
themselves because of a lack of phenolic groups (90). The high oxidative stability
of sesame oil came mainly from the transformation products of these sesame lig-
nans. Sesamol, which possesses antioxidative activity and is usually present in trace
amount in the oil of raw sesame seed, may be released from sesamolin during the
roasting process of sesame seed prior to pressing for oil (46). Sesamol could also be
formed from the hydrolysis of sesamolin after heating at frying temperature for 12
hours (47). During sesame oil refining, the antioxidative sesaminol was formed in
high concentration from sesamolin under the acidic anhydrous condition of bleach-
ing (acid clay are used for bleaching) as reported by Fukuda et al. (41).
Sesamolin is first decomposed to sesamol by protonolysis to form an oxonium
ion, and then the carbon-carbon bond is formed; thus, it is hypothesized that sesa-
minol is formed from sesamolin by intermolecular group transformation (91). The
conversion of sesamolin to sesamol and sesaminol is illustrated in Figure 10. Both
sesamol and sesaminol are strong antioxidants; they contribute to the superior
oxidative stability of refined or roasted sesame oil.
Sesamol is unstable to heat and is completely destroyed when the roasted sesame
oil is heated at the deep frying temperature of 180 C for 4 hours. Sesaminol,
562 SESAME OIL
O
O O O
O O
O
O O
Protonolysis
OH OH
O
O
H O O
O Oxonium ion
O O
O O O
Sesamolin Sesamol Sesaminol
Figure 10. Conversion of sesamolin to sesamol and sesaminol.
however, is more heat stable. The retention of sesaminol in the roasted sesame oil
after heating at 180 C for 6 hours was 40.5% (52).
The other sesame lignan, sesamin, will undergo epimerization upon heating
with acid (92). A marked change in the sesamin content of sesame oil was observed
after bleaching and deodorization. The decrease in sesamin was accompanied by
the formation of epi-sesamin. The deodorization process of oil refining will also
destroy the thermally liable lignan sesamol. Sesamol produced from sesamolin dur-
ing the bleaching step was lost in the next deodorization step (93). Only trace
amounts (<20 ppm) of sesamol were found in the commercially deodorized sesame
oil (41).
The changes in the contents of sesame lignans during industrial refining of
unroasted sesame oil, which included alkaline treatment, warm water washing,
bleaching with acid clay, and deodorization, are listed in Table 11. These data
clearly revealed that the significant changes in the sesame lignans contents
occurred at the bleaching step. There were epimerization of sesamin (41%),
disappearance of sesamolin, and formation of sesamol, sesaminol, epi-sesaminol,
and a minor amount of sesamol dimer. It was also evident that the contents of
TABLE 11. Contents of Sesame Lignans and Tocopherol in Unroasted Sesame Oil
During Industrial Refining Process (mg/100-g Oil).a
Sesamol Sesaminol
Refining Stage Sesamin Epi-Sesamin Sesamolin (Sesamol Dimer) (Epi-Sesaminol) g-Tocopherol
Crude sesame oil 813.3 0 510.0 4.3 0 33.5

(0) (0)
Alkaline-refining 730.6 0 458.0 2.5 0 23.4
(0) (0)
Warm water 677.8 0 424.8 0.7 0 22.6
washing (0) (0)
Bleaching 375.5 277.6 0 46.3 33.9 21.8
(trace) (48.0)
Deodorizing 258.3 192.6 0 1.7 28.4 18.4
(trace) (34.4)
a
TABLE 12. Effect of Processing Method on the Retention of Sesamin and Sesamolin in Sesame Oil (mg/100-g Oil).a
Sesamin Sesamolin

Coated Seed Dehulled Seed Coated Seed Dehulled Seed

Processing Method Fresh Storedb Fresh Storedb Fresh Storedb Fresh Storedb
Raw seed 649 20d,x 584 15d,y 610 21d,x 461 16d,y 183 7d,x 123 6d,y 168 5d,x 117 5d,y
Roasting 576 14e,x 436 10g,y 489 15f,x 315 14f,y 146 5e,x 73 2f,y 119 3f,x 55 1g,y
Steaming 601 18e,x 514 14e,y 531 16e,f,x 325 13f,y 129 5f,x 88 4e,y 108 4g,x 52 1g,y
Roasting plus steaming 583 15e,x 506 13e,f,y 555 18e,x 411 15e,y 146 6e,x 115 7d,y 139 4e,x 106 3e,y
Microwaving 590 17e,x 475 12f,y 520 12e,f,x 422 16e,y 123 3f,x 71 3f,y 129 2e,f,x 75 2f,y
a
b
The extracted oil was stored at 65 C for 35 days.
c
Results are mean values of three determinations SD. Values in each column with different superscripts (d-g ) are significantly p < 0:05 different from one another. Values
of fresh and stored oil with different superscripts (x and y ) are significantly p < 0:05 different from each other.
564 SESAME OIL
Roasted Sesame Oil Sesame Seed Refined Sesame Oil
roasting deodorizaiton
Epi-sesamin Sesamin Epi-sesamin
roasting heating
Sesamol Sesamolin Sesamol deodorizaiton Sesamol dimer
(heat unstable)
[o] bleaching Sesaminol
Sesamol (heat stable)
dimer
decomposition Samin and sesamol
[o]
Sesamol
dimmer
quinone
Figure 11. Changes of sesame lignan during processing.
sesaminol and its epimer did not decrease by deodorization as much as sesamol. In
refined unroasted sesame oil, sesaminol, epi-sesaminol, and g-tocopherol are thus the
antioxidative substances responsible for its excellent oxidative stability (41).
Shahidi et al. (94) investigated the effect of different processing methods, in-
cluding roasting (200 C for 20 min), steaming (100 C for 20 min), roasting
(200 C for 15 min) plus steaming (100 C for 7 min), and microwaving
(2450 MHz for 15 min) on the endogenous antioxidants in the resultant sesame
oil and upon storage. Sesamin content in oil was well retained (nearly 90%) in
oil from coated seed immediately after processing, but the decrease was more
pronounced (nearly 50%) in oil from dehulled seed especially after the oil was
stored (65 C for 35 days). The roasting process resulted in the highest loss of
sesamin. The corresponding changes in sesamolin contents were more drastic
than sesamin (Table 12).
The changes of sesame lignans during processing is summarized in Figure 11.
Recently, Asakura et al. (95) have prepared the ortho methylene-bridged and
direct-link oligomers from sesamol. The structures are shown in Figure 7. The
methylene-bridged oligomers showed much stronger antioxidant activities on
the autoxidation of lard than the sesamol monomer because of a greater average
number of hydroxyl groups per sesamol unit. The direct-linked oligomers prepared
in acidic conditions were better antioxidants for lard than the sesamol monomer,
whereas oligomers prepared under neutral and alkaline conditions did not improve
the antioxidant effect of sesamol.
7. NUTRITIONAL CHARACTERISTICS
7.1. Effect on Polyunsaturated Fatty Acid Metabolism

Linoleic acid and a-linolenic acid are essential fatty acids and are the important
fatty acids involved in the metabolic pathway of prostaglandin synthesis.
NUTRITIONAL CHARACTERISTICS 565
Converting linoleic acid to g-linolenic acid and dihomo-g-linolenic acid (DGLA) is

catalyzed by 6 -desaturase, whereas 5 -desaturase catalyzes the transformation of
DGLA to arachidonic acid. Shimizu et al. (96) reported that sesame oil could cause
an accumulation of DGLA acid in the cell. Sesamin was discovered to be the active
component in sesame oil; it can inhibit the activity of 5 -desatursase (97). When
rats were fed sesamin, there was an accumulation of DGLA in liver phospholipids
and the ratio of DGLA to arachidonic acid increased. Arachidonic acid is the
precursor of eicosanoids such as 2 series prostaglandin and 4 series leukotriene.
Consequently, sesamin tended to reduce the production of eicosanoids from
arachidonic acid (98), and the plasma concentration of PGE2 was decreased (99).
Fujiyama-Fujiwara et al. (100) also reported that sesame lignans (sesamin and epi-
sesamin) inhibited 5 desaturation from DGLA n-6 to arachidonic acid n-6, but
not from 20 : 4n-3 to eicosapentaenoic acid (EPA, n-3) in cultured rat hepatocytes,
and Umeda-Sawada et al. (101) confirmed this finding in vivo and also found that
dietary sesame lignans decreased arachidonic acid content and increased n-6/n-3
ratio. Umeda-Sawada et al. (102) further examined the effect of dietary sesame lig-
nans on hepatic metabolism and n-6/n-3 ratio of essential fatty acids in rats; they
concluded that sesame lignans could inhibit extreme changes of n-6/n-3 ratio and
function to bring it close to the appropriate n-6/n-3 ratio. Epidemiological and clin-
ical studies have shown that the plasma n-6/n-3 ratio is associated with the preva-
lence of thrombosis (99, 103). Therefore, sesame lignans would be beneficial to the
prevention of thrombosis.
7.2. Hypocholesterolemic Effect of Sesame Lignans

Sesame oil was reported to lower the absorption of fatty acid and cholesterol in
lymph by 50% when rats were fed diet containing 24% sesame oil as compared
with control diet containing no sesame oil (104). As the lymphatic system is the
major route for the transport of absorbed fatty acids and cholesterol, serum and liver
cholesterol levels were significantly reduced, especially LDL-cholesterol (105).
Crude lignan fraction separated from sesame oil was found to have a weak but signi-
ficant hypocholesterolemic activity (98). The cholesterol-lowering activity depen-
ded on the dietary level of the lignans. With purified sesame lignan (sesamin),
the hypocholesterolemic effect was clearly demonstrated (106). As shown in
Table 13, sesamin (0.5%) significantly reduced the serum cholesterol in rats fed
a cholesterol-enriched diet (Exp. I) or a commercial chow diet (Exp. II). Sesamin
lowered intestinal absorption of cholesterol by precipitating cholesterol from the
bile acid micelles, and thus the serum cholesterol level is reduced. Table 13 also
shows that liver cholesterol concentration was also significantly lowered when
rats were fed a sesamin-containing diet because of the reduction in the activity
of liver microsomal 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-
CoA reductase), the key enzyme in the cholesterol synthesis pathway of liver. Sesa-
min thus possess a unique function in that it can simultaneously inhibit cholesterol
synthesis and absorption. It is, therefore, a potential hypocholesterolemic agent of
natural origin.
566 SESAME OIL
TABLE 13. Effect of Sesamin on the Concentrations of Serum Cholesterol, Liver

Cholesterol, and the Liver Enzyme Activity.1,2
Serum Cholesterol Liver Cholesterol Liver HMG-CoA Reductase

Diet (mg/dl) (mg/g liver) (pmol/min/mg protein)
Exp. I
Purified diet 108 4a 2.54 0.13a 203 12a
Diet sesamin (0.5%) 110 5a 1.95 0.06b 151 11b
Diet cholesterol (0.5%) 136 8b 20.8 2.2c 51.6 2.0c
Diet cholesterol (0.5%) 102 5a 9.13 1.02d 29.0 2.4d
and sesamin (0.5%)
Exp. II
Commercial chow 69.1 5.2a 2.86 0.19a 269 27a
Chow sesamin(0.5%) 55.5 3.0b 1.82 0.04b 172 13b
1
2
Values are means SEM (n 6 8). Values with different letters are significantly different in each
experiment p < 0:05. Male Wistar rats were fed experimental diets for 4 weeks.
The hypocholesterolemic effect of sesamin could be enhanced by a-tocopherol

(107). Data shown in Table 14 clearly indicated that rats fed sesamin together with
tocopherol (1%), the serum cholesterol-lowering effect of sesamin, could be
demonstrated at a much lower level (0.05%). This synergistic effect was found to
be related to both the levels of sesamin and cholesterol in the diet. The combination
of a-tocopherol with sesamin has a practical value for the treatment of hyper-
cholesterolemia. The cholesterol-lowering effect of sesamin has also been demon-
strated in humans with dietary supplementation of sesamin at 64.8-mg/day
level (108).
With regard to the mechanism underlying the hypocholesterolemic effect of
dietary sesamin, Hirose et al. (109) demonstrated in rats that it increased fecal
TABLE 14. Combined Effects of Sesamin and a-Tocopherol

on Serum Cholesterol Levels of Rats.1,2
Group Serum Cholesterol (mg/dl)
Cholesterol diet 490 94a

Diet 1.0% tocopherol 460 70a
Diet 0.05% sesamin 437 76a
Diet 0.05% sesamin 1.0% tocopherol 244 23b,c
Diet 0.2% sesamin 371 28a,c
Diet 0.2% sesamin 0.2% tocopherol 243 5c
Diet 0.2% sesamin 1.0% tocopherol 149 9c
1
2
Values are means SE (n 6 9). Values with different letters are signi-
ficantly different in each experiment p < 0:05. Male Wistar rats were fed
experimental diets for 4 weeks.
cholesterol excretion and reduced the hepatic activity of HMG-CoA reductase. In

addition, Ashakumary et al. (110) examined the effect of sesame lignan (a 1 : 1
mixture of sesamin and episesamin) on hepatic fatty acid oxidation in rats. They
concluded that sesame lignan greatly increased the activity and gene expression
of hepatic fatty acid oxidation enzymes and thus increased the rate of fatty acid
b-oxidation through the activation of peroxisome proliferator activated receptor
(PPAR)a. Sesame lignan was also demonstrated to decrease the hepatic fatty acid
synthesis in rats by decreasing the activity and gene expression of many hepatic
enzymes involved in fatty acid synthesis (111) because sesame lignan contains
both sesamin and episesamin. The effect of each component was examined by
Kushiro et al. (112). They found that episesamin caused a larger magnitude of
increase in the activity and gene expression of enzymes in fatty acid oxidation
than sesamin. Sesamin and episesamin showed no difference, however, in lowering
the activity and gene expression of hepatic lipogenic enzymes.
7.3. Effect on Vitamin E

Sesame seed has long been regarded as a health food for longevity. Namiki et al.
(113115) examined the effect of sesame seed in aging by using a senescence-
accelerated mouse, and they have found that the advancement of senescence was
suppressed by long-term feeding of sesame seed. Vitamin E is recognized as a
food component that may exert an anti-aging effect (116). Sesame seed, however,
contains mainly g-tocopherol whose Vitamin E activity is only 616% that of
a-tocopherol (117, 118), although it exhibits a stronger antioxidative activity in
vitro than a-tocopherol (119, 120). The effects of sesame seed and sesame lignans
on Vitamin E activity were, therefore, studied extensively to elucidate if sesame is a
good source of Vitamin E.
Yamashita et al. (121) first reported that sesame seed and its lignans could raise
the bioactivity of g-tocopherol to almost the same level as a-tocopherol in rats.
Later, they reported that sesame seed lignans could also act synergistically with
a-tocopherol to enhance its Vitamin E activity in rats fed a low a-tocopherol diet
(122). Kamal-Eldin et al. (123) showed that feeding rats with sesamin, a lignan
from sesame oil, increased g-tocopherol and g-/a-tocopherol ratio in the plasma,
liver, and lung. Sesamin appears to enhance the bioavailabity of g-tocopherol in
rat plasma and tissues, and this effect persists in the presence of a-tocopherol. Diet-
ary sesame seed can also elevate the tocotrienol concentration in the adipose tissue
and skin of rats fed tocotrienol-rich diet (124). The effect of sesame lignans on the
levels of tocopherols was also demonstrated in humans. In a study with 40 healthy
Swedish women (mean age 26), serum g-tocopherol concentrations were raised sig-
nificantly after consuming a diet that contained 22.5 g/day of sesame oil (125).
Coonery et al. (126) gave muffins containing equivalent amounts of g-tocopherol
from sesame seeds, walnuts, or soy oil to nine volunteers; they observed that con-
sumption of as little as 5 mg of g-tocopherol per day over a 3-day period from
sesame seeds but not from walnuts nor soy oil significantly elevated serum g-toco-
pherol levels in the volunteers.
568 SESAME OIL
7.4. Effect on Blood Pressure

Sesamin, the most abundant lignan present in sesame seed and sesame oil, was
demonstrated to suppress the development of hypertension in rats induced by
deoxycorticosterone acetate (DOCA) and salt (127). Dietary sesamin was also
reported to effectively prevent the elevation of blood pressure and cardiac hypertro-
phy in two-kidney, one-clip (2k, 1c) renal hypertensive rats (128). In the stroke-
prone spontaneously hypertensive rats (SHRSP), sesamin feeding was much
more effective as an anti-hypertensive regimen in salt-loaded SHRSP (with 1%
salt in drinking water) than in unloaded SHRSP (129).
7.5. Antioxidative Effect in Biological System

In the development of atherosclerosis, oxidative modification of low-density lipo-
protein (LDL) is the critical step and is therefore a target for interventions aimed at
slowing down the progression of atherogenesis (130). Antioxidants such as Vitamin E,
probucol, and N,N0 -diphenylphenylenediamine (DPPD) were suggested to prevent
the oxidation of LDL (131133). Sesame oil is highly resistant to oxidative dete-
rioration because of the presence of endogenous antioxidants such as sesaminol,
sesamolinol, pinoresinol, and P1. Sesaminol exerted a strong inhibitory effect on
the 2,20 -azobis (2,4-dimethylvaleronitrile) (AMVN)-induced peroxidation of LDL
by acting as a chain breaker in the lipid peroxidation cascade in vitro (134). In inhi-
biting either Cu2-induced or 2,20 -azobis (2-amidinopropane) dihydrochloride
(AADH)-induced lipid peroxidation in LDL, sesaminol was found to be more effec-
tive than a-tocopherol and probucol. Sesaminol was also the strongest antioxidant
among the sesame lignans (sesamolinol, pinoresinol, and P1) for protecting LDL
from oxidative modification (135). The reason for the strong antioxidative effect
of sesaminol is possibly because of its highly lipophilic nature that makes it act
within the LDL particle to exert a sparing effect on tocopherol (122, 123).
The in vivo antioxidative activity of sesame lignan was examined in an animal
model (136). When SD rats were fed a diet containing 1% sesamolin, the lipid per-
oxidation activity (measured as 2-thiobarbituric acid reactive substances, TBARS)
in the liver and kidney was significantly lowered. The amount of 8-hydroxy-20 -
deoxyguanosine, a DNA base-modified product generated by reactive oxygen spe-
cies and a good marker for oxidative damage (137), was also significantly lower in
the sesamolin-fed rats. Sesamolin is one of the major sesame lignans present in the
oil fraction of sesame; however, it does not possess any appreciable in vitro anti-
oxidant activity (138). The significant in vivo antioxidative activity of sesamolin
came from its metabolites, sesamol and sesamolinol, when sesamolin was supple-
mented in rats diet (136). Feeding rats with a diet containing 40% of dietary energy
as either sesame, soybean, olive, or canola oils for 7 weeks, sesame oil was shown
to be the most effective one in lowering lipid peroxidation (139). Sesame seeds rich
in sesame lignans, sesamin and sesamolin, could lower the activities of enzymes
involved in fatty acid synthesis, and thus the serum triacylglycerol levels were
lower in rats fed diets high in sesame lignans (140).
Sesame seeds contain two types of lignans, the oil-soluble lignans such as sesa-
min and sesamolin and the water-soluble lignan glycosides including pinoresinol
glucosides (141) and sesaminol glucosides (142). Both of the glucosides were lower
in peroxyl radical scavenging activity than their corresponding aglycones because
of the lack of phenolic group. Using hypercholesterolimic rabbit as the animal
model, Kang et al. (143) were able to demonstrate that dietary defatted sesame flour
(containing 1% sesaminol glucoside ) could decrease the peroxidation in liver and
serum. Sesaminol, the principal metabolite of sesaminol glucoside and the active
antioxidant, was found in abundant quantities in the serum and liver of rabbit
(143). In an insulin-resistance animal model, rats were fed with high fructose
diet in order to develop insulin-resistance, which was accompanied by a high oxi-
dative stress status (144). When the insulin-resistant rats were given 1.0 g/kgBW of
crude lignan glycosides, liver TBARS were significantly lowered and the insulin
sensitivity was improved, indicating an alleviation of oxidative stress (145).
7.6. Effect on Cancer

Antioxidants are well recognized to play an important role in the defense against
oxidative stress, which may cause damage to membrane, nucleic acid, and protein
resulting in circulatory ailments, senility, mutation, and cancer (146). As sesame
lignans possess antioxidative ability, their effect on the model systems for in vivo
peroxidation, such as the peroxidation of ghost membranes of rabbit erythrocyte
and the peroxidation of rat liver microsome, were investigated (147). Sesame
lignans were found to suppress lipid peroxidation equal to or stronger than toco-
pherol in these systems. One of the sesame lignan, sesaminol, was observed to
be as strongly suppressive as tocopherol in mutagenicity of E. Coli WP2s induced
by peroxidation of membrane lipid of erythrocytes (147).
As mentioned earlier that sesame lignans, especially sesamin and epi-sesamin,
could influence the metabolism of polyunsaturated fatty acid and the production of
prostaglandins. As prostaglandin is one of the most influential factors for mammary
carcinogenesis, Hirose et al. (99) studied the effect of sesamin on dimethylbenz-
anthracene (DMBA)-induced mammary cancer. Their results showed that sesamin
at a dietary level of 0.2% considerably reduced the cumulative number and
mean number of mammary cancer; the effectiveness of sesamin was similar to a-
tocopherol.
The anti-tumor promotion activity of topically and orally admistered sesame
components was tested in ICR mice using a two-stage skin tumorigenesis model
(148). Skin tumor was initiated with 7,12-dimethylbenz [a]-anthracene (DMBA)
and promoted with 12-o-tetra-decanoylphorbol-13-acetate (TPA). The sesame com-
ponents applied topically after TPA treatment were able to delay the formation of
papilloma remarkably. It was suggested that sesame components had radical
scavenging ability toward the reactive oxygen species or peroxidized molecules
generated by TPA. Therefore, the inhibition of tumorigenesis by sesame compo-
nents was the result of metabolic inactivation. When sesame components were
admistered orally, the formation of skin papilloma was also inhibited effectively,
570 SESAME OIL
indicating that the sesame components could be absorbed and remained active even
after passing through digestive organs (149).
Sesamin, however, did not significantly reduce the number of N-nitrosobis-
(2-oxopropyl)-amine(DOP)-induced pancreatic cancer in hamsters (150). It was
noticed that 2% sesamol in the diet exerts forestomach carcinogenic activity in
rats and mice (151). Fortunately, human beings do not have a forestomach and daily
ingestion of sesamol is much lower than 2%.
7.7. Effect on Liver Function

Sesamin fed to rats at a level above 0.5% caused a temporary liver enlargement
because of an increase in liver phospholipids; no specific histological changes
were observed, and the activities of serum GOT and GPT remained unchanged
(99, 106). It was suggested that sesamin could act as a stimulus to the liver function,
particularly in the endoplasmic reticula. When mice were exposed to a high concen-
tration of carbon tetrachloride or continuously inhaled ethanol to cause liver
damage, sesamin was able to improve the liver function (152). Furthermore, rats
previously given sesamin were found to reduce their plasma ethanol levels more
rapidly than the control rats. This effect of sesamin on alcohol metabolism was
studied in human trials. Male adults given sesamin (100 mg/day for 7 days) were
found to have a significantly faster rate of ethanol reduction in their blood (153).
The effect of dietary sesamin and sesaminol on the ethanol-induced modulation
of immune indices related to food allergy has also been studied. Although chronic
ethanol drinking would increase the plasma IgA, IgM, and IgG concentrations,
0.2% sesamin in the diet could suppress this increase of IgA and IgM, whereas
sesaminol was not effective. In addition, the increase in relative liver weight
because of ethanol consumption was alleviated by dietary supplementation of
sesamin but not by sesaminol (154).
REFERENCE
1. D. Bedigian and J. R. Harlan, Econ Bot., 40, 137 (1986).

2. N. M. Nayar, in N. W. Simmonds, ed., Sesame: Evolution of Crop Plants, Longman,
London and New York, 1976, pp. 231233.
3. T. Kobayashi, in A. Ashri, ed., Sesame: Status and Improvement Proceedings of Expert
Consultation, The Type Classification of Cultivated Sesame Based on Genetic Char-
acters, FAO Plant Production and Protection Paper 29 Rome, 1981, pp. 8689.
4. N. M. Nayar and K. L. Mehra. Econ. Bot., 24, 20 (1970).
5. A. Ashri, in G. Roebblen, R. K. Downey, and A. Ashri, eds., Oil Crops of the Word,
McGraw Hill, New York, 1989, p. 375.
6. J. J. Ochse, J. Soule, M. J. Dijkman, and C. Wehlburg, Tropical and Subtropical
Agriculture Vol. II, Macmillan, New York, 1961, pp. 10891093.
REFERENCE 571
7. J. W. Purseglove, Tropical Crops: Dicotyledons, Longman, Harlow, Essex, 1982, pp.

430435.
8. D. G. Langham, J. Hered., 36, 245 (1945).
11. T. W. Culp, J. Hered., 51, 146 (1960).
12. D. R. Langham and T. Wiemers, in J. Janick and A. Whipkey, eds., Trends in New Crops
and New Used, ASHS Press, Alexandria, VA, 2002, p. 157.
13. Food and Agriculture Organization of the United Nations (FAO). (2003). Available:
http://faostat.fao.org/faostat/collections?subsetagriculture.
14. D. H. Siraj, Palm Oil Link, 12, 15 (2002).
15. D. A. Dashak and C. N. Fali, Food Chem., 47, 253 (1993).
16. G.-C. Yen, S.-L. Shyu, and J.-S. Lin, J. Agric. Forest. (Taiwan), 35, 177 (1986).
17. U. E. Inyang and C. U. Nwadimkpa, J. Amer. Oil Chem. Soc., 69, 819 (1992).
18. J. E. Kinsella and R. R. Mohite, in A. M. Altsechull and H. L. Wilcke, eds., New Protein
Foods, Academic Press, New York, 1985, 435456.
19. L. A. Johnson, T. M. Suleiman, and E. W. Lusas, J. Amer. Oil Chem. Soc., 56, 463
(1979).
20. A. Kamal-Eldin and L.-A. Appelqvist, J. Amer. Oil Chem. Soc., 71, 135 (1994).
21. T. Tashiro, Y. Fukuda, T. Osawa, and M. Namiki, J. Amer. Oil Chem. Soc., 67, 508
(1990).
22. H. Baydar, I. Turgut, and K. Turgut, Tr. J. Agric. Forest., 23, 431 (1999).
23. H. Baydar and I. Turgut, Tr. J. Agric. Forest., 18, 387 (1994).
24. Codex Alimentarius, Named Vegetable Oils 8, Codex Standard 210, (2001).
25. R. T. OConnor and S. F. Herb, J. Amer. Oil Chem. Soc., 47, 186A (1970).
26. H. Yoshida, J. Shigezaki, S. Takagi, and G. Kajimoto, J. Sci. Food Agric., 68, 407
(1995).
28. R. J. Sims, J. A. Fioriti, and M. J. Kanuk, J. Amer. Oil Chem. Soc., 49, 298 (1972).
29. T. A. Miettinen, T. Puska, H. Glyying, H. Vanhanen, and E. Vartianen, N. Engl. J. Med.,
333, 1308 (1995).
30. A. B. Awad, Y.-C. Chen, C. S. Fink, and T. Hennessey, Anticancer Res., 16, 2797 (1996).
31. R. L. Von Holtz, C. S. Fink, and A. B. Awad, Nutr. Cancer, 32, 8 (1998).
32. A. Downie, C. S. Fink, and A. B. Awad, FASEB, 13, A433 (1999).
33. C. H. Lea and R. J. Ward, J. Sci. Food Agric., 10, 537 (1959).
34. G. W. Burton and M. G. Traber, Ann. Rev. Nutr., 10, 375 (1990).
35. M. S. Narasinga Rao in H. V. Srivastava, S. Bhaskaran, B. Vatsya, and K. K. G. Menon,
eds., Oilseeds Production: Constraints and Opportunities, Oxford and IBH, New Delhi,
India, 1985, pp. 625634.
36. M.C.S. Sastry, N. Subramanian, and H.A.B. Parpia, J. Amer. Oil Chem. Soc., 51, 115
(1974).
37. T. A. El-Adawy, Food Chem., 59, 7 (1997).
38. G. I. O. Badifu and E. M. Akpagher, Plant Foods for Human Nutrition, 49, 119 (1996).
572 SESAME OIL
39. P. Budowski and K. S. Markely, Chem. Rev., 48, 125 (1951).

40. M. Namiki, Food Rev. Int., 11, 281 (1995).
41. Y. Fukuda, M. Nagata, T. Osawa, and M. Namiki, J. Amer. Oil Chem. Soc., 63, 1027
(1986).
42. P. K. J. P. D. Wanasundara, F. Shahidi, and V. K. S. Shukla, Food Rev. Int., 13, 225
(1997).
43. Y. Fukuda, T. Osawa, S. Kawakishi, and M. Namiki, Nippon Shokuhin Kogyo Gakkaishi,
35, 483 (1988).
44. W. A. Jones and M. Beroza, J. Org. Chem., 27, 3232 (1962).
45. A. Kamal-Eldin and G. Yousif, Phytochemistry, 31, 2911 (1992).
46. T. Osawa, M. Nagata, M. Namiki, and Y. Fukuda, Agric. Biol. Chem., 49, 3351 (1985).
47. M. Nagata, T. Osawa, M. Namiki, and Y. Fukuda, Agric. Biol. Chem., 51, 1285 (1987).
48. Y. Fukuda, T. Osawa, M. Namiki, and T. Ozaki, Agric. Biol. Chem., 49, 301 (1985).
49. H. Katsuzaki, M. Kawasumi, S. Kawakishi, and T. Osawa, Biosci. Biotech. Biochem.,
56, 2087 (1992).
50. H. Katsuzaki, S. Kawakishi, and T. Osawa, Phytochemistry, 35, 773 (1994).
51. K.-I. Kuriyama, K.-Y. Tsuchiya, and T. Murui, Nippon Nogeikagaku Kaishi, 67, 1693
(1993).
52. T. Osawa, H. Kumon, M. Namiki, S. Kawakishi, and Y. Fukuda, in M. W. Pariza, H. U.
Aeschbacher, J. S. Felton, and S. Sato, eds., Mutagens and Carcinogens in the Diet,
Wiley-Liss, New York, 1990, p. 233.
53. S. N. Ryu, C.-T. Ho, and T. Osawa, J. Food Lipids, 5, 17 (1998).
54. G.-C. Yen, S.-L. Shyu and T.-C. Lin, Food Sci. (Chinese), 13, 198 (1986)
55. G.-C. Yen, and S.-L. Shyu, J. Chinese Agric. Chem. Soc., 26, 50 (1988).
56. A. J. Shi, Food Res. Develop. (China), 22, 43 (2001).
57. G.-C. Yen, and Y.-E. Tsai, Food Sci., 17, 87 (1990).
58. Oil Technological Research Institute, Profiles in Oil Technology, OTRI, Anantpur, India,
1977, p. 157.
59. B. D. Green, in E. W. Lusas, D. R. Erickson, and W. Nip, eds., Food Uses of Whole Oil
and Protein Seeds, Amer. Oil Chem. Soc., Champaign, Illinois, 1989, p. 218.
60. B. S. Ramachandra, M. C. S. Sastry, and L. S. Subbarao, J. Food Sci. Technol. (Mysore),
7, 127 (1970).
61. Y. G. Moharram, A. M. Moustafa, and H.O.A. Osman, Lebensm. Wiss. Technol., 14, 137
(1981).
62. G. M. Yehai, O. A. Hussein, and I. A. Y. Abou-El-Khier, Food Nut. Bull, 121, 1 (1985).
63. D. K. Nag, D. Pandit, M. Sen, D. K. Bhattacharya, M. M. Chakraborty, and
S. Mukherjee, Sci. Culture, 65, 163 (1999).
65. L.-W. Chang, W.-J. Yen, S. C. Huang, and P.-D. Duh, Food Chem., 78, 347 (2002).
66. H. A. Abou-Gharbia, F. Shahidi, A. A. Y. Shehata, and M. M. Youssef, J. Amer. Oil
Chem. Soc., 74, 215 (1997).
67. G.-C. Yen and S.-L. Shyu, Food Chem., 31, 215 (1989).
68. G.-C. Yen, J. Sci. Food Agric., 50, 563 (1990).
REFERENCE 573
69. H. Yoshida and S. Takagi, J. Sci. Food Agric., 75, 19 (1997).

70. H. Yoshida, Y. Hirakawa, and S. Takagi, J. Sci. Food Agric., 80, 1495 (2000).
71. Y. Fukuda, M. Nagata, T. Osawa, and M. Namiki, Agric. Biol. Chem., 50, 857 (1986).
72. Y. Koizumi, Y. Fukuda, and M. Namiki, Nippon Shokuhin Kagaku Kogaku Kaishi, 43,
689 (1996).
73. Y. Fukuda, Y. Koizumi, R. Ito, and M. Namiki, Nippon Shokuhin Kagaku Kogaku
Kaishi, 43, 1272 (1996).
74. G.-C. Yen and S.-L. Shyu, Food Sci., 19, 25 (1992).
75. H. Yoshida, J. Sci. Food Agric., 65, 331 (1994).
76. H. A. Abou-Gharbia, A. A. Y. Shehata, and F. Shahidi, Food Res. Int., 33, 331 (2000).
77. H. Yoshida, S. Abe, Y. Hirakawa, and S. Takagi, J. Sci. Food Agric., 81, 620 (2001).
78. H. Yoshida and G. Kajimoto, J. Food Sci., 59, 613 (1994).
79. D. K. Salunkhe, J. K. Chavan, R. N. Adsule, and S. S. Kadam, World Oilseeds:
Chemistry, Technology and Utilization, Van Nostrand Reinhold, New York, pp. 371402
(1992).
80. H. G. Muralidhara, in A Panorama of the World of Oils, National Education Society,
Shimoga, India, 1981, pp. 5657.
81. P. Brandt, E. Hollstein, and C. Franzke, Lebensm. Ind., 20, 31 (1973).
82. G.-P. Dong, China Oil Fat, 18, 55 (1993).
83. S. Y. Tsai and Y. H. Chu, Report of FIRDI (Taiwan), 21 (1991).
84. S. B. Mukhopadhyay, P. K. Gupta, and A. K. Basu, J. Amer. Oil Chem. Soc., 69, 294
(1992).
85. L. L. Richardson, J. Amer. Oil Chem. Soc., 55, 777 (1978).
86. H. B. W. Patterson, J. Amer. Oil Chem. Soc., 53, 339 (1976).
87. K. Y. Liew, A. H. Yee, and M. R. Nordin, J. Amer. Oil Chem. Soc., 70, 539 (1993).
88. Y.-Y. Chang and C.-I. Lin, J. Chem. Eng. Jpn., 34, 1 (2001).
89. S.-P. Lin, C.-I. Lin, and H.-K. Chen, J. Chem. Eng. Jpn, 34, 737 (2001).
90. P. Budowski, F. G. T. Menezenes, and F. G. Doller, J. Amer. Oil Chem. Soc., 27, 377
(1950).
91. Y. Fukuda, M. Isobe, M. Nagata, T. Osawa, and M. Namiki, Heterocycles, 24, 923
(1986).
92. M. Beroza, J. Amer. Oil Chem. Soc., 78, 5082 (1956).
93. P. Budowski, J. Amer. Oil Chem. Soc., 41, 280 (1964).
94. F. Shahidi, R. Amarowicz, H. A. Abou-Gharbia, and A. Adel Y. Shehata, J. Amer. Oil
Chem. Soc., 74, 143 (1997).
95. K. Asakura, E. Honda, S. Matsumura, S. Osanai, and K. Kawada, J. Oleo. Sci., 51, 97
(2002).
96. S. Shimizu, K. Akimoto, H. Kawashima, Y. Shinmen, and H. Yamada, J. Amer. Oil
Chem. Soc., 66, 237 (1989).
97. S. Shimizu, K. Akimoto, Y. Shinmen. H. Kawashima, M. Sugano, and H. Yamada,
Lipids, 29, 512 (1991).
98. M. Sugano, T. Inoue, K. Koba, K. Yoshida, N. Hirose, Y. Shinmen, K. Akimoto, and
T. Amachi, Agric. Biol. Chem., 54, 2669 (1990).
574 SESAME OIL
99. N. Hirose, F. Doi, T. Ueki, K. Akazawa, K. Chijiiwa, M. Sugano, K. Akimoto,

S. Shimizu, and H. Yamada, Anticancer Res., 12, 1259 (1992).
100. Y. Fujiyama-Fujiwara, R. Umeda-Sawada, and O. Igarashi, J. Nutr. Sci. Vitaminol., 38,
353 (1992).
101. R. Umeda-Sawada, N. Takahashi, and O. Igarashi, Biosci. Biotech. Biochem., 59, 2268
(1995).
102. R. Umeda-Sawada, M. Ogawa, and O. Igarashi, Lipids, 33, 567 (1998).
103. H. O. Bang, Lancet, 1, 1143 (1971).
104. S. Satchithanandam, M. Reicks, R. J. Calvert, M. M. Cassidy, and D. Kritchevsky,
J. Nutr., 123, 1852 (1993).
105. S. Satchithanandam, R. Chanderbhan, A. T. Kharroubi, R. J. Calvert, D. Klurfeld,
S. A. Tepper, and D. Kritchevsky, Int. J. Vit. Nutr. Res., 66, 386 (1995).
106. N. Hirose, T. Inoue, K. Nishihara, M. Sugano, K. Akimotok, S. Shimizu, and H. Yamada,
J. Lipid Res., 32, 629 (1992).
107. A. Nakabayashi, Y. Kitagawa, Y. Suwa, K. Akimoto, S. Asami, S. Shimizu, N. Hirose,
M. Sugano, and H. Yamada, Int. J. Vit. Nutr. Res., 65, 162 (1995).
108. F. Hirata, K. Fujita, Y. Ishikura, K. Hosoda, T. Ishikawa, and H. Nakamura, Athero-
sclerosis, 122, 135 (1996).
109. N. Hirose, T. Inoue, K. Nishihara, M. Sugano, K. Akimoto, S. Shimizu, and H. Yamada,
J. Lipid Res., 32, 629 (1991).
110. L. Ashakumary, I. Rouyer, Y. Takahashi, T. Ide, N. Fukuda, T. Aoyama, T. Hashimoto,
M. Mizugaki, and M. Sugano, Metabolism, 48, 1303 (1999).
111. T. Ide, L. Ashakumary, Y. Takahashi, M. Kushiro, N. Fukuda, and M. Sugano,
Biochimica Biophysica Acta, 1534, 1 (2001).
112. M. Kushiro, T. Masaoka, S. Hageshita, Y. Takahashi, T. Ide, and M. Sugano, J. Nutr.
Biochem., 13, 289 (2002).
113. M. Namiki, K. Yamashita, and T. Osawa, in K. Yagi, ed., Active Oxygens, Lipid
Peroxides, and Antioxidants, Japan Scientific Society Press, Tokyo/CRC Press, Boca
Raton, Florida, 1993, p. 319.
114. K. Yamashita, Y. Kawagoe, Y. Nohara, M. Namiki, T. Osawa, and S. Kawagishi, J. Jpn.
Soc. Nutr. Food Sci., 43, 445 (1990).
115. K. Yamashita and M. Namiki, in T. Takeda, ed., The SAM Model of Senescence,
Excerpta Medica Amsterdam, London-New York-Tokyo, 1994, p. 153.
116. B. Loo, R. Labugger, C. P. Aebischer, J. N. Skepper, M. Bachschmid, V. Spizer,
J. Kilo, L. Altwegg, V. Ullrich, and T. F. Luscher, Circulation, 105, 1635 (2002).
117. I. R. Peake and J. G. Bieri, J. Nutr., 101, 1615 (1971).
118. J. G. Bieri and R. P. Evarts, J. Nutr., 104, 850 (1974).
119. G. G. Duthie, B. M. Gonzalez, P. C. Morrice, and J. R. Arthur, Free Radic. Res.
Commun., 15, 35 (1991).
120. T. Saldeen, D. Li, and J. L. Mehta, J. Amer. Coll. Cardiol., 34, 1208 (1999).
121. K. Yamashita, Y. Nohara, K. Katayama, and M. Namiki, J. Nutr., 122, 2440 (1992).
122. K. Yamashita, Y. Iizuka, T. Imai, and M. Namiki, Lipids, 30, 1019 (1995).
123. A. Kamal-Eldin, D. Pettersson, and L.-A. Appelqvist, Lipids, 30, 499 (1995).
124. S. Ikeda, K. Toyoshima, and K. Yamashita, J. Nutr., 131, 2892 (2001).
REFERENCE 575
125. M. Lemcke-Norojarvi, A. Kamal-Eldin, L.- A. Appelqvist, L. H. Dimberg, M. Ohrvall,

and B. Vessby, J. Nutr., 131, 1195 (2001).
126. R. V. Cooney, L. J. Custer, L. Okinaka, and A. A. Franke, Nutr. Cancer, 39, 66 (2001).
127. Y. Matsumura, S. Kita, S. Morimoto, K. Akimoto, M. Furuya, N. Oka, and T. Tanaka,
Biol. Pharm. Bull., 18, 1016 (1995).
128. S. Kita, Y. Matsumura, S. Morimoto, K. Akimoto, M. Furuya, N. Oka, and T. Tanaka,
Biol. Pharm. Bull., 18, 1283 (1995).
129. Y. Matsumura, S. Kita, Y. Tanida, Y. Taguchi, S. Morimoto, K. Akimoto, and T. Tanaka,
Biol. Pharm. Bull., 21, 469 (1998).
130. D. Steinberg, J. Biol. Chem., 272, 20,963 (1997).
131. R. M. Dieber, H. Puhl, G. Waeg, G. Striegl, and H. Esterbauer, J. Lipid Res., 32, 1325
(1991).
132. J. Regnstrom, G. Walldius, L. A. Carlson, and J. Nilsson, Atherosclerosis, 82, 43 (1990).
133. C. P. Sparrow, T. W. Doebber, J. Olszewski, M. S. Wu, J. Ventre, K. A. Stevens, and
Y. S. Chao, J. Clin. Invest., 89, 1885 (1992).
134. M.-H. Kang, H. Katsuzaki, and T. Osawa, Lipid, 33, 1031 (1998).
135. M.-H. Kang, M. Naito, K. Sakai, K. Uchida, and T. Osawa, Life Sci., 66, 161 (2000).
136. M.-H. Kang, M. Naito, N. Tsujihara, and T. Osawa, J. Nutr., 128, 1018 (1998).
137. B. Halliwell and O. I. Aruoma, DNA and Free Radicals, Ellis Horwood, Chichester,
U.K., 1993.
138. P. Budowski, J. Amer. Oil Chem. Soc., 27, 264 (1950).
139. N. H. Baba, Z. Ghossoub, and Z. Habbal, Nutr. Res., 20, 1113 (2000).
140. S. Sirato-Yasumoto, M. Katsuta, Y. Okuyama, Y. Takahashi, and T. Ide, J. Agric. Food
Chem., 49, 2647 (2001).
141. H. Katsuzaki, M. Kawasumi, S. Kawakishi, and T. Osawa, Biosci. Biotechnol. Biochem.,
56, 2087 (1992).
142. H. Katsuzaki, S. Kawakishi, and T. Osawa, Phytochemistry, 35, 773 (1994).
143. M.-H. Kang, Y. Kawai, M. Naito, and T. Osawa, J. Nutr., 129, 1885 (1999).
144. F. S. Facchini, M. H. Humphreys, C. A. DoNascimento, F. Abbasi, and G. M. Reaven,
Am. J. Clin. Nutr., 72, 776 (2000).
145. S.-F. Kang, M. S. Thesis, Grad. Inst. of Food Sci. and Tech., National Taiwan University,
Taipei, Taiwan, 2001, p. 72.
146. B. Halliwell, R. Aeschbach, J. Loliger, and O. I. Aruoma, Food Chem. Toxicol., 33, 601
(1995).
147. T. Osawa, M. Namiki, and S. Kawakishi, in Y. Kuroda, D. M. Shankel, and M. D.
Waters, eds., Antimutagenesis and Anticarcinogenesis Mechanisms II, Plenum,
New York, 1990, p. 139.
148. A. Mimura, K. Takebayashi, M. Niwano, H. Tokuda, H. Nishino, A. Iwashima, and
Y. Takahara, Proc. Jpn. Cancer Assoc., 57 (1991).
149. A. Mimura, K. Takebayashi, M. Niwano, Y. Takahara, T. Osawa, and H. Tokuda, in C.-T.
Ho, T. Osawa, M.-T. Huang, and R.-T. Rosen, eds., Food Phytochemicals in Cancer
Prevention II, American Chemical Society, Series 547, American Chemical Society,
Washington, D.C., 1994, p. 281.
150. T. Ogawa, T. Makino, N. Hirose, and M. Sugano, Carcinogenesis, 13, 1663 (1994).
576 SESAME OIL
151. S. Tamano, M. Hirose, H. Tanaka, E. Asakawa, K. Ogawa, and N. Ito, Jpn. J. Cancer
Res., 83, 1279 (1992).
152. K. Akimoto, Y. Kitagawa, T. Akamatsu, N. Hirose, M. Sugano, S. Shimizu, and
H. Yamada, Ann. Nutr. Metab., 37, 218 (1993).
153. M. Nakamura, H. Nagai, M. Nakagawa, and M. Sugano, 45th Annual Meeting of the
Japanese Society of Nutrition and Food, 1991, p. 168.
154. M. Nonaka, K. Yamashita, Y. Iizuka, M. Namiki, and M. Sugano, Biosci. Biotech.
Biochem., 61, 836 (1997).
13
Soybean Oil
Earl G. Hammond, Lawrence A. Johnson, Caiping Su,
Tong Wang, and Pamela J. White
Iowa State University
Ames, Iowa
1. INTRODUCTION
The amounts of soybeans and total vegetable oil crops have been rising for a
number of years. World production of soybeans in 2003 was estimated to be
184.49 million MT out of 317.89 million MT total for vegetable oil crops, making
soybeans the worlds largest oilseed crop, rivaled only by palm oil (1). The 2003
crop of soybeans was expected to yield 29.85 million MT of soybean oil out of a
total of 91.79 million MT of vegetable oil worldwide. The U.S. production of
soybean oil was estimated at 8.59 million MT for 2002, of which 7.86 million
MT was consumed domestically. During 20022003, Brazil produced 4.90 million
MT and Argentina 4.12 million MT of soybean oil (2). The U.S. price of crude
soybean oil has varied from $0.24/kg to $0.62/kg over the past 5 years with the
lower prices being more recent (1).
Soybeans owe their dominance of the oilseed market to the value of their protein,
which is much greater than that of other oilseeds. Of the oilseed meals produced in
2003, 129.58 million MT out of a total of 185.69 milllion MT was soybean meal
(1). Of the money made on extracting soybeans, the meal accounted for between
51% and 76% of the total in the last 10 years. Soybean oil of typical composition
performs well as a salad oil, but it is usually hydrogenated for use as a margarine
stock or frying oil. Soybean oils stability to oxidation also is limited by its content
577
578 SOYBEAN OIL
of linolenic acid. Recent decades have witnessed numerous attempts to manipulate

the fatty acid composition of soybean oil to help it compete better in various uses,
but the cost of growing, segregating, and testing special varieties and resistance to
genetically modified oils have limited the appeal of these altered varieties.
2. COMPOSITION OF SOYBEANS
Table 1 shows the average composition of soybean seed (oil, protein, and some
amino acids) grown in the United States during recent years (3). Aside from varietal
differences, the composition is affected by various geographic/environmental fac-
tors. According to Hurburgh (5), oil is much more variable than protein from year-
to-year. States most distant from the center of the Corn Belt (probably those with
the greatest weather extremes) experience the most variability in composition.
Table 2 lists some of the environmental and cultivation practices that have an
TABLE 1. Typical Composition (wt% std. dev.)

of Soybeans (dry weight basis) (3).
Protein 40.69 0.51

Lysine 2.56 0.11
Methionine 0.57 0.03
Cysteine 0.72 0.06
Tryptophane 0.52 0.05
Threonine 1.54 0.07
Oil 21.38 0.64
Ash 4.56 0.34 (4)
Carbohydrate 29.4 3.29 (4)
TABLE 2. Soybean Protein and Oil Responses to Various

Environmental and Cultivation Practices (3).
Variable Protein Oil

a
High temperatures ?
Early season drought
Late season drought
Early frost/cold temperature b
Additional soil nitrogen
Increased fertility (P,S)
Late planting
Insect defoliation
Insect depodding ?
Rhizobium inoculation
a
? inconclusive; increase; decrease.
b
Oil is reduced because of refining loss to remove chlorophyll.
COMPOSITION OF SOYBEANS 579
observable effect on soybean protein and oil percentages. Maestri et al. (6) grew
soybean cultivars in several regions of Argentina and concluded that the protein
and oil contents were positively correlated with altitude. Protein was negatively cor-
related with latitude and precipitation, and oil was negatively correlated with tem-
perature and precipitation. Oil content in soybeans tends to be negatively correlated
with protein, but breeding soybeans for high protein while maintaining oil content
has been a priority of the U.S. soybean producers, and some progress has been
achieved (7, 8). The variety Prolina reportedly produces 22.7% oil and 45.5% pro-
tein on a dry weight basis. There also has been interest in reducing the oligosac-
charides that cause flatulence and reduce the digestibility and nutritive value of
soybeans.
Isoflavones are minor constituents of soybeans whose consumption is believed to
have beneficial effects (911). The benefits of isoflavones have encouraged the
direct consumption of soy protein in the United States. The concentration of
isoflavones changes with variety and growing conditions and has been reported
to be 1.22.5 mg/g in U.S. beans (9), 0.52.3 mg/g in Korean beans (10), and
0.23.5 mg/g in Japanese beans (11).
Table 3 shows the typical composition of the lipid phase of soybeans. Triacyl-
glycerols are the primary component. The 3.7% phospholipids content in the soy
beans is higher than that usually found in hexane-extracted oil, which is typically
TABLE 3. Typical Composition of Crude Soybean Oil.
Component % Std. Dev.

a
Triacylglycerol 94.4 1.4
Phospholipids 3.7b 1.2 (12)
Unsaponifable matter (1315) 1.31.6
Sterolsc (16) 0.236 0.053
Campesterol 0.059 0.018
Stigmasterol 0.054 0.013
-Sitosterol 0.123 0.027
5-Avenasterol (17) 0.005
7-stigmasterol (17) 0.005
7-avenasterol (17) 0.002
Tocopherols (16) 0.123 0.040
Alpha 0.0093 0.0044
Beta 0.0018 0.0028
Gamma 0.0834 0.036
Delta 0.029 0.010
Hydrocarbons (14, 15) 0.38
Free fatty acids (18) 0.30.7
Trace metals (18) ppm
Iron 13
Copper 0.030.05
a
By difference.
b
Based on 23 varieties chosen to represent a wide fatty acid composition.
c
Based on 13 varieties chosen to represent a wide range of composition.
580 SOYBEAN OIL
1.852.75% (19). Of the unsaponifiable matter of soybean oil, typically about

1.45% of the oil, 16% is sterols, 8.5% is tocopherols, and 26% is hydrocarbons.
The remaining 50% of the unsaponifiable matter consists of other minor and uni-
dentified products. The sterols are about 52% -sitosterol, 25% campesterol, and
23% stigmasterol. Maestrl et al. (4) reported similar proportions on the three major
sterols but also reported 5.4% 0.82 of 5-avenasterol, 3.8% 0.76 of 7-
stigmasterol, 1.3% 0.42 of 7-avenasterol, and traces of cholesterol in the total
sterols. The tocopherols are about 7.6% a, 1.5% b, 67.8% g, and 23.6% d. There is
considerable variation among plant varieties in the amounts and proportions of
molecular species of sterols and tocopherols (4, 16, 20). Vlahakis and Hazebroeck
(21) also have investigated the effects of planting locations and temperature on the
sterol and total tocopherol contents of a number of soybean varieties. They found
that growth temperature can cause as much as a 2.5-fold difference in sterol
content, with higher temperatures favoring higher amounts of sterols, increasing
the campesterol/-sitosterol ratio and decreasing total tocopherols. McCord et al.
(22) examined a number of soybean lines with low and normal contents of linoleate.
The low-linolenate lines averaged about 6% lower in tocopherol than the high-
linolenate lines, but some reduced-linolenate lines were not significantly different
from normal-linolenate lines in tocopherols. The a- and g-tocopherols tend to be
concentrated in the soybean germ, whereas d-tocopherol is concentrated in the
endosperm (23). The hydrocarbon fraction of soybeans consists of n-hydrocarbons
of chain length 14 to 33 plus squalene and small amounts of hexahydrofarnesyla-
cetone (14, 15). The squalene content is reported to be about 0.014% of the oil.
There seems to be considerable variation in the distribution of the hydrocarbon
chain lengths with plant variety, judging from the two examples in the literature.
Free fatty acids vary considerably with the age and soundness of the beans
but are seldom lower than the 0.1% of the crude oil (18). Damaged beans can
contain 18% free fatty acid as well as elevated iron and copper, 37 ppm and
0.080.18 ppm, respectively.
Refined oil usually retains little phospholipid, but damaged beans can have a sig-
nificant content of phosphatidic acid, and the amount of iron in the oil is related to
the amount of phosphorus (24). During deodorization, considerable amounts of
sterol and tocopherol may be removed from the oil. The proportion removed
depends on deodorization conditions, but a 30% to 40% decrease is not unusual
(25). Much of the hydrocarbons and squalene are lost to the deodorizer distillate
as well. Free fatty acids in fully refined oil are required to be <0.05% and unsapo-
nifiable matter <1.5% (26).
Table 4 shows the percentages and standard deviations of the methyl esters of 21
typical refined soybean oil samples. This composition is typical of most presently
commercial soybean varieties. The typical composition probably has been selected
through plant breeding because it is associated with good yield and other important
agronomic properties. It has been possible to change the composition of soybean oil
considerably, and Table 4 also shows the ranges of percentages that have been
reported for each methyl ester. Many of the changes in composition can be achieved
without great losses in yield or oil content, but lines with high or low palmitate
TABLE 4. The Averages and Standard Deviations of Methyl Esters from Typical Soybean
Oils and the Range Reported for each Methyl Ester.
Methyl Ester Typical Value %a (27) Range Achieved %
Myristate 0.04 0.5 (27) trace0.03 (4)

Palmitate 10.57 0.43 (27) 3.226.4 (33, 34)
Palmitoleate 0.02 0.04 (27) trace0.7 (29)
Stearate 4.09 0.34 (27) 2.632.6 (33, 35)
Oleate 22.98 2.01 (27) 8.679.0 (36, 37)
Linoleate 54.51 1.54 (27) 35.264.8 (3537)
Linolenate 7.23 0.78 (27) 1.719.0 (38, 39)
Arachidate 0.33 0.14 (27) trace0.7 (28)
Gondoate 0.18 (28) trace0.6 (4)
Behenate 0.25 0.20 (27) trace1.0 (4)
Lignocerate 0.1 (29)
Furanoid IIb 0.014 0.0086 (30) 0.00330.0290 (30)
Furanoid IIIc 0.015 0.0076 (30) 0.00840.0272 (30)
Saponification Value 190.4 (31, 32) 188.5201.6 (31, 32)
Iodine Value 132.7(31, 32) 114.0138.5 (31, 32)
a
Based on 21 commercial samples.
b
10,13-epoxy-11,12-dimethyloctadeca-10,12-dienoate.
c
12,15-epoxy-13,14-dimethyloctadeca-12,14-dienoate.
percentages tend to have reduced oil contents (33, 40, 41). Lines with high stearate
percentages suffer from low yields and sporadically from poor germination. Wang
et al. (42) tested lines with elevated palmitate or stearate in a number of tests of
germination and seedling vigor at three temperatures and found that, although
the high-saturate seed did well in these tests, vigor was negatively correlated
with saturate percentage. Most of the changes reported in Table 4 were attained
by traditional plant breeding or use of mutagenic agents. The high-oleic mutant
is an exception and was attained by direct genetic manipulation (37). High-oleate
lines developed by traditional plant breeding have been reported, but their oleate
percentage varies widely with growth environment, which limits their commercial
value (38).
The fatty acid composition of soybean oil changes considerably with maturity
and with seed oil deposition (15, 35, 43, 44). In typical soybean triacylglycerols,
the palmitate and linolenate tend to decrease with maturity, whereas linoleate
increases. Oleate tends to increase to a maximum and then decline slightly. Soy-
beans selected for atypical fatty acid compositions show quite different patterns
of change with maturity from typical soybeans.
Seitz (31) and Wesolowski (32) measured the saponification and iodine values of
a number of samples from various geographic locations, and their ranges and typi-
cal values are shown in Table 4.
Harp and Hammond (45) explored the stereospecific distribution of acyl groups
on the three positions of the glycerol molecule for soybean triacylglycerols with a
wide range in fatty acid composition. They found that the amount of an acyl group
582 SOYBEAN OIL
on a particular position was linearly related to the amount of that acyl group in the
whole triacylglycerol. At low concentrations of palmitate, stearate, oleate, and
linoleate in the total triacylglycerols, the amounts on the sn-1 > sn-3, but the
reverse was true at higher total concentrations. Palmitate and stearate were confined
to the sn-1 and sn-3 positions, whereas the oleate concentration was similar on all
three positions. Linoleate concentrations at the sn-2 position were generally greater
than those at the sn-1 and sn-3 positions, but the amount of linoleate on the sn-2
position seemed to be strongly and negatively correlated with the amounts of satu-
rates on the sn-1 and sn-3 positions. Plots of linolenate concentrations at particular
positions versus the concentrations in the whole triacylglycerol showed consider-
ably more scatter than plots for the other acyl groups, but they generally showed
the amount of linolenate on sn-2 > sn-1 > sn-3. The saturate percentages also
seemed to influence the amounts of linolenate on the sn-2 position positively and
on the sn-3 position negatively. Table 5 shows the stereospecific distribution of typi-
cal soybean triacylglycerols.
Theoretically, stereospecific data can be used to predict the acylglycerol struc-
ture using the 1-random-2-random-3-random distribution theory (47), if one
assumes the fatty acid composition of the three glycerol positions are individually
controlled but that the combinations of the three positions are random. However,
the change in fatty acid composition with maturity, described in the previous
paragraph, shows that the triacylglycerol composition is unlikely to be truly random
in the combination of the three glycerol positions. In addition, soybeans from the
same plant or pod can have slightly different acyl group compositions, so pooled oil
from many seeds and plants is unlikely to be exactly random in its glycerol position
combinations. Thus, such a calculation can lead only to approximate compositions.
Neff et al. (48, 49) partially separated the triacylglycerols of soybean oils
with a wide range of fatty acid compositions using high-performance liquid
TABLE 5. Stereospecific Distribution of Acyl Groups in the Triacylglycerols, Phospha-

tidylcholine, Phosphatidylethanolamine, and Phosphatidylinositol of a Typical Soybean
(45, 46).
Compound/Acyl group 16:0 18:0 18:1 18:2 18:3
Triacylglycerols 11.8 4.6 29.4 47.1 7.2

sn-1 19.3 7.5 25.4 39.6 7.8
sn-2 2.9 0.8 27.9 61.1 7.4
sn-3 14.9 6.4 34.8 38.9 5.0
Phosphatidylcholine 11.2 11.9 8.6 58.6 9.9
sn-1 16.0 22.6 7.3 38.3 6.0
sn-2 4.1 3.7 9.9 71.1 11.2
Phosphatidylethanolamine 16.0 8.3 6.8 57.3 11.7
sn-1 28.5 17.1 5.0 42.4 7.1
sn-2 3.5 2.4 8.7 73.8 11.7
Phosphatidylinositol 22.2 19.3 6.1 43.4 9.3
sn-1 45.1 35.2 5.3 17.1 2.4
sn-2 4.6 3.4 5.9 70.9 15.3
chromatography, and their results are shown in Table 6. These data show how the
amounts of the triacylglycerol species change with fatty acid composition.
The primary phosphatides of soybean oil are phosphatidylcholine, phosphatidyl-
ethanolamine, and phosphotidylinositol, which generally make up 55.3%, 26.3%,
and 18.4% of the total phosphatides, respectively (50). The stereospecific distribu-
tion of the acyl groups in these phospholipids for a typical soybean lipid is shown in
Table 5. In all the phospholipids, the saturated acyl groups are concentrated in the
TABLE 6. Acyl and Triacylglycerol Composition in mol% of Soybean Oils Having a Wide
Range of Fatty Acyl Compositions (48, 49).
Sample Number

Acyl group 1 2 3 4 5
Palmitate (P) 3.9 21.4 23.6 28.2 8.5

Stearate (S) 3.3 3.3 19.0 3.9 26.5
Oleate (O) 28.5 23.6 9.3 13.9 18.0
Linoleate (L) 61.8 49.0 38.0 43.8 38.9
Linolenate (Ln) 2.5 2.7 10.0 10.2 8.2
Triacylglycerol Species
LnLL 2.0 1.2 1.4 3.2 2.6
LnLnO 0.1 0.1 0.2 0.1
LnLnP 0.4 0.6 0.1
LLL 30.0 11.5 3.7 9.6 6.5
LnLO 1.7 1.5 0.8 2.0 1.9
LnLP 0.4 1.7 6.9 9.4 2.2
LLO 26.9 14.4 3.6 8.7 7.1
LnOO 0.4 0.5 0.1 0.3 0.3
LLP 6.4 20.7 17.5 21.4 11.6
LnOP 0.1 1.0 1.4 2.1 0.4
LnPP 0.1 1.5 2.0 0.1
LOO 13.9 7.3 1.3 3.1 2.5
LLS 3.6 2.6 7.7 2.0 13.0
LOP 3.7 16.3 7.4 12.2 6.4
PLP 0.8 8.6 13.8 14.8 2.0
OOO 4.6 2.1 1.0 0.8 1.1
LOS 2.6 2.3 3.9 1.3 11.8
POO 0.9 3.2 0.6 1.0 0.5
SLP 0.8 2.2
LnSS 16.0 3.0 8.8
POP 0.2 1.7 1.2 1.5 0.3
PPP 0.1
SOO 0.7 0.5 0.3 0.2 2.1
SLS 0.3 0.2 6.5 0.3 12.3
SOP 0.1 0.4 1.3 0.2 1.4
PPP 0.6 0.6
SOS 0.6 3.4
PSS 0.2 0.1
SSS 0.1
584 SOYBEAN OIL
sn-1 position and the unsaturated acyl groups, especially linoleate, on the sn-2 posi-
tion. Phosphatidylinositol tends to be richest in palmitate and stearate, whereas
phosphatidylcholine has the least palmitate. Wang et al. (51) reported the stereospe-
cific distribution of the acyl groups in the various phospholipids types and the
amounts of particular acyl combinations for soybean lipids with a wide variety
of fatty acid compositions. Some phosphatidic acid and lysophospholipids also
may be present as a result of hydrolysis of the phospholipids (52). The amounts
of the hydrolytic products usually increase with age and damage to the beans (53).
Soybeans also contain 170 47 ppm of cerebrosides in which the sugar is glu-
cose and the chief fatty acid is 2-hydroxy palmitic acid (54). Traces of ceramides
also are present. These are believed to play a role in cell signaling in the soybean
plant.
Crude soybean oil contains about 1.9 ppm of Vitamin K1 or phylloquinone (55).
This vitamin plays a role in blood coagulation and bone metabolism. During refin-
ing, some Vitamin K1 may be lost (56, 57), especially during deodorization. Hydro-
genation of the fat converts some of the Vitamin K1 to 20 ,30 -dihydrovitamin K1 (58).
Wilson et al. (7, 8, 39, 59) have reviewed the genetic control of fatty acid bio-
synthesis in soybeans and discussed the advantages of soybean oil with special
compositions. Oil with reduced palmitate is available presently in a limited market.
The commercial introduction of low-linolenate soybeans has been inhibited by the
availability of corn oil, which has a composition like very low-linolenate soybean
oil. The price differential between these oils often is smaller than the costs of con-
tract growing, segregating, and processing low-linolenate soybeans. High-oleate
soybean oil is stable under frying conditions, but this trait alters the flavor of the
fried products (60). The acceptance of high-oleate soybean oil also suffers from
public concern about the growth and consumption of plants produced by direct
genetic modification.
There are small amounts of two acyl groups containing furan rings in soybeans
(30). These oils are reported to be the sources of the odorous compound 3-methyl-
2,4-dione by photo-oxidation (61), but Kao et al. (62) were not able to find differ-
ences in flavor of photo-oxidized varieties with high and low content of these acyl
groups.
3. PHYSICAL PROPERTIES OF SOYBEAN OIL
The physical properties of fatty acids vary with their chain length, unsaturation, and
other substituents and change with temperature. Numerous attempts have been
made to develop equations that will predict these properties. Soybean oils proper-
ties should reflect its constituents and, especially, its fatty acid composition, and
physical properties have frequently been measured for typical soybean oils, but
there have been fewer measurements of soybean oils with modified fatty acid com-
positions.
Table 7 shows the values of physical properties of soybean oil of typical com-
position. Seitz (31) examined 77 samples of soybean oil from various parts of the
PHYSICAL PROPERTIES OF SOYBEAN OIL 585
TABLE 7. Some Physical Properties of Typical Soybean Oil.
Density 20 C 0.9165 to 0.9261 g/mL (31, 32) Decreases 0.000643 to

0.000668 g/mLC (6367)
Specific Heat Capacity 20 C 0.448 cal/g C Increases 0.000616 cal/g C (68)
Melting Point 0.6 C (35)
Cloud Point 9 C (69)
Pour Point 12 to 16 C (69, 70)
Heat of Combustion 9450-9388 cal/g (71)
9135 91 cal/g (72)
Heat Transfer Coefficient 269.7 watts/ K M2 at 180 C (73)
Surface Tension 30 C 27.6 dyne/cm Decreases 0.077 dyne/cm C (63, 64)
Viscosity 20 C 58.562.2 cP (31)
Refractive Index nD20 C 1.47331.4760 (32)
Vapor Pressure 1m at 254 C (74)
Heat of Vaporization 44,200 cal/mol (74)
Electrical Resistivity 24 C
Dry 23.7 Tohm cm (75)
Water Saturated 7.25 Tohm cm (75)
Smoke Point 245 C (76)
Flash Point 324 C (76)
Fire Point 360 C (76)
world over a seven-year period and reported densities at 20 C ranging from
0.9165 g/mL to 0.9210 g/mL. Wesolowski (32) examined the density of 53
Polish soybean oils at 19.9 C and reported values ranging from 0.9202 g/mL to
0.9165 g/mL. The following correlations of density and other variables were found:
with refractive index 0.62, with iodine value 0.64, with saponification value 0.34,
and with acid value 0.59. Yokota and Tachimori (77, 78) also reported a close
relation between density and iodine value. Halvorsen et al. (79) and Rodenbush
et al. (80) developed equations to predict the density of vegetable oils that took their
fatty acid compositions into account and predicted densities of soybean oils with
<0.1% error. The density of vegetable oils changes approximately linearly with
temperature, and Kravchenko et al. (65, 66) found the density decreased
0.000668 g/mL C between 0 C and 100 C, whereas Alvarado (63, 64) found a
value 0.000643 between 20 C and 70 C, and Noureddini et al. (67) found
0.0006674 between 23.9 C and 110 C.
The densities of soybean oil-solvent mixtures at various temperatures are impor-
tant for engineering calculations and have been reported for hexane, ethylene
dichloride, and tricholoroethylene at 25 C, 37.8 C, and 50 C (81); Skellysolve B
at 20 C, 10 C, 0 C, 10 C, 25 C, and 40 C (82); dichloromethane at 25 C (83);
and hexane at 25 C (84).
The specific heat capacity of soybean oil was measured by Clark et al. (85)
and varied from 0.448 cal/g C to 0.666 cal/g C between 1 C and 271 C. Specific
heat increased linearly with temperature at 0.00070 cal/g C. Tochitani and Fujimo-
to (68) measured the specific heat capacity of soybean oil from about the
586 SOYBEAN OIL
approximate melting point to 150 C and found a linear increase that fit the follow-
ing equation:
Sp: Heat Capacity in cal=g C 0:4353 0:000616 T; 1
where T is the temperature in C. Their data agreed closely with those of Clark et al.
(85) but were slightly higher than those reported by Kasprzycka-Guttman et al.
(86), who made measurements between 70 C and 140 C. Wang and Briggs (87)
estimated the heat capacities of soybean oils of various compositions based on
an equation by Morad et al. (88). They calculated that high-oleate oils should have
a slightly higher heat capacity and low-saturate oils a slightly lower heat capacity
than typical soybean oil, and the change with temperature should be 0.00057 cal/
g C. Their equation agreed with their experimental values within 5%.
Miller et al. (73) determined the heat-transfer coefficient for soybean oil at fry-
ing temperatures and found that they varied from 261.3 watts/ KM2 to 276.2 watts/

KM2 between 170 C and 190 C, where M2 is square meters of surface.
The melting of natural fats and oils usually occurs over a considerable tempera-
ture range, and soybean oils typical melting range is below 0 C. The availability of
differential scanning calorimetry (DSC) at low temperatures has made information
on melting of soybean oil available, and interest in using vegetable oils as fuels has
also sparked measurements of their cloud and pour points. Table 8 (41) gives the
temperatures of onset, maximum, and end of melting for various types of soybean
oil. Table 7 gives the cloud and pour points of typical soybean oil. Wang and Briggs
(87) also gave DSC curves for the melting of high-oleate, low-saturate, and low-
linolenate soybean oil. Hagura and Suzuki (89, 90) used the change in electrical
capacitance of oil samples to obtain the melting range of soybean oil and found
the results agreed with those obtained by DSC.
Seitz (31) measured the viscosity at 20 C of 77 soybean oils from four
geographic locations, and the range of variation was 58.1cP to 62.2cP (Table 7).
Viscosity decreases with temperature, and the relation is not linear. Kinematic
values (viscosity/density) have been reported at 20 C and 80 C by Chioffi (91)
and by Miller et al. (73) at frying temperatures (170190 C); dynamic viscosities
have been reported between 0 C and 100 C by Kravchenko et al. (65, 66), between
23.9 C and 110 C by Nourreddini et al. (67), between 20 C and 70 C by Alvarado
TABLE 8. The Onset, Maximum Rate, and Termination of Melting Temperatures of

Soybean Oil with Various Fatty Acid Compositions % (41).
Class Onset Maximum Termination 16:0 18:0 18:1 18:2 18:3
Typical 39.6 9.4 0.6 11.4 4.2 26.1 50.3 7.9

18:0 " 13.7 18.3 20.7 10.1 22.8 17.3 42.2 7.7
16:0 & 18:0 " 17.1 16.8 18.9 24.6 18.7 8.6 37.5 10.7
16:0 " 21.8 8.4 11.6 28.0 4.7 13.8 42.1 11.4
16:0 # 46.1 13.8 8.1 3.4 2.6 18.0 64.8 11.2
PHYSICAL PROPERTIES OF SOYBEAN OIL 587
(63, 64), and between 1 C and 60 C by Arissen (92). Dahlberg et al. (93) were able
to predict the viscosity of soybean and other oils from the Fourier transform infra-
red spectra. Rodenbush et al. (80) calculated the viscosity of oils by relating visc-
osity to a function they termed the reduced density, which they could calculate from
the fatty acid composition.
Several authors have fit their viscosity-temperature data to equations (6367, 87,
94). Some of these come with a claim of theoretical significance, but all have
enough variables to fit the data well. One of Alavarados equations (63, 64) is
ln m ln m0 E=RT; 2
where E/R was 3262 and ln m0 was 6.997 for soybean oil. Wang and Briggs (87)
reported graphically the change of viscosity with temperature from 10 C to 90 C
for soybeans with altered fatty acid compositions. They found the viscosity of high-
oleic soybean oil higher and low-saturated soybean oil lower than that of typical
soybean oil.
Miller et al. (73) determined the kinematic viscosity of soybean oil at temp-
eratures of 170 C, 180 C, and 190 C, and obtained values of 3.151 cm2/sec,
2.880 cm2/sec, and 2.614 cm2/sec, respectively. The viscosities of soybean oil-hex-
ane (Skellysolve B) mixtures at temperatures between 20 C and 40 C were inves-
tigated by Magne et al. (84). Ibemesi and Igwe (95) examined the reduced viscosity
(viscosity/concentration) of solutions of soybean oil in toluene, xylene, cyclohex-
ane, and tetrahydrofuran. They found an anomalous reduced viscosity increase at
concentrations below about 0.12 g/mL that they attributed to clustering of the fat
molecules in the solvent. Erhan et al. (96) determined the kinematic viscosity of
blends of typical soybean oil with polyalphaolefins and isobutyrl oleate and
high-oleic soybean oil with isotrideceyl adipate and mineral oil to achieve viscos-
ities suitable for lubricants.
The surface tension of soybean oil at 20 70 C was reported by Alvarado (63, 64)
and is given in Table 7. The surface tension decreased linearly with temperature at
0.077 dyne/cm C.
Wesolowski (32) examined the refractive index of 53 samples of soybean oil
from Poland, and the range is given in Table 7. Sietz (31) reported average values
for samples from several geographic locations, and these values (1.47471.4752)
fall near the mean of Weslowskis samples. Refractive index depends on chain
length and unsaturation (97) and often has been used to follow hydrogenation
(98102). Refractive index also has been used to follow autoxidation (103). A clo-
sely related quantity, the dielectric capacitance also has been used to assess the
quality of frying oil (104). Perry et al. (74) measured the vapor pressure of soybean
oil at various temperatures and found that the data fit the equation:
log P 18:3 9650=T; 3
where P is the pressure in microns and T is in K. The also estimated the heat of
vaporization (Table 7).
588 SOYBEAN OIL
Tomoto and Kusano (105, 106) measured the solubility of carbon dioxide,
nitrogen, hydrogen, and oxygen in soybean oil between 0.2 atm and 1 atm and
between 30 C and 70 C. The Bunsen coefficient (volume of gas at standard condi-
tions / volume of soybean oil at 760 mm) at 30 C was 1.018 for carbon dioxide,
0.086 for nitrogen, and 0.048 for hydrogen. The Bunsen coefficient of oxygen at
30 C was 0.141 but increased with temperature, probably because of oxidation
during the measurement. The Bunsen coefficient decreases linearly to zero at
zero gas pressure. The natural logarithm of the Bunsen coefficient versus 1/T in
K is linear, and the constant is the heat of solution of the gases divided by the
gas constant. These heats of solution are 2.42 kcal/mol for carbon dioxide,
2.58 kcal/mol for nitrogen, and 3.86 kcal/mol for hydrogen. From this relation,
one can calculate the solubility at any temperature and pressure in the range of the
study. Comparison of the values for soybean oil with olive and linseed oil suggested
that the Bunsen coefficients are influenced by the degree of unsaturation of the oil.
The viscosity of soybean oil decreased with the amount of carbon dioxide
dissolved, but dissolved nitrogen slightly increased the viscosity.
Loncin (107) reviewed the data on the solubility of water in fats and oils. For
typical soybean oil, the solubility of water was 0.11% by weight at 22 C and
rose to 0.19% at 60 C. The solubility of water decreases with fatty acid chain
length and increases with the percentage of free fatty acids.
The vapor pressures of soybean oil-hexane mixtures between 75 C and 120 C
were reported (108, 109), and similar data for soybean oil with commercial hexanes
was reported by Smith (110). Arnold and Breuklander (83) measured the boiling
point of dichloroethylene-soybean oil mixtures and found the log (V.P.) was a linear
relation of the mole fraction of oil. Kusano (111, 112) measured the vapor pressure
(P) of soybean oil-solvent mixtures that included hexane, benzene, and carbon tet-
rachloride between 20 C and 50 C and found linear relations between log P and 1/
T. Anikin et al. (113, 114) measured the vapor pressure of mixtures of soybean oil
with the khladon 113 (trichlorotrifluoroethane) between 30 C and 100 C. Aeber-
hard and Spekuljak (115, 116) measured the vapor pressure of hexane in hexane-
soybean oil mixtures and found the vapor pressure at 25 C could be predicted by
the equation
P 9128x 0:2807x2 0:004695x3 ; 4
where P is the vapor pressure in Torr and x is the weight percentage of solvent in the
mixture.
Tekin and Hammond (75) measured the resistivity of soybean oil and found it
decreased logarithmically with temperature from about 100 Tohm cm at 5 C to
0.251 Tohm cm at 100 C. The resistivity was decreased by saturating the oil with
water and the addition of oleic acid, a-tocopherols, b-carotene, phospholipids, and
monoacylglycerol.
The smoke, flash, and fire points of soybean oil have been determined by the
Cleveland Cup method and show considerable variation. Dickhart (117) reported
a smoke point of 138 C while Detwiler and Markley (76) reported 241250 C.
GRADING 589
Detwiler and Markley (76) found that the smoke point varied considerably with the
degree of refining, especially the removal of free fatty acids, and also with the mode
of oil extraction. Yen et al. (118) found a smoke point of 191 C, which was raised
several degrees by the addition of phenolic antioxidants. The flash point of soybean
oil, the temperature at which vapors coming from the oil will catch fire from an
ignition source, were reported as 304 C (117), 326331 C (76), 174 C (69),
318 C (70), and 320 C (119). The low value reported by Ali et al. (69) was
obtained by using a Pensky-Martens closed tester and ASTM method 093-90.
The flash points of hexane-soybean oil mixtures were determined and correlated
with headspace gas chromatography data (120).
Fire points or self-ignition temperatures (SITs) for soybean oil by using the
Cleveland Cup method, which uses a brass cup, were reported to be 356363 C
(76) and 400 C using a stainless-steel cup apparatus (71). The burning rate of soy-
bean oil was 4.3 g/m2sec, flame height 129 mm, and irradiance 0.153 kW/m2 (71).
Kowalski (119) studied the self-ignition temperature in a differential scanning
calorimeter heated at rates of 4090 C/min and under 8002800 kPa of oxygen
pressure and found values of 260290 C for soybean oil. He found the addition
of copper wire to the sample decreased the self-ignition temperature by 515 C.
The self-ignition temperature was inversely related to oxygen pressure. Wakakura
(121, 122) used a scanning calorimeter at an oxygen pressure of 980 kPa with
soybean oil spread on glass wool and in bulk and found self-ignition temperatures
of 147 C and 376 C, respectively.
4. GRADING
To facilitate soybean marketing, the U.S. Federal Grain Inspection Service (FGIS)
established grading standards for soybeans (Table 9) (123), and the FGIS website
(124) provides much more detailed information than can be provided here (124).
TABLE 9. Official Grades and Grade Requirements of the Federal Grain Inspection
Service, United States Department of Agriculture.
Maximum Limits

Minimum
Damaged Kernels
Test Soybeans
Weight Heat Foreign of Other
per Bushel Damaged Total Material Splits Colors
Grade (lbs) (%) (%) (%) (%) (%)
U.S. No. 1 56.0 0.2 2.0 1.0 10.0 1.0

U.S. No. 2 54.0 0.5 3.0 2.0 20.0 2.0
U.S. No. 3 52.0 1.0 5.0 3.0 30.0 5.0
U.S. No. 4 49.0 3.0 8.0 5.0 40.0 10.0
U.S. Sample Grade
590 SOYBEAN OIL
Soybeans are classified into two classes based on color, Yellow Soybeans and
Mixed Soybeans. There are four numerical grades (U.S. No. 1, 2, 3, and 4) and a
U.S. Sample Grade for each class. Sample Grade designates those soybeans that do
not meet the requirements of any of the numerical grades. Six factors are consid-
ered in assigning a grade designation: test weight, amounts of beans that are
damaged or heat damaged, and amounts of foreign material, splits, and soybeans
of other colors. Although important to processors because they affect yields and
qualities of finished products, the FGIS official grades do not consider moisture,
protein, and oil contents, but these factors may be specified on contracts in some
markets. Near infrared transmission (NIT) spectroscopy is widely used to rapidly
estimate (within less than 2 min after sampling and without any sample preparation
required) moisture, protein, and oil contents. Brumm and Hurburgh (125) devel-
oped a computer program to estimate the process value of soybeans based on their
composition and selling prices of oil and meal. In some cases, price premiums are
offered for soybeans high in oil content or high in both oil and protein contents, and
details of the program are available on the Internet (126).
Beans low in test weight may contain less oil. Test weight is the weight in
pounds of grain per Winchester bushel (35.2 L) and is determined by using an Offi-
cial Test Weight Apparatus and a 11/4-quart (1.18 L) sample before removing for-
eign material. All other grading factors are measured as percentages of total sample
weight. Foreign material, which is other grains, weed seeds, pods, leaves, stems,
etc., reduces oil and protein contents and storage life. Foreign material is deter-
mined by sieving a sample. All materials, including soybeans and soybean pieces
that readily pass through an 8/64-inch (3.2-mm) round-hole sieve and all material
other than soybeans remaining on the sieve after sieving are considered to be for-
eign matter. Split soybeans, which result from mechanical damage during handling
and over drying, reduce storage life and oil yield, and increase losses during oil
refining. Splits (typically the cotyledon splits into two halves) and broken beans
(more than two pieces) increase free fatty acid (FFA), phosphatides, iron, and per-
oxide contents of the crude oil. Heat-damaged beans have high-FFA content and
darken the oil color, both changes in oil quality increase refining loss (127). Splits
are defined as beans with more than one-fourth of the bean removed and are not
damaged. Splits are determined by sieving a portion of the grain after removing
the foreign material. Damaged beans reduce the storage life of the beans and oil
yield in processing, cause the oil to be dark-colored and poor in flavor, and increase
losses during oil refining (128). Soybeans and soybean pieces that are badly
damaged by the ground, weather, frost, heat, insects (stinkbug-stung kernels are
considered at one-fourth the actual percentage), mould, or sprouting are considered
to be damaged. Damaged beams are determined by hand picking after removing
foreign material. Soybeans of other colors may affect oil color by contributing
undesirable pigments and are those beans that are green, black, brown, or have mul-
tiple colors.
Almost 27 million MT of soybeans were exported from the United States during
the 2002 crop year, of which 4.8% was U.S. No. 1, 94.6% was U.S. No. 2, 0.4% was
U.S. No. 3, and 0.1% was U.S. No. 4. By comparison, Brazilian soybeans are
RECOVERY OF OIL FROM SOYBEANS 591
typically slightly higher in oil content (6-yr average of 1.2% higher oil content),
foreign matter, damage, free fatty acid, and moisture contents and lower in test
weight (129).
5. RECOVERY OF OIL FROM SOYBEANS
Soybeans are economically important because of their high qualities and quantities
of oil and protein. From one bushel of soybeans (60 lb, 27.2 kg), crushers typically
recover 11.1 lb (5.0 kg) of crude oil, 44.3 lb (20.1 kg) of meal (48% protein), and
3.3 lb (1.5 kg) of hulls with the remainder being shrinkage. According to the U.S.
Department of Agriculture statistics, the oil accounts for about one-third of the
returns in processing soybeans with the protein in the form of meal accounting
for the remainder (130). Over the past five years, the meal (48% protein) has ranged
in yearly average prices of $153289/MT (6.913.1 cents/lb), whereas the oil has
ranged $311569/MT (14.125.8 cents/lb). Hulls have limited outlets, mostly in
cattle feeds, and sell for about $66/MT (3 cents/lb) and return $4.04/MT of
soybeans ($0.11/bu). During the same period, the average price of soybeans in
the United States ranged from $167270/MT ($4.547.35/bu) and crushing
margins, the difference in soybean price and crusher returns, averaged $23.1
56.2/MT ($0.631.53/bu).
Farmers often store their soybeans in metal bins on the farm or in concrete silos
at local elevators for a fee. This allows farmers to sell their crop later in the year
when prices usually increase. Soybeans should be stored at less than 13% moisture
to assure safe storage and preservation of the quality. This moisture content is
usually achieved by drying in the field before harvesting. Lower moisture contents
increase the tendency of soybeans to split during handling to form two half pieces
of cotyledon. Higher moisture content during storage can lead to mold damage or
heating damage due to seed respiration (131). These forms of damage can affect
soybean grade and oil quantity and quality when processed.
The processing of soybeans has been described in more detail elsewhere than
can be done here (132134). Oil is recovered today by either mechanical means
or through the use of organic solvents. In the preindustrial revolution period, soy-
beans were merely pressed with lever or animal-driven screw-operated batch
presses. Around the turn of the Twentieth Century, when soybeans became a viable
commercial crop in the United States, steam-powered hydraulic batch presses were
used. Today, electric-powered continuous screw-presses, often referred to as
expellers (but this is a trademarked name for screw presses manufactured by one
supplier), or continuous countercurrent solvent extractors are used.
In either case, soybeans are pretreated prior to oil recovery to either make oil
recovery easier or more complete, or to increase the value of the defatted solids
known as meal. Usually, soybeans arriving from the farm or elevator are cleaned
to remove stems, leaves, pods, broken grain, dirt, stones, and extraneous seeds
using shaker screens and aspirators. It is usually advantageous to remove the major
portion of the hulls because they are low in oil (<1%) and protein. The hulls of
592 SOYBEAN OIL
soybeans account for 78% of the weight. Dehulling reduces the material going
downstream into costly operations and increases the protein content of the meal.
Dehulling raises the meal protein content by about four percentage points (i.e.,
from 44% for undehulled solvent-extracted soybean meal to 4849%) and reduces
fiber content (from 7.0% to <3.3%). The formulated feed market prefers high-pro-
tein and low-fiber meal, especially in manufacturing swine and poultry feeds. The
hulls are relatively easy to remove from soybeans compared with those of other oil-
seeds, simply cracking the bean into 68 pieces to free the hull using corrugated
roller mills and aspirating the hulls away from the oil- and protein-rich cotyledon,
known as meat, is effective. Consistent bean size is important to proper cracking
and drought-caused shrinking and wrinkling make dehulling much more difficult
and less efficient (135). Often, the aspirated hulls go to gravity tables to scavange
any small meats aspirated with the hulls. Usually, cleaned soybeans are conditioned
prior to cracking to improve dehulling efficiency by heating and drying the beans to
about 9.5% moisture and allowing the moisture to equilibrate for 17 days within
the bean to loosen the hull. Various hot-dehulling schemes have also been devised to
increase dehulling efficiency, and are often used in northern latitudes where the pro-
tein contents of soybeans, and, consequently, meal protein levels, may be lower and
specified protein levels cannot be achieved without more complete hull removal.
In the 1930s, soybeans were widely processed by screw pressing after cooking
the seed. A typical process diagram for screw pressing soybeans is shown in
Figure 1 and a plant photo is shown in Figure 2. The beans are heated and the
oil is squeezed out. The pressed oil usually goes to settling basins to reduce fine
Soybeans
CLEANING Foreign Matter
CRACKING ASPIRATING Hulls

( optional)
Meats
DRYING
OR COOKING
Foots SCREW PRESSING Cake MEAL GRINDING
SETTLING MEAL COOLING
POLISH FILTERING Foots Partially Defatted

Meal
Crude Oil
Figure 1. Process flow diagram for screw pressing soybeans.
Figure 2. Photograph inside a modern soybean screw-press plant (courtesy of West Central
Cooperative, Ralston, IA). (This figure is available in full color at http://www.mrw.interscience.
wiley.com/biofp.)
solids content, with the fines being recycled to the screw press. The oil then goes to
polish filters before being placed into storage for shipment to a refinery. Today, in
the United States, there are less than a half-dozen traditional screw press plants
(excluding extrusion-expelling, which will be discussed later). Only one screw-
press plant crushing more than 800 MT/day exists, and it is located in Ralston,
IA. Under optimum processing, the meal can contain as low as 46% residual
oil, which contributes metabolizable energy to livestock consuming screw-pressed
meal. As a result of the heat treatment during cooking and screw pressing,
increased rumen-bypass characteristics improves feed efficiency in high producing
dairy cattle. Thereby, the meal may sell for premium prices over solvent-extracted
meal when adequate numbers of dairy animals are located nearby. As this meal is
used to feed ruminants, the beans are not usually dehulled.
Direct solvent extraction is the most widely used oil-recovery method for
soybeans, but it also requires considerable capital and large scale to compete. In
actual practice, solvent extraction is used to crush over 98% of the soybean pro-
cessed in the United States. Process flow diagrams are shown in Figures 3 and 4.
Most soybean solvent-extraction plants process more than 2,500 MT/day (Figure 5),
and some are capable of processing as much as 5,000 MT/day (especially newly
constructed plants in Brazil). Direct-solvent-extraction plants smaller than 1,000
MT/day have difficulty competing in the United States. At various times, soybeans
have been extracted commercially with petroleum distillate fractions that resemble
gasoline, acetone, carbon disulfide, ethanol, trichloroethylene, and even water,
594 SOYBEAN OIL
Soybeans
CRACKING
ASPIRATING GRAVITY Hulls
TABLING
( optional)
Meats
Cracked Meats
CONDITIONING
FLAKING
Flakes
EVAPORATING EXPANDING
( optional)
Solvent
Miscella Collets
(oil and solvent)
SOLVENT
STRIPPING
EXTRACTING
Crude Oil Marc FLASH

(solids and solvent) DESOLVENTIZING Solvent
( optional) White Flakes
GRINDING
MEAL Enzyme-active
DESOLVENTIZING Flour
TOASTING
COOLING
GRINDING
Toasted Meal
Figure 3. Process flow diagram for direct solvent-extracting soybeans.
which is not a true solvent but facilitates oil separation by creaming. A petroleum
distillate containing a mixture of hexane isomers having a typical boiling range of
65 C to 71 C is the only solvent used today. These products typically contain 45%
to 70% n-hexane. n-Hexane is considered a neurotoxin in the United States and has
proven toxicity at high concentrations. The U.S. Occupational and Safety Admin-
istration has set the maximum workplace exposure level at 500 ppm and a time-
weighted average not to exceed 50 ppm (136). In recent years, there has been con-
siderable interest by the soybean industry in alternative solvents to hexanes because
of increasing environmental and safety concerns. Alternative solvent technologies
have been extensively reviewed (137139).
Figure 4. Depiction of equipment and process flow diagram for direct solvent-extracting soybeans (courtesy of French Oil Mill Machinery Co., Piqua, OH).
595
596 SOYBEAN OIL
Figure 5. Photograph of a modern soybean-extraction plant (courtesy of Bunge North America,

Council Bluffs, IA). (This figure is available in full color at http://www.mrw.interscience.wiley.
com/biofp.)
Cleaned and dehulled soybeans are conditioned by heating to 74 C to soften the
meat prior to flaking using smooth roller mills. Proper cracking and conditioning
are important to achieve the desired cell distortion or cell rupture that is necessary
for efficient extraction and to prevent production of excessive amounts of fine meat
particles that impede proper flaking or extraction. Highly distorted cells are desired
(140) so that cell walls and pseudo-membranes around oil bodies are sufficiently
ruptured, and the oil can be easily contacted by the solvent and leached out. Soy-
beans are typically flaked to 0.25 mm (1012 thousandths of an inch) to achieve the
desired distortion (141). The flakes may be conveyed directly to the extractor or to
an expander. In recent years, expanders have been adopted to achieve increased cell
distortion and to produce an easily extractable porous pellet (collets) that is more
dense than flaked soybeans. Thereby, more mass of material can be placed into the
fixed volume of the extractor, the oil is more quickly extracted reducing extraction
time, and the solvent drains more completely reducing the load on meal desolven-
tizing equipment. All of these factors increase plant throughput capacity (142144).
Plants vary in the amounts of flakes that are expanded, typically about one-third of
the flake production, but in a few cases, all flakes are expanded. Although there is
not universal agreement, expanding may also improve oil quality by quickly inac-
tivating phospholipases, which cause phospholipids to become nonhydratable. In
the authors opinion, adoption of expanders is the most significant change in solvent
extraction during the past quarter century.
Figure 6. Photograph inside a modern direct solvent-extraction plant processing soybeans

(courtesy of Crown Iron Works, Minneapolis, MN). (This figure is available in full color at
http://www.mrw.interscience.wiley.com/biofp.)
Soybeans are exclusively extracted in the percolation mode as opposed to the

immersion mode used during early days of soybean extraction. A photograph of
a modern chain extractor is shown in Figure 6. The percolating solvent flows by
gravity through the bed. Solvent is always passed countercurrent to the transport
of meal solids. There are several different types of extractors, including chain
and basket types, and shallow- and deep-bed types. Soybean flakes or collets are
extracted for 30 45 min in six or more stages.
The best quality oil, low in non-triacylglycerole components, is extracted first,
and with more exhaustive extraction, poorer quality oil is recovered. Thus, at low-
residual oil levels, the proportions of phosphatides, free fatty acid, and pigments
that are extracted are greater and so is the refining loss. However, the current indus-
try practice strives for the most complete extraction possible, typically in the range
or 0.5% to 1.25% residual oil. For this reason, exhaustive laboratory devices, such
as a Soxhlet extractor with ground material, are not very useful when trying to
achieve oil that is representative of that produced by a commercial extractor,
and, for best results, the solvent should be percolated in stages through a bed of
flaked material.
The full miscella (oil-rich extract) containing 2030% oil drains from the fresh-
est flaked or expanded meats and is sent to solvent-recovery operations. The opera-
tions include two-stage evaporators and an oil stripper. The oil content exiting the
first-stage evaporator is 6570% oil and is heated with vapors from the desolventi-
zer-toaster. After the second-stage evaporator, the oil content is 90 95% oil. The oil
stripper uses steam-injection vapor, high heat, and high vacuum to remove the
598 SOYBEAN OIL
solvent to less than 0.2% remaining in the oil. The temperature of the oil in the
stripper should not exceed 115 C to prevent scorching the oil and causing dark col-
or. Flash point determination is an easy method to assure that the solvent-evapora-
tion equipment is operating as it should and the flash point should exceed 150 C.
All evaporated solvent is recycled to the extractor. The oil should be sent to a
vacuum dryer to remove any residual stripping steam condensate and the dry oil
immediately cooled prior to placing into storage.
As a result of natural antioxidants (i.e., phoshpahtides, tocopherols), crude soy-
bean oil can be stored for a long time in large tanks provided the oil is first cooled to
ambient temperature and has limited access to air. The crude oil should be low in
moisture to prevent hydrolysis. Gummy deposits of phosphatides may spontaneously
form in the bottoms of storage tanks and tank cars used for shipping crude oil.
There has been much speculation about using supercritical carbon dioxide
because using this technology eliminates safety issues as carbon dioxide is not
flammable and the oil is better quality (139), but no such plants have been con-
structed to process soybeans. This is due to the absence of a commercially feasible
means of continuously feeding soybean flakes into a high-pressure vessel and
removing the spent flakes. Recently, one company has developed a screw press in
which supercritical carbon dioxide is injected into the barrel. This equipment has been
successfully used to produce soybean meal with lower residual oil contents than typi-
cally produced by screw pressing and with little heat denaturation of the protein.
The spent flakes or collets are sent to a meal desolventizer-toaster (DT). Newer
equipment incorporates countercurrent steam usage. The Schumacher-type deso-
lventizer/toaster/dryer/cooler has become widely accepted in the soybean industry,
and, with this equipment, residual levels of hexane should be less that 500 ppm.
Both indirect and direct steam heating are used. Steam vapor and a modest vacuum
carry away the solvent vapors for condensing. Condensed solvent is recycled to the
extractor after separating water from the hexane. A desolventizer-toaster is a series
of trays through which the meal flows. Soybean meal is unique in that it must be
toasted to inactivate protease inhibitors (especially trypsin inhibitor) that would
reduce feed efficiency if not denatured and inactivated. Urease activity is used as
a measure of adequate heating. The toasted meal typically has low-protein solubi-
lity as measured by protein dispersibility index (typically 45 PDI). The meal is then
sent to a dryer-cooler to reduce the meal temperature for safe storage. The moisture
content should be about 12% and the residual fat content less than 1.5%. The free
extractable oil after extraction is less than 1.0%, but heating during desolventizing-
toasting frees some bound fat that previously was not extractable with hexane.
Overtoasting may reduce digestibility and nutritional value of the meal. The meal
is then ground with a hammer mill to produce meal with uniform particle size.
If dehulling is employed, as is typical for plants in the United States, the meal
will contain around 48% protein. Additionally, dehulling reduces the fiber content
of the meal by over 50%. In some plants, a portion of the soybean hulls may be
added back to the meal prior to grinding to adjust and precisely control meal protein
content. Livestock feeders are concerned about having uniform protein and fiber
contents in order to formulate minimum-cost feeds for maximum feed efficiency.
The meal is generally ground so that 95% passes a U.S. 10-mesh screen and a max-
imum of 3% to 6% passes through a U.S. 80-mesh screen.
Some plants divert part of their spent flake production away from a desolventi-
zer-toaster to a flash desolventizer, which is designed to produce white flakes with
high-protein solubility (PDI 70 90). White flakes are used as the starting material
for producing protein isolates or concentrates, which contain >90% and 65% pro-
tein, respectively, and are used as food ingredients.
Some soybean extraction plants also degum their oil before shipping to centra-
lized refineries. There is not sufficient market to make it profitable to recover all of
the soybean phosphatides and market them as soy lecithin. The gums are added
back to the meal in the toaster to evaporate the water. The gums contribute to
the metabolizable energy content of the meal and the soybean crusher can get
meal prices for crude phosphatides.
Quality standards and trading rules for solvent-extracted soybean meal and oil
are designated by the National Oilseed Processors Association and are available
at a website (145). Soybean products are remarkably uniform in their quality char-
acteristics compared with alternative sources of oil and meal.
Recently, a third process, known as extruding-expelling (or Express Systems as
trademarked by the equipment manufacturer), was developed (Figures 7 and 8)
(146, 147). In this process, a dry extruder, which generates heat solely through fric-
tion of the beans in the extruder, replaces steam generating and steam heating the
beans. The heated beans then go to a screw press and the rest of the process is
the same as in screw pressing. The plants typically process 550 MT/day.
Soybeans
CRACKING ASPIRATING Hulls

( optional)
Meats
EXTRUDING
EXPELLING Cake
Foots (screw pressing) MEAL GRINDING
SETTLING MEAL COOLING
POLISH FILTERING Foots Partially Defatted

Meal
Crude Oil
Figure 7. Process flow diagram for extruding-expelling soybeans.
600 SOYBEAN OIL
Figure 8. Photograph inside a modern extruding-expelling plant processing soybeans. (This

figure is available in full color at http://www.mrw.interscience.wiley.com/biofp.)
Approximately 70 extruding-expelling plants have been built over the past 10 years
for crushing soybeans. Usually, these plants are farmer-owned and provide meal to
nearby livestock feeders (148). The oil is sold to the large oil refineries, often at a
discount despite the oil being of excellent or superior quality because high costs are
incurred in handling small lots of oil. These plants are ideally suited to identity-
preserved processing. There are niche opportunities for these plants to market
certified organic or nonGMO soybean oil, for which there is a lucrative market
in some countries. Other opportunities reside with genetically enhanced soybean
oils and meals, such as low-linolenate, high- and low-saturates, and high-oleate
oils. This process has even been proposed for producing soybean products during
interplanetary exploration (149). NASA plans to grow soybean in space because
some missions, such as Mars exploration, cannot be supported without growing
food in space.
6. QUALITIES OF SOYBEAN OILS AND MEALS EXTRACTED

BY DIFFERENT METHODS
Wang and Johnson (150) compared the qualities of soybean oils and meals obtained
by the three processing methods. Soybean oil and meal samples were collected
at three times within a one-year period from 13 extruding-expelling plants, eight
QUALITIES OF SOYBEAN OILS AND MEALS EXTRACTED 601
TABLE 10. Quality Characteristics of Soybean Meals Produced by Different Oil-Extrac-

tion Processes.
Processing Method

Property Solvent Extraction Screw-Pressing Extruding-Expelling
Moisture, % 11.65 11.03 6.94

Residual oil1, % 1.2 6.3 7.2
Protein1, % 48.8 43.2 42.5
Urease, pH 0.04 0.03 0.07
Protein solubility in KOH, % 89.1 61.6 88.1
Protein dispersibility index 44.5 10.6 18.1
Rumen-bypass protein, % 36.0 48.1 37.6
Hunter L color 69.1 51.5 65.8
Trypsin inhibitor, mg/g 5.46 0.3 5.52
Trypsin inhibitor, TIU/g 5280 2000 12,250
1
Reported at 12% moisture basis.
solvent-extraction plants, and one continuous screw-press plant. Their results are
shown in Tables 10 and 11. Solvent extraction is by far the most efficient method
of recovering oil from soybeans, typically only about 1.2% residual oil is left in the
meal. Screw-pressing is slightly more efficient in recovering oil than is extruding-
expelling, leaving 6.3% oil in screw-pressed meal compared with a mean of 7.2%
for extruded-expelled meals. Most solvent-extraction plants dehull soybeans to pro-
duce soybean meal with 48% or more protein and carefully control the moisture
content at 12%. Solvent-extracted soybean meal is highly uniform, often much
more so than either screw-pressed or extruded-expelled meal. The high-protein
and low-fiber contents of solvent-extracted soybean meal are desired when feeding
poultry and swine, which consume 46% and 25% of the soybean meal produced,
respectively. Most extrusion-expelling and screw-press plants have not invested
in dehulling equipment, as their meal generally goes into feeding ruminant animals.
Protein dispersibility indices, a measure of protein denaturation that is used in
the food industry, are lower for extruded-expelled and screw-pressed meals. Protein
TABLE 11. Quality Characteristics of Soybean Oils Recovered by Different Processes.
Processing Method

Property Solvent Extraction Screw-Pressing Extruding-Expelling
FFA, % 0.31 0.33 0.21

Phosphorus, ppm 277 463 75
Tocopherols, ppm 1365 1217 1257
Moisture, % 0.08 0.05 0.08
PV, meq/kg 0.96 1.76 1.73
AOM stability, h 39.8 36.2 23.9
Lovibond color, red 11.1 17.5 10.2
602 SOYBEAN OIL
solubilities in potassium hydroxide solution, a measure of protein denaturation and

an indicator of overcooking that is used in the feed industry, are similar for
extruded-expelled and solvent-extracted meals, but higher than that of screw-
pressed meal (62%). Rumen-bypass protein values are higher for the screw-pressed
meals, indicating that more protein escapes the rumen and is not converted to
microbial protein that has a lower nutritive value than the original soybean protein.
All meals examined by Wang and Johnson (151), regardless of the processing meth-
od employed, had low-trypsin-inhibitor activity, which is important to proper pro-
tein digestion. Soybean trypsin inhibitors, especially in unheated soybeans, can
inhibit the protease enzymes trypsin and chymotrypsin, reducing protein hydrolysis
during digestion. There are two trypsin inhibitors in soybeans, Kunitz inhibitor and
Bowman-Birk inhibitor. The Kunitz inhibitor is relatively easily inactivated by
moist heat, comprises about 85% of the inhibitory activity, and acts only on trypsin;
the Bowman-Birk inhibitor is much more stable to heat (due to six disulfide cross
linkages) and acts on both trypsin and chymotrypsin. The activity of the enzyme
urease (easily measured as pH change) is often used as a quick and easy indicator
of adequate cooking. A valuable resource for characteristics of soybean meal is
http://www.stratsoy.uiuc.edu/epv/.
Oil properties vary considerably between different types of plants (Table 11) and
among plants of the same type and sampling times. The free fatty acid (FFA) con-
tent, a measure of hydrolytic degradation during seed storage and oil extraction, of
extruded-expelled oil is significantly lower than that of solvent-extracted oil, which
may be due to the rapid inactivation of lipases during extrusion. Screw-pressed
soybean oil typically contains 0.33% FFA, which is similar to that of typical sol-
vent-extracted oil. The amounts of phospholipids in the oils after settling are much
lower in extruded-expelled oil (75 ppm phosphorus) than in solvent-extracted oil
(277 ppm phosphorus). Screw-pressed oil has much higher phospholipid content
(463 ppm phosphorus) than does solvent-extracted oil. The phospholipid in
extruded-expelled oil is readily hydratable and easy to settle, which are attributed
to the rapid heat inactivation of the phospholipases. The tocopherol contents of
crude extruded-expelled oils are slightly lower than those of crude solvent-extracted
oil.
Peroxide values (PVs), a measure of primary lipid oxidation products, are sig-
nificantly higher for crude extruded-expelled oil than for crude solvent-extracted
oil, which is attributed to the high temperature used in extruding-expelling, the
long period typically allowed for oil cooling, or the often poor oil-storage condi-
tions and longer storage times at extruding-expelling plants. Oxidative stability,
as measured by the Active Oxygen Method (AOM), of extruded-expelled oil is sig-
nificantly lower than that of solvent-extracted oil, probably because of the higher
PV value and lower contents of phosphorus (phosphatides) and tocopherols in crude
extruded-expelled oil. The colors of extruded-expelled and solvent-extracted oils
are significantly different. Although solvent-extracted oil tends to be slightly darker
than extruded-expelled oil, screw-pressed oil is much darker in color than are the
other two types, probably because of the more severe heat treatment of the screw-
pressed oil before pressing.
SOY PROTEIN INGREDIENTS 603
7. SOY PROTEIN INGREDIENTS
Defatted soybean meal (white flakes) may be heated to produce a variety of solu-
bility and enzyme-activity characteristics, ground and sized to produce grits or
flour, and used as a food ingredient in bakery products, soymilk, and meat products.
A historical accounting of the development of these products was published by
Johnson et al. (151, 152). Soy flour may be relecithinated or refatted with refined,
bleached, and deodorized oil to achieve desirable functional properties. Soy flour
can also be texturized by using an extruder to produce meat-like products called
TVP (texturized vegetable protein) that are often used to extend ground meat.
Enzyme-active soy flour is used in bread at 0.5% of the wheat flour. Lipoxygenase
in the soy flour bleaches the carotenoids of wheat flour to produce a whiter crumb
and improves dough-mixing properties. White flakes may be processed into soy
protein isolates or concentrates (132, 153). Soy protein is poorly soluble in water
at pH 4.5, the isoelectric point, and highly soluble at pH >8.0. These solubility
characteristics can be used to isolate or concentrate soy protein.
Untoasted and flash-desolventized meal in which the protein is undenatured and
highly soluble (>70 PDI and preferably >90 PDI) is the preferred starting material
in manufacturing soy protein isolates. Under some conditions, extruded-expelled
meal can be used, but the yield of soy isolate is reduced. The meal is ground in
water adjusted to pH 8.0 with sodium hydroxide and centrifuged to remove insolu-
ble fiber. The soluble fraction is acidified to pH 4.5, and the protein precipitates.
The precipitated protein curd is separated from the soluble sugars by centrifuging.
The protein curd may be washed, neutralized, and spray-dried.
High protein solubility is not needed for protein concentrates and heating to
insolubilize the protein and facilitate extracting the solubles (mostly sugars) with
water is one way that has been used to prepare soy protein concentrates. Concen-
trates today, however, are normally made by extracting the sugars with either acid
(pH 4.5) or aqueous ethanol (6080%). Aqueous ethanol is most frequently used
because it produces the blandest product, but ethanol denatures the protein and
leaves the protein with reduced functional properties unless the product is refunc-
tionalized by jet cooking (154, 155) or by homogenizing under alkaline conditions
(156). Soy protein concentrate must contain >65% protein on a dry basis.
The soybean storage proteins glycinin and b-conglycinin, which often are recog-
nized in the older literature as 11S and 7S proteins, respectively, based on their
sedimentation during ultra centrifuging, comprise 6580% of the protein. Methods
have even been developed to separate soy protein into fractions rich in individual
proteins (157, 158). Some believe b-conglycinin has greater health benefits than
glycinin.
Soy protein isolates are used in dairy analogs (milk replacers and beverage pow-
ders), meat-pumping solutions, luncheon meats, and infant formulas, whereas soy
protein concentrates are used in dairy analogs (milk replacers, beverage powders,
cheeses, coffee whiteners, frozen desserts, whipped toppings), baked goods, and
meat products (156). These protein products are used for their functional properties
such as solubility, water absorption and binding, viscosity control, gelation,
604 SOYBEAN OIL
cohesion-adhesion, elasticity, emulsification, fat absorption and binding, foaming,

and color control. The solubility and thermal properties of these products were
recently compared by Lee et al. (159). Some products have high solubility even
though they were largely denatured.
Many health benefits have been attributed to soy protein products, either because
of the proteins or accompanying phytochemicals, such as isoflavones, saponins, etc.
There is a growing body of evidence that soy protein products may impact hyper-
tension and heart disease, osteoporosis and bone health, and certain cancers. The
perception of such nutritional benefits is driving an increased interest by food com-
panies in the incorporation of soy protein products. In October 1999, the U.S. Food
and Drug Administration (FDA) authorized a health claim for soy protein in cardi-
ovascular disease. U.S. food labeling laws now permit a statement on the label that
Diets low in saturated fat and cholesterol that include 25 grams of soy protein a
day may reduce the risk of heart disease. One serving of (name of food) provides
(list number) grams of soy protein. The health claim allowance is reported in the
Federal Register (160) and is posted on the FDA website (161).
8. BASIC PROCESSING OPERATIONS
As discussed in the previous section on soybean oil composition and Table 11,
crude soybean oil can contain phospholipids, free fatty acids, lipid oxidation
products, and unsaponificable matter, which includes chlorophyll and carotenoid
pigments, tocopherols, sterols, and hydrocarbons. Some of these components nega-
tively affect oil quality, and some may play positive roles in nutrition and function-
ality. The goal of oil refining is to remove the undesirable components so that a
bland, stable, and nutritious product can be obtained. The basic processing opera-
tions in oil refining are (1) degumming, (2) neutralization, (3) bleaching, (4) hydro-
genation, (5) deodorization, and (6) winterization or crystallization. These steps are
outlined in a flow chart as shown in Figure 9.
8.1. Degumming
Crude soybean oil contains a relatively high concentration of phospholipids com-
pared with other vegetable oils. Degumming is a process of removing these com-
ponents from crude soybean oil to improve its physical stability and facilitate
further refining. Phospholipids can lead to dark-colored oils and they can also serve
as precursors of off-flavor (162) compounds. Free fatty acids, pigments, and other
impurities are also partially removed by degumming. Soybean oil can also be neu-
tralized directly without degumming if gum or lecithin recovery is not desired. Con-
ventional belief holds that the loss of neutral oil in refining crude oil by direct
neutralization is less than the combined losses of degumming and caustic refining
of the degummed oil.
The quality of crude soybean oil influences the efficacy of degumming. Phos-
pholipids can exist in a hydratable form, which can be readily removed by addition
BASIC PROCESSING OPERATIONS 605
Crude Soybean Oil
Water
FILTERING Foots
GUMS HYDRATING CENTRIFUGING
Alkali NEUTRALIZING
Soapstock GUMS DRYING Moisture
CENTRIFUGING (free fatty acids,
phosphatides)
Water WASHING Lecithin
Wash-water
CENTRIFUGING (residual
soapstock)
VACUUM DRYING Moisture
Bleaching BLEACHING
Earth
Spent Bleaching
FILTERING Earth (color, residual
soapstock)
Steam DEODORIZING DISTILLATE CONDENSING
POLISH FILTERING Deodorizer Distillate

(off-flavor compounds, minor
volatiles, free fatty acids)
Salad & Cooking Oils
Figure 9. Diagram of conventional soybean oil refining.
of water, or in a nonhydratable form, which cannot be easily hydrated and removed.

The nonhydratable phospholipids are considered to be the calcium and magnesium
salts of phosphatidic acids, which are formed by enzymatic hydrolysis of the origi-
nal phospholipids. This degradation can result from seed damage during storage
and improper handling. List et al. (53) studied the factors promoting the formation
of nonhydratable phospholipids in soybeans and showed that they are promoted by
four interrelated factors: (1) moisture content of beans or flakes, (2) phospholipase
D activity, (3) heat applied to beans or flakes prior to and during extraction, and
(4) disruption of the cellular structure by cracking or flaking. These results suggest
that a nonhydratable-phosphatide formation can be minimized by control of the
moisture of beans or flakes entering the extraction process, inactivation of phospho-
lipase D, and optimizing the temperature during conditioning of cracked beans or
flakes. Normal quality soybean oil from the conventional solvent extraction
contains about 90% hydratable and 10% nonhydratable phospholipids. Phosphoric
or citric acid can be used as a pretreatment to achieve more complete removal of
nonhydratable phospholipids, but their presence in the gum will darken it and
reduce its quality. The total phospholipid content in crude soybean oils ranges
from 1.85% to 2.75% (19) and partially depends on the seed preparation and extrac-
tion methods employed. Use of an expander or the Alcon process to cook the flakes
prior to extraction will increase total phospholipids content in the crude oil and the
phosphatidylcholine percentage in the gum (163).
606 SOYBEAN OIL
Degumming can be achieved in a batch or continuous fashion. In batch degum-

ming, soft water at the same percentage as total phospholipid is added to oil heated
to 70 C and mixed thoroughly for 3060 min, followed by settling or centrifuging.
In continuous water degumming, heated oil is mixed with water by an in-line
proportioning and mixing system and the mixture is held in a retention vessel for
1530 min before centrifugation. The phosphorus content is typically lowered to
12170 ppm (164). A well-degummed soybean oil should contain less than
50 ppm of phosphorus, which is well below the 200 ppm level specified in the
National Oilseed Processors Association (165) trading rules for crude degummed
soybean oil. Degumming for physical refining, as opposed to alkali refining of soy-
bean oil, requires more complete removal of the phospholipids to prevent darkening
during fatty acid distillation. For more complete phospholipid removal, several
modified degumming methods can be employed (166, 167).
Recently, polymeric ultrafiltration membranes were used for degumming crude
soybean oil and removing phospholipids from the crude oil/hexane miscella (168).
Crude soybean oil also can be de-acidified by methanol extraction of the free fatty
acids and the extract separated into fatty acids and solvent by a membrane filter
(169). A surfactant-aided membrane degumming also has been applied to crude
soybean oil, and the degummed oil contained 2058 ppm of phosphorus (170).
Supercritical carbon dioxide extraction was shown to be an effective means of
degumming (171). In this process, soybean oil countercurrently contacted supercri-
tical carbon dioxide at 55 MPa and 75 C. The phosphorus content of the oil was
reduced from 620 ppm to less than 5 ppm. Ultrasonic degumming was also success-
fully used to reduce the gum content of soybean oil (172).
8.2. Neutralization
Neutralization is also referred to as de-acidification and alkali or caustic refining.
Neutralization is achieved by treating the soybean oil with aqueous alkaline solu-
tion (most commonly, sodium hydroxide) to neutralize the free fatty acids in a batch
or continuous system. The soap formed in the reaction also adsorbs natural pig-
ments, the gum and mucilaginous substances not removed by degumming. Natural
settling or centrifugation is used to remove the soap. Crude soybean oil also can be
netralized directly without degumming. When this is practiced, the oil commonly is
pretreated with 3001000 ppm of 75% phosphoric acid to facilitate removal of
phospholipids. The percentage of excess sodium hydroxide solution required for
crude oil is higher than that for degummed oil (173).
The quality changes, such as lipid oxidation and reduction of tocopherols and
phytosteols during neutralization, are considerable compared with the other proces-
sing steps as shown by Wang and Johnson (174), and also as presented in Table 12.
The further phospholipid removal (below 2 ppm phosphorus) also reduces the oxi-
dative stability of soybean oil (175) due to the antioxidant property of these phos-
pholipids.
One of the new developments in neutralization is the use of silica-based adsor-
bent to remove the residual soap instead of using water washing. Water usage and
TABLE 12. Effect of processing on content of tocopherols, sterols, and squalene

in soybean oil (25).
Tocopherols Sterols Squalene

Processing
Step ppm % Loss Ppm % Loss ppm % Loss
Crude 1132 3870 143

Degummed 1116 1.4 3730 3.6 142 0.7
Neutralized 997 11.9 3010 22.2 140 2.1
Bleached 863 23.8 3050 21.2 137 4.2
Deodorized 726 35.9 2620 32.3 89 37.8
waste generation is greatly reduced by this practice. Sodium silicate also was used
as a mild neutralizing agent to refine specialty oils (176). Its agglomerating ten-
dency allowed the removal of the soap by filtration, and its low alkalinity mini-
mized saponification of neutral oil and loss of minor nutrients. Other adsorbents,
such as magnesium silicate, also were shown to be effective in reducing free fatty
acids, as well as reducing primary and secondary oxidation products in the treated
oil (175, 177).
Physical refining or steam refining is a process similar to steam deodorization.
Steam distillation is typically used for oil with a high free-fatty acid content to
reduce the refining loss, which would be significant if caustic refining was used.
Acid-aided degumming produces soybean oil with very low phosphorus content
and makes the distillation of free fatty acids possible. Nevertheless, the relatively
difficult task of removing sufficient phospholipids from soybean oil has prevented
extensive use of this technique in the United States. Physical refining, however, has
virtually replaced caustic refining of palm oil in Malaysia.
8.3. Bleaching
Bleaching is a process designed not only to remove the oxidation-inducing pig-
ments such as chlorophylls, but more importantly to decompose the peroxides pro-
duced by oxidation into lower molecular weight carbonyl compounds that can be
removed by subsequent deodorization. Bleaching also removes other impurities
such as soap and metal ions. In soybean oil refining, color reduction occurs at
each step, nevertheless, the most significant reduction of chlorophylls occurs in
the bleaching step. Acid-activated bleaching clay is most effective in adsorbing
chlorophylls and decomposing peroxides, and it is commonly used for soybean
oil. The chlorophyll content in normal crude soybean oil (11.5 ppm) can be
reduced by 25% by alkali refining, and bleaching with acid earth further reduced
chlorophylls to 15 ppb (178) The subsequent hydrogenation and deodorization
remove or degrade red and yellow pigments more than chlorophyll, so incomplete
chlorophyll removal by bleaching will cause the refined oil to appear greenish. The
refined and bleached oil is particularly susceptible to oxidation and is less stable
than the crude, degummed, refined, or deodorized oils (178).
608 SOYBEAN OIL
The desired bleaching endpoint is typically zero peroxide, although a color spe-
cification is often used as an important measure. The amount of bleaching earth
should be adjusted based on the quality of oil to be bleached, and it usually ranges
from 0.3% to 0.6% for a typical soybean oil. Low contents of phosphorus (510
ppm P) and soap (1030 ppm) in the neutralized oil are essential to maximize
the bleaching effect. Successful bleaching can be achieved by atmospheric batch
bleaching, vacuum batch bleaching, or continuous vacuum bleaching at tempera-
tures between 100 C and 120 C for 2030 min. More details of soybean oil bleach-
ing are described by Erickson (179).
Recently, silica-based synthetic materials have been used in bleaching. The nat-
ural bleaching earth, fullers earth, a hydrated aluminum silicate, mostly has been
replaced by acid activated clays, which are sulfuric- or hydrochloric-acid-treated
bentonites or montmorillonites. Manufacturers continuously improve the quality
and develop new bleaching earths to meet the markets needs. Higher activity
and filterability are the main focuses of such development.
8.4. Hydrogenation
The high degree of unsaturation, particularly the relatively high content of linole-
nate, of soybean oil significantly limits its food applications because of low oxida-
tive stability. Hydrogenation is used to improve oxidative stability as well as to
increase the melting temperature of soybean oil. A great proportion of soybean
oil is hydrogenated to produce cooking oil, bakery/confectionery fats, and shorten-
ing.
When oil is treated with hydrogen gas in the presence of a catalyst (typically
nickel) and under appropriate agitation and temperature conditions, it becomes
more saturated and forms a semisolid or plastic fat that is suitable for many food
applications. Selectivity is a term used to describe the relative reaction rate of the
fatty acids from the more unsaturated to the more saturated forms. Perfect selectiv-
ity would provide sequential elimination of linolenate, linoleate, and then oleate. To
completely hydrogenate linolenate while minimizing changes in the other acyl
groups, a high ratio of the reaction rates of linolenate to linoleate compared with
linoleate to oleate is desirable. Generally, selectivity increases with temperature and
catalyst concentration and with decreases in hydrogen pressure and agitation rate
(180). The effect of pressure on hydrogenation selectivity of soybean oil was
reported by List et al. (181), who found that the linoleate-containing triacylglycer-
ols were reduced at a slower rates than the linolenate-containing triacylglycerols
under selective condition. At higher pressures (500 psi), the reaction was truly non-
selective; whereas at 50 psi, the reaction became selective. Impurities in soybean
oil, such as phosphorus, oxidation products, carotene, and metal ions can poison
the catalyst and cause slower hydrogenation (182). A particular limitation with
nickel catalyst is its low selectivity for linolenate over linoleate, and copper-con-
taining catalysts have greater selectivity for linolenate acid than the conventional
nickel catalysts (183). The use of copper catalyst can produce soybean oil that
has a low degree of hydrogenation (iodine value of 110115) but has less than
1% linolenate. However, copper catalysts are not as active as nickel catalysts; they
are also easily poisoned (184). Furthermore, any trace of residual copper in the fully
processed oil will promote lipid oxidation.
The most common tests for degree of hydrogenation are congeal point and the
iodine value as determined by refractive index. Refractive index is a valuable tool
for iodine values above 95, but when the oil is further hydrogenated, refractive
index becomes an inadequate measurement for melting prediction because
increased amount of trans-isomers results in harder oil than the refractive index
would indicate (185). For margarine or shortening, the solid fat index (SFI), as
determined by dilatometry, or solid fat content (SFC), determined by nuclear mag-
netic resonance, is the most appropriate method to measure the consistency of the
hydrogenated oil. These indices predict the workability and creaming ability at a
particular temperature.
Double-bond isomerization or trans-fatty acid formation is the most important
side-reaction that occurs during hydrogenation, and it has a strong impact on the
physical and possibly the nutritional properties of the products. Trans-double bonds
are thermodynamically a more favorable configuration than their cis-counterpart; so
trans-bonds are produced in significant quantities if the hydrogenation does not go
to completion. The trans-fatty acids have a much higher melting point than their
cis-isomers, therefore a fat product with considerable trans-acyl groups will have
an elevated melting point, which is desirable in shortening and margarine applica-
tions. A partially hydrogenated soybean oil can have at least 30 different one-, two-,
and three-double-bond isomers that will result in more than 4000 different triacyl-
glycerol molecules. This complexity allows the production of a great variety of oils,
margarines, and shortenings that have a wide range of physical and functional prop-
erties. However, the established relationship between trans-fat consumption and
health has prompted research to minimize trans-double formation in fats and oils.
Hydrogenation of soybean oil may be carried out in a batch or a continuous sys-
tem. In the United States, batch operations are typical. More comprehensive
reviews on hydrogenation and formulation can be found in Erickson and Erickson
(180), Hastert (186), and Kellens (187).
8.5. Deodorization
Deodorization is usually the last step in conventional oil processing. It is a steam-
stripping process in which good quality steam (13% of oil) generated from de-aer-
ated and properly treated feed water is injected into soybean oil under high tem-
perature (252266 C) and high vacuum (<6 mm Hg) to decompose peroxides
and vaporize the free fatty acids and odorous compounds. Deodorization relies
on the large differences in volatility between the triacylglycerols and other undesir-
able components under certain conditions. The musty and earthy odor produced
from bleaching and the hydrogenation odor and flavor are effectively removed by
deodorization. The free fatty acids, typically ranging from 0.1% to 0.5% in neutra-
lized oil and 0.5% to 5% in oil to be physically refined, are also reduced to below
0.03%, a value used as an indicator for deodorization efficiency. Zero peroxide
610 SOYBEAN OIL
value is another indicator for effective deodorization. Heat bleaching is achieved by

holding the oil for 1560 min at high temperature to ensure considerable decom-
position of carotenoid pigments.
During the deodorization process, many desirable reactions take place, but some
undesirable reactions, such as lipid hydrolysis, polymerization, and isomerization,
also occur. Therefore, the deodorization temperature is carefully controlled to
achieve optimum quality of the finished soybean oil product. The effect of refining
condition on trans-fatty acid content in refined vegetable oils was investigated by
Okamoto et al. (188). Trans-fatty acid contents of deodorized oils increased with
prolonged exposure to high temperature, and trans-formation was higher in
oils containing greater proportions of polyunsaturates. The isomerization rate of
linolenate was 6.5- to 16.3-fold higher than that of linoleate in soybean oil. Kemeny
et al. (189) studied kinetics of the formation of trans-linoleic acid and trans-
linolenic acid in vegetable oils deodorized at temperatures from 204230 C for
286 h. Their data can give good estimates of the trans-level of refined oils for
given deodorization conditions. Deodorization has also been modified to retain
more nutrients and prevent other undesirable reactions. Mathematical models
have been established describing the influence of different process parameters
such as time, temperature, steam rate, and pressure on tocopherol stripping, produc-
tion of oxidized and polymeric triacylglycerols, and trans-fatty acid formation dur-
ing physical refining of soybean oil (190). Tocopherol removal was mainly
influenced by processing temperature and steam rate, whereas oxidized and poly-
merized triacylglycerols were not significantly affected by any of the investigated
process parameters.
There are three types of deodorization operations. The batch process is the least
common because of its low efficiency and inconsistent product quality. The
semicontinuous and continuous deodorizers have improved processing efficiency.
There are several configurations of the continuous deodorizer, including the
single-shell cylindrical vessel type, the vertically stacked-tray type, and the thin-
film packed-column type. The thin-film system provides excellent fatty acid
stripping with minimum use of steam, but it does not achieve the desired heat
bleaching or effective deodorization because of its relatively short retention time.
A retention vessel held at high temperature has to be used after the column distillation to
achieve bleaching (191).
The overall oil quality change during refining of soybean oil was examined by
Jung et al. (178), and their results are shown in Table 13. A study of oxidative
TABLE 13. Effect of Processing Steps on Quality of Soybean Oil (178).
Phosphorus Iron Chlorophyll Peroxide Value Tocopherol Free Fatty

Refining Step (ppm) (ppm) (ppm) (meq/kg) (ppm) Acid (%)
Crude 510 2.9 0.30 2.4 1670 0.74

Degummed 120 0.8 not available 10.5 1579 0.36
Refined 5 0.6 0.23 8.8 1546 0.02
Bleached 1 0.3 0.08 16.5 1467 0.03
Deodorized 1 0.3 0.00 0.0 1138 0.02
ALTERNATIVE REFINING METHODS 611
stability of soybean oil at different stages of refining indicated that crude oil was the
most stable and highly purified oil was the least stable (192). The influence of the
refining steps on the distribution of free and esterified phytosterols in soybean and
other oils was reported by Verleyen et al (193). A significant reduction in free ster-
ols was found after neutralization. Deodorization removed free sterols and also pro-
moted steryl ester formation when the oil was physically refined due to a heat-
promoted esterification reaction between free sterols and free fatty acids.
8.6. Fractionation and Winterization

Fractionation or winterization is a process in which the more saturated molecular
species in the oil are solidified and removed by a low-temperature treatment, which
increases the cold storage physical stability of the oil. Partially hydrogenated soy-
bean oil with 110115 iodine value (IV) that is intended for use as salad and cook-
ing oil should be fractionated. By doing so, the more saturated molecules and some
high-melting trans-isomers are removed to produce clear oil that meets low-tem-
perature storage requirements. The formation of large and easily filterable crystals
and the removal of the crystallized fraction from the liquid oil can be challenging
tasks. The temperature of the oil should be lowered slowly to prevent small crystal
formation. Nucleation occurs when the oil is supercooled to a temperature that is
much lower than the thermodynamic equilibrium temperature. Heterogeneous
nucleation, i.e., the formation of nuclei on to foreign substances, typically takes
place around dust particles or on the walls of the crystallizer. The crystal growth
rate depends on the degree of supercooling and polymorphic form. In order to
have continuous and uniform crystallization, an intense but nondestructive agitation
is required. To produce salad oil with good cold stability, soybean oil is usually
hydrogenated to an iodine value of 100110 (linolenate content of 23%) and win-
terized at 23 C. To produce a cooking and frying oil, hydrogenation to an iodine
value less than 90 (linolenate content of less than 0.5%) is more desirable, and the
stearine fraction obtained from winterization of such oil is a good shortening and
margarine base. Crystal separation can be done by filtering, centrifuging, or decan-
tating. More details about these systems are presented by Krishnamurthy and
Kellens (194).
9. ALTERNATIVE REFINING METHODS
Although oil extraction by mechanical pressing of soybeans accounts for a very

small percentage of soybean processing, it is used by many farm cooperatives or
family-owned on-farm operations in the United States, primarily for using protein
meals as animal feed. There is an increasing use of extrusion-expelling technology
to produce identity-preserved soybean oil and protein products for niche market.
The advantages of small tonnage requirement, no flammable solvent used, low initi-
al capital investment, and unique products have made this processing technology
very appealing for many soybean growers and processors.
612 SOYBEAN OIL
Alternative techniques are being developed for refining soybean oil produced by
mechanical means. Simple refining methods were explored to process extruded-
expelled (E-E) soybean oils with various fatty acid compositions (174, 177). E-E
oils can be easily water degummed to very low phosphorus levels. Free fatty
acid content was reduced to 0.04% by adsorption treatment with Magnesol1, a
commercial magnesium silicate product from Dallas Group of America (Jefferson-
ville, IN). This material also adsorbed primary and secondary oil oxidation pro-
ducts. A mild steam deodorization as the last processing step produced good-
quality soybean oil. This adsorption refining procedure was much milder than con-
ventional refining, as indicated by little formation of primary and secondary lipid
oxidation products and less loss of tocopherol during refining.
10. COPRODUCTS AND UTILIZATION
10.1. Lecithin
Soybean lecithin is the predominant source of food and pharmaceutical lecithin
because of its availability and outstanding functionality. The composition of crude
soy lecithin is shown in Table 14. As a result of the presence of a large amount of
neutral oil, crude lecithin is usually de-oiled to improve its functionality. De-oiling
is based on the solubility difference of neutral and polar lipids in acetone, in which
the phospholipids are precipitated and separated. Alcohol fractionation of de-oiled
lecithin can further separate lecithin into an alcohol-soluble fraction that is enriched
with phosphatidylcholine and an alcohol-insoluble fraction enriched with phospha-
TABLE 14. Composition of Commercial Soy Lecithin in Comparison with Egg Lecithin,
wt % (195).
Compounds Soy Lecithin Egg Lecithin
Phosphatidylcholine 1015 6570

Phosphatidylethanolamine 912 913
Phosphatidylinositol 810
Phosphatidylserine 12
Phosphatidic acid 23
Lysophosphatidylcholine 12 24
Lysophosphatidylethanolamine 12 24
Phytoglycolipids 47
Phytosterines 0.52.0
Other phosphorus-containing lipids 58
Sphingomyelin 23
Carbohydrate 23
Free fatty acids max 1 max 1
Mono-, diacylglycerols max 1 Trace
Water max 1.5 max 1.5
Triacylglycerols 3540 1015
COPRODUCTS AND UTILIZATION 613
TABLE 15. Typical Composition (%) of Commercially Refined Soy Lecithin

Products (196).
Lecithin Lecithin Lecithin

Oil-Free Alcohol-Soluble Alcohol-Insoluble
Phosphatidylcholine 29 60 4
Phosphatidylethanolamine 29 30 29
Phosphatidylinositol and glycolipid 32 2 55
Neutral oil 3 4 4
Others 7 4 8
Emulsion type favored w/o or o/w o/w w/o
tidylinositol. The phosphatidylcholine-enriched fraction is an excellent oil-in-water

emulsifier, and the phosphoinositol-enriched fraction is a good water-in-oil emulsi-
fier that is often used in the chocolate industry. The typical composition of de-oiled
and fractionated lecithin products is shown in Table 15.
Supercritical CO2 extraction also has been used to selectively extract phospha-
tidylcholine from de-oiled soybean lecithin (197). The effects of temperature, pres-
sure, and amount of ethanol on phosphatidylcholine extraction were examined, and
a high-purity product could be produced with optimized conditions.
Lecithin recovered from solvent-extracted soybean oil had different phospho-
lipid class compositions from those produced by mechanical pressing (198). The
percentage of phosphatidylcholine was considerably higher in lecithin recovered
from extruded-expelled oil than from solvent-extracted oil. The phosphatidylcho-
line- and phosphatidylinositol-enriched fractions produced by ethanol extraction
of the crude lecithin also showed different functional properties (199).
Soybean lecithins can be chemically altered to modify their emulsifying proper-
ties and improve their dispersibility in aqueous systems. Phospholipids may be
hydrolyzed by acid, base, or enzyme to achieve better hydrophilic and emulsifica-
tion properties. Hydroxylation of lecithin improves its oil-in-water emulsification
property and water dispersibility. Acetylation creates improved fluidity and emul-
sification, water dispersion properties, and heat stability (200).
10.2. Deodorizer Distillate

Deodorizer distillate is the material collected from the steam distillation of oils. It is
a mixture of free fatty acids (especially during physical refining) tocopherols, phy-
tosterols and their esters, hydrocarbons, and lipid oxidation products. The quality
and composition of deodorizer distillate depends on the feedstock oil composition
and processing conditions. Tocopherols and sterols are the most valuable compo-
nents that can be recovered from the distillate, and they are used in the nutrition
supplement and pharmaceutical industries (201). Typical soybean deodorizer distil-
late contains about 33% unsaponifiable matters, of which 11% is tocopherol and
18% sterol (202).
614 SOYBEAN OIL
Soybean tocopherols are the major source of natural fat-soluble antioxidants and
Vitamin E. The Vitamin E activity of natural d-a-tocopherol is much greater that
that of synthetic Vitamin E, which is a mixture of eight stereoisomers (203). Phy-
tosterols are used as raw materials for over 75% of the worlds steroid production.
The more recent application of phytosterol, phytostanol, and their fatty acid esters
in margarine and table spreads is based on the blood cholesterol-lowering effect of
these compounds (204, 205). The recent development of functional foods containing phy-
tosterols has been reviewed by Hollingsworth (206) and Hicks and Moreau (207).
The preparation of high-purity tocopherols and phytosterols involves steps such
as molecular distillation, adduct formation, liquid-liquid extraction, supercritical
fluid extraction, saponification, and chromatography (175). The extraction of toco-
pherols from soybean oil deodorizer distillate by urea inclusion and saponification
of free fatty acids resulted in good recovery of tocopherols (208). To improve the
separation of sterols and tocopherols, Shimada et al. (209) used a lipase to esterify
sterols with free fatty acids. Then the steryl esters and tocopherols were separated
better by molecular distillation. Chang et al. (210) used supercritical fluid CO2
extraction to recover tocopherols and sterols from soybean oil deodorizer distillate.
A patent by Sumner et al. (211) advocated treatment of the distillate with methanol
to converted free fatty acids and other fatty acid esters to methyl esters that can then
be removed by a stripping operation. Then separation of sterols and tocopherols
could be carried out by molecular distillation.
10.3. Soapstock
Soap is recovered from alkaline neutralization of the crude or degummed soybean
oil. Soap consists of water, free fatty acids, neutral oil, phospholipids, unsaponifi-
able matter, proteins, and mucilaginous substances. Its composition depends on
seed quality and oil extraction and refining conditions. Soapstock is the least valu-
able byproduct from oil processing, and it is generated at a rate of about 6% of the
volume of crude soybean oil refined (212), amounting to as much as 0.8 million MT
in the United States annually. The majority of the soap or acidulated soap is used as
a feed ingredient contributing metabolizable energy. Soybean oil can be refined
using potassium hydroxide and acidulated with sulfuric acid, followed by neutrali-
zation with ammonia rather than sodium hydroxide to produce a fertilizer (213).
Soybean oil methyl esters can also be produced from soapstock (214218) for bio-
diesel applications.
11. FOOD AND BIOBASED PRODUCT USES OF SOYBEAN OIL
11.1. Distribution of Soybean Oil Utilization

In 20012002, when 8.32 million MT (18,300 million pounds) of soybean oil was
used in the United States, over 97% (8.09 million MT, 17,800 million pounds) was
used for food, with the remainder used in nonfood products (219). Among the food
FOOD AND BIOBASED PRODUCT USES OF SOYBEAN 615
uses, about 48% (3.89 million MT, 8,570 million pounds) was for shortening, 43%
(3.58 million MT, 7,897 million pounds) for cooking and salad oils, 7% (0.56 mil-
lion MT, 1,237 million pounds) for margarine, and 1% (0.06 million MT, 125 mil-
lion pounds) for other food uses. Soybean oil is used to produce about 95% of the
total margarine and 83% of the total shortening consumed in the United States.
Among the 0.24 million MT (519 million pounds) used in nonfood products,
about 16% (0.04 million MT, 85 million pounds) was for resins and plastics,
12% (0.03 million MT, 60 million pounds) for paint and varnish, 13% (0.03 million
MT, 68 million pounds) for fatty acids, and 59% (0.14 million MT, 306 million
pounds) for a myriad of other inedible uses. The use of soybean oil in lubricants
(220), oleochemicals (221), and bioplastics (222), and the production of methyl
soyate for environmentally friendly solvents (223, 224) and for blending with diesel
fuel to produce biodiesel (20% methyl soyate/80% diesel fuel) (225) are significant
parts of the soy oil used in nonfood applications (226). Usage of soybean oil to
make biodiesel is likely to increase in future years because several new plants
are planned for construction as a result of the recent Farm Bill of 2002 providing
financial incentives for producing biobiesel. Some states, notably Minnesota, have
enacted legislation that provides biodiesel tax incentives. Biodiesel interests have
become organized as the National Biodiesel Board (Jefferson City, MO) and the
Renewable Fuels Association (Washington, DC), and exercise considerable politi-
cal influence. During 2002, 57 million liters (15 million gal) of biodiesel were pro-
duced in the United States (227), almost three times that which was produced in
2001.
The usage of soybean oil in food products is similar to other oils, and these uses
and products are discussed in more detail for all oils in other chapters of this edi-
tion. This chapter will focus on specifics of soybean oil in those uses. The major
products in which soybean oil is consumed are cooking and salad oils, frying oils
and fats, baking shortenings, and margarine. Only minor amounts of soybean oil are
used in vegetable dairy products and confectionery products.
11.2. Trading Rules for Crude and Refined Soybean Oils

As the U.S. government does not have trading rules, the National Oilseed Proces-
sors Association (NOPA, Washington, D.C.) has established them, including quality
specifications, to facilitate trade and marketing of three types of oils: crude
degummed, once-refined, and fully refined soybean oils (Table 16). These rules
are also available on the Internet (228). Factors that impact grade of crude
degummed and once-refined soybean oils are moisture and volatile matter content,
flash point, free fatty acid content, smoke point, unsaponifiable matter content,
green color, phosphorus content, and refined bleached color. The flash point reflects
the presence of residual hexane, and the other factors reflect expected refining loss.
For fully refined soybean oils, the flavor, cold test values, peroxide value, and AOM
(Active Oxygen Method) are additional considerations that reflect crystallization
at low temperatures and stability to oxidation. Crude soybean oil is sold as
degummed oil because the gums tend to spontaneously hydrate and settle out during
616 SOYBEAN OIL
TABLE 16. Trading Specifications for Crude Degummed, Once-Refined and Fully Refined
Soybean Oils (228).
Methods of
Factor Crude Degummed Once-Refineda Fully Refineda Analysisb
Moisture and volatile 0.3 max.c 0.10 max. 0.10 max.d Ca 2d-25
matter and
insoluble impurities (%) (up to 0.15 with (up to 0.15 with Ca 3a-46
discount) discount)
Flash point ( C) 121 min. 121 min. Cc 9c-95
Free fatty acids 0.75 max. 0.10 max. 0.05 max. Ca 5a-40
(% as oleic)
(up to 1.25 with (up to 0.15 with
discount) discount)
Unsaponifiable 1.5 max. 1.5 max. 1.5 max. Ca 6a-40
matter (%)
Presence of fish and Neg. 28.121
marine animal oils
Phosphorus (%) 0.02 max. Ca 12-55
(up to 0.025 with
discount)
Refined bleached color 3.5 Red max. 20 Yellow, 2.0 Cc 8e-63
(Lovibond) Red, max. Cc 13b-45
Green color None
Flavor Bland
Cold test (hr) 5.5 min. Cc 11-53
Peroxide value 2.0 max. Cd 8-53
(meg/kg)
AOM Stability 8 min. Cd 12-57
(hr to 35 PV)
a
The oil shall be clear and brilliant in appearance at 2129 C (7085 F) and free from settlings in this
temperature range.
b
Analyses in accordance with the Official and Tentative Methods of the American Oil Chemists Society except
for presence of fish and marine animal oils in accordance with Association of Official Analytical Chemists
methods.
c
Includes insoluble impurities as determined by AOCS Method Ca 3-46.
d
Oil shall be free of settlings or foreign matter of any kind.
transportation and storage, which cause numerous handling problems. Once-refined

soybean oil is seldom traded anymore because most buyers do their own refining or
purchase fully refined oil. End-users typically have their own specifications for fully
refined soybean oil and use the NOPA values as bases for their more stringent spe-
cifications (136).
11.3. Cooking and Salad Oils

In most parts of the world, both cooking and salad oils from soybeans are refined to
have bland taste and light color. For other oils, distinct flavors and dark colors may
be acceptable. Important distinctions between salad oils, cooking oils, and frying
oils, however, reflect their differences in oxidative and thermal stabilities. Cooking
and frying oils need to be more stable to oxidation than salad oil because of the
higher temperatures to which cooking oils are exposed. Temperature stability is
especially required in fats and oils used in deep-fat frying. Salad oils must be phy-
sically stable so that they do not crystalize at refrigerated temperatures.
As soybean oil contains relatively great amounts of the polyunsaturates, notably
unstable linoleate (61%) and linolenate (7.8%), partial hydrogenation is customary
to make cooking or salad oils more stable to oxidation. Typical specifications for
different cooking and salad oils are shown in Table 17.
Synthetic antioxidants, such as butylated hydroxyanisole (BHA), butylated
hydroxytoluene (BHT), propyl gallate (PG), ascorbyl palmitate, and tertiary-butyl-
hydroquinone (TBHQ), are used in soybean cooking oils and frying fats (230).
These antioxidants are typically added at 0.01% for one antioxidant and 0.02% total
for two or more. Natural antioxidants, derived from sage, rosemary, and green tea,
are increasingly popular because of consumer preferences for natural food ingredi-
ents (231).
Salad oils differ from cooking oils in their tolerance to cold temperatures with-
out crystallizing. Salad oils must not crystallize, cloud, or leave deposits of any kind
when stored at refrigerator temperatures (4.4 C) and are defined as such. Soybean
oil used as a salad oil should not cloud or produce any visible crystals and remain
brilliant and clear for a minimum of 5.5 hr at 0 C. Fully refined soybean oil can be
directly used as salad oil because it will normally meet this specification, whereas
other oils, such as sunflower and corn, must be dewaxed before they can meet typi-
cal salad oil specifications. Soybean oil may be partially hydrogenated and then
winterized to achieve greater oxidative stability and still not crystallize nor lose
proper emulsion properties when refrigerated, although most of the soybean oil
used in commercial dressings is not hydrogenated.
New nutrition-oriented salad and cooking oils have been developed in recent
years. LoSatSoy is an oil low in saturated fatty acids that was developed at Iowa
State University, licensed to Pioneer Hybrid International (Johnston, IA), and com-
mercialized as a salad or cooking oil. This specialty soybean oil has one-half the
amount of saturated fatty acids in normal soybean oil (7% versus 15%); therefore, it
is promoted as having improved nutritional and health benefits.
Other specialty soybean oils, low (<2% or <1%) in linolenate and with
improved oxidative stabilities in salad and cooking oil applications, are comparable
with typical soybean oil that is partially hydrogenated. Today, low-linolenic-acid
soybean oil is an attractive alternative to hydrogenated oil that contains trans-fatty
acids. Beginning in 2006, labeled food products must disclose both the grams of saturated
fat and grams of trans-fat per serving (232). This is inducing food companies to
eliminate or significantly reduce trans-fatty acid contents of their products.
All specialty soybean oils require identity-preserved soybean production, crush-
ing, and refining systems. As financial incentives are needed all along the produc-
tion process to compensate for increased costs of identity preservation, specialty
soybean oils command premium consumer prices and have been slow to impact
soybean oil markets.
618 SOYBEAN OIL
TABLE 17. Trading Specifications for Soybean Cooking and Salad Oils (229).
Cooking and Salad Oil

Factor Refined, Fully, Fully, HWb Analytical
Deodorized Refineda Refineda Soybean Methodc
Source of specifica- Fedd,e NSPA ASCSf Fedd

tions
Moisture and volatile 0.06 max. 0.10 max. 0.10 max. 0.06 max. Ca 2d-25
matter (max) (%) (0.14 with
discount)
Unsaponifiable con- 1.5 max. Ca 6a-40
tent (%)
Flash point, C 228 min. Cc 9b-55
Free fatty acids (wt%) 0.05 max.g,h 0.05 max. 0.05 max. 0.05 max.g,h Ca 5a-40
as oleic
Red color (Lovibond) 4 max. 2.0 max. 2.0 (2.6 with 2.0 max. Cc 8b-52
discount)
Cc 8e-63
Cc 13b-45
Yellow color (Lovi- 35 max. 20 max. 20 max. 20 max. Cc 8b-52
bond)
Cc 8e-63
Cc 13b-45
Peroxide value (meg/ 1.0 max.h 2.0 max. 0.5 max. 1.0 max.h Cd 8-53
kg) (1.0 with
discount)
Fat stability by AOM
methodi
(a) Peroxide value 35 max. 35 max. Cd 1257
after 8 hr
(b) Peroxide value 15 min.h 25 min.h Cd 12-57
of 100 or less at
indicated no. of hr
Cold test (hr)
Free from sediment Yesj Yes Yes Yesj Ca 3-46
and foreign matter of
any kind
Clear and brilliant at Yes Yes Yes Yes
2129 C
k
Fish oil and marine Neg.
animal oil test
Iodine value 105115 Cd 1-25
Linolenic acid (wt%) 3.0 max. by Cd 7-58
or 3.5 max. Cd 1-62
by
Odor and flavor 1 1 1 1 m

Additives n o p n, q
/preservative
Permitted/required
a
Typically a refined, bleached, and deodorized oil.
b
Refined, bleached, partially hydrogenated, winterized, and deodorized, pure soybean oil.
c
Analyses in accordance with the Official and Tentative Methods of the American Oil Chemists Society
Champaign, Illinois, unless indicated otherwise.
d
Federal specifications No. JJJ-S-30G dated March 24, 1978, issued by U.S. General Services Administration,
Washington, D.C.
e
The salad oil may contain properly refined and deodorized cottonseed, corn, peanut, soybean, sesame,
sunflower, or safflower vegetable oils or a mixture of these oils. Olive oil shall not be used. Edible vegetable
oils not specified may also be used provided they are in accordance with good commercial practice.
f
Specifications per announcement PV-501 dated June 17, 1976, issued by Agricultural Stabilization and
Conservation Service, U.S. Department of Agriculture, Shawnee Mission, Kansas.
g
0.05% will be acceptable if propyl gallate has been added as an antioxidant or as a component in an
antioxidant.
h
Determination will be made within 7 days after packaging each lot.
i
Active oxygen method.
j
Exclusive of particles of resinous flux material from can manufacture.
k
Association of Official Analytical Chemists Method No. 28.107.
l
The oil after heating shall be bland and free from beany, rancid, painty, musty, soapy, fishy, metallic, and other
undesirable or foreign flavors and odors when tested by the method prescribed in footnote m within 7 days
after packaging each lot.
m
Approximately 50 g of the finished product shall be placed in a clean 150-mL Pyrex glass beaker and heated
to a temperature 177 3 C. The oil shall be examined for odor at this temperature, and for flavor, each
cooling to approximately 38 C. From Federal Specification JJJ-S-30G.
n
Heavy metal scavengers, antifoaming agents, and antioxidant materials may be added to improve the
keeping quality and use performance of the oils. The ASCA specifications also permit the addition of
oxystearin. Such additives should be of a kind and at levels permitted in edible oil products under the federal
Food, Drug, and Cosmetic Act and regulations promulgated thereunder.
o
Preservatives generally recognized as safe are permitted.
p
During the cooling stage of deodorization, 0.005% of citric acid or 0.006% of monoisopropyl citrate shall be
added to the oil.
q
The packaging gas shall be of food-grade quality and may consist of pure nitrogen or a mixture of nitrogen
and approximately 10% of carbon dioxide plus other inert gases in the atmosphere, but it shall contain no
more than 0.005% oxygen. Maximum permissible oxygen content of the headspace gas within 15 min after
the oil is packaged is 0.50% as measured at standard temperature and pressure. Measurement shall be made
at time of packaging or within 15 min thereafter. For method of analysis, see Bulletin 916, issued in 1963 by
American Dry Milk Institute, Chicago, IL.
11.4. Frying Oils and Fats

In addition to its use as a common household cooking oil, soybean oil is used
widely in home and commercial deep-fat frying procedures. The popularity of fried
foods among U.S. consumers has created a large market for stable frying oils and
for fast-food establishments. Typical untreated cooking and salad oils, including
soybean oil, are not suitable for frying applications because they oxidize too
quickly. Thus, the oils must be altered to make them stable to the frying treatment.
Heat treatments, such as commercial and household frying, accelerate autoxida-
tion. The heat itself causes oxidation and breakdown of the fat. In addition, when
620 SOYBEAN OIL
fats are heated in the presence of moisture, as often is the case in food applications,
fatty acids are released via hydrolysis of the ester linkages (233). The free fatty
acids, in turn, can accelerate oxidation of the oil. Decomposition and condensation
of hydroperoxides also produces a multitude of nonvolatile monomeric products,
including di- and tri-oxygenated esters, and dimeric and polymeric materials, espe-
cially at elevated temperature. Many of these dimers and polymers are known to be
rich sources of volatile carbonyl compounds and decrease the flavor and oxidative
stability of soybean oil (234). These high-molecular-weight materials also can pro-
duce a series of physical and chemical changes to the oil and food products, includ-
ing increased viscosity, polarity, free-fatty acid content, development of dark color,
and an increased tendency of the oil to foam (233).
A typical soybean oil shortening is generally hydrogenated to enhance its
stability, making it suitable for frying procedures. In addition, polydimethylsiloxane
is routinely added at a level of 0.022 ppm as an antifoaming agent, which greatly
extends the frying life of soybean oil (235). The antioxidants mentioned in the
subsection on Cooking and Salad Oils provide oil stability prior to frying and
can enhance the oxidative stability of the fried food. Even though most antioxidants
are volatile at frying temperatures, with their concentration decreasing during
frying, some antioxidant is transferred to and retained in the food (carry through),
thus providing antioxidant protection in the food during storage. In tests,
heated palm olein with no frying lost 70% of its original BHT and 60% of the
original BHA after 8 hr (236). TBHQ being the highest molecular weight (lowest
volatility) of the typical antioxidnts, provides the greatest carry-through benefit
(237).
Extensive hydrogenation produces flaked fats or shortening-like products for fry-
ing applications, which offer convenience in filling fryers and excellent frying sta-
bility. Unfortunately, the process of hydrogenation creates trans-fatty acids as
byproducts of the reaction As noted elsewhere in this chapter, recent concerns about
the presence of trans-fatty acids in our diets, and the subsequent new labeling
requirements for trans-fatty acids (232), have prompted food manufacturers and
oil producers to explore alternative treatments to create soybean oil that is stable
to frying.
One procedure to increase stability without creating trans-fatty acids involves
adding a small amount of a fully hydrogenated oil (hardstock) to a typical soybean
oil. The blended oil is then interesterified to create a stable frying oil without
trans-fatty acids. In a recent study, the low-linolenate soybean oil noted in the
subsection on Cooking and Salad Oils, when blended with 5% of a soybean oil
hardstock, was as stable as a traditional trans-fat-containing soybean oil that
had been stabilized for deep-fat frying, while still retaining excellent flavor
characteristics (238). Another approach to enhance frying stability of soybean
oils is to increase the oleate concentration in the soybean oil created by the
plant, either through traditional plant breeding or biotechnological methods. The
resulting oil, however, when used in frying, creates a fried food with a stale,
waxy-like flavor that lacks the desirable flavor components typical of a fried
food (239, 240).
11.5. Mayonnaise and Salad Dressing
In the United States, mayonnaise, salad dressing, and French dressing are defined
by Standards of Identity issued by the U.S. Food and Drug Administration (FDA;
Code of Federal Regulations, Section 21, 169.140) (241). The Food, Drug and Cos-
metic Act of 1930 and later revisions and amendments were promulgated to prevent
adulteration and misrepresentation of certain food products by establishing
Standards of Identity.
Mayonnaise is defined as a semisolid food prepared with not less than 65% vege-
table oil, and egg yolk and vinegar. Most mayonnaise in the United States, however,
contains 7582% oil, to get the proper texture (242). Soybean oil is usually used in
mayonnaise but winterized cottonseed, corn, and canola and hydrogenated soybean
oil also can be used. Mayonnaise is an oil-in-water emulsion with oil droplets mea-
suring 12 mm in diameter. The higher the oil content, the more tightly the oil dro-
plets are packed in the continuous water phase and thus, the greater the viscosity
and rigidity. Mayonnaise production is partly an art because of the difficulty of pro-
ducing an oil-in-water emulsion in which the dispersed phase has seven times more
volume than the continuous phase. The protein in the egg yolk solids is the only
emulsifier allowed and processing conditions play critical roles in achieving
high-quality and high-stability mayonnaise.
Salad dressings are also oil-in-water emulsions and were developed as alterna-
tives to mayonnaise. The Standard of Identity (21 CFR, 1699.150) requires that sal-
ad dressings contain not less than 30% vegetable oil (but most contain 3550% oil),
vinegar, 4% egg yolk, and starch. For texture and viscosity, salad dressings rely
on starch, in contrast to mayonnaise, which depends on greater oil content. The oils
used in salad dressings are selected using the same criteria for mayonnaise.
The qualities of mayonnaise and salad dressing are determined by the physical
and oxidative stability of its lipid components. Phase separation or emulsion
breakdown is caused by mechanical shock, agitation, extreme temperatures, or
fat crystallization. Oxidation of vegetable oil and egg lipid also can occur. As the
quality of oil plays a major role in the flavor stability of these products, only the
best quality salad oil should be used. It is particularly important to use salad oils
with long cloud point times (high cold test hours). If fat crystals form during storage
at refrigerated temperatures, the emulsion will break and the product will become
unsightly with visible free oil. Crystal inhibitors, such as oxystearine, lecithin, and
polyglycerol esters, are allowed to prevent crystallization and emulsion breakdown.
Although mayonnaise and salad dressings are spoonable products due to their
high viscosity, French dressing is a pourable oil-in-water dressing. French dressing
must contain 35% oil as defined by a Standard of Identity (21 CFR, 169.115). Egg
products are optional. Other dressings, such as Thousand Island, are not subject to
Standards of Identity, and any ingredients can be used. Pourable dressings can be in
two different finished forms; emulsion or two phases depending on whether the pro-
duct is homogenized. The oil used in these products is predominantly soybean salad
oil in the United States. In Canada and Europe, other salad oils are often used,
depending on the availability and costs of those vegetable oils in each specific region.
622 SOYBEAN OIL
As the oil contents of mayonnaise, salad dressings, and French dressing are high,
it is important to prepare them from salad oils that taste bland and are relatively
stable to oxidation. Peroxide values of the oil should be <2 meq/kg. Even early
stages of oxidation can be detected in mayonnaise and salad dressings as grassy
and beany flavors. Packaging with an inert headspace is important to prevent
oxidation during distribution, retailing, and consumer storage. Storage under refrig-
eration is important once the package is opened and the headspace gas becomes
replaced with air.
11.6. Margarine
Margarine was first produced in 1869 by the French chemist Hippolyte Megge
Mouries. During the Franco-Prussian War, he was awarded a prize and patent for
his invention of a butter substitute. It was not until the 1940s, however, that mar-
garine became widely used. Until then, the powerful dairy industry in the United
States prevented the sale of colored margarine in many states, and consumers did
not readily accept white table margarine. Today, more than twice as much margar-
ine is consumed as butter per capita in the United States, and margarine is no longer
considered a cheap imitation of butter. Unlike butter, margarine can be formulated
from a variety of fats and oils to give a variety of physical and functional properties,
which are needed in many food applications today.
In the United States, margarine or oleomargarine is also controlled by an FDA
Standard of Identity (21 CFR, 166.110), requiring at least 80% fat. Soybean oil is
predominantly used in the United States, followed by cottonseed and corn oils. The
other 20% of the margarine formulation may be made up of water and other
optional ingredients, including milk products, soy protein isolate, salt, selected
emulsifiers (up to 0.5%), mold inhibitors, antioxidants, color additives, flavorings,
and acidulants. Margarine is a water-in-oil emulsion.
The traditional retail form of margarine is stick margarine, but margarine is now
also marketed as pourable and soft tub products. Margarine may also be sold as a
whipped product in which air or an inert gas is incorporated. Still other margarine-
like forms, including polyunsaturated and low-fat spreads, have been developed to
satisfy consumer demands for improved convenience and reduced saturated fat and
calories. In addition to the traditional use as a table food, margarine is also widely
used in baking applications such as in cookies and as roll-in fats for puff and Danish
pastries.
A significant recent consumer trend is increased demand for margarine-like
spreads that are not controlled by a Standard of Identity and that contain much
less fat. Most spreads contain 4060% fat with 40% fat spreads being more popular
in Europe and 60% in the United States. During the past 15 years, however, very
low-fat spreads containing less than 20% fat have been introduced. As a result of
these trends, there are significantly fewer 80%-fat margarine products available in
the United States today than in the previous decades. Stabilizing these high levels of
aqueous phases in such a small amount of fat as the continuous phase requires spe-
cial equipment to generate the necessary shear and higher amounts of emulsifiers.
Moustafa (243) reports that the aqueous droplets must no longer be spherical but
rather polyhedral when loading levels of the aqueous phase exceed 74%.
Margarine processing includes blending the fats separately from the aqueous
phase ingredients and water, dispersing and emulsifying the aqueous phase within
the fat phase, chilling to solidify the fats, pin working the solidified mass, resting,
forming, and packaging. The ingredients are emulsified before being fed into a
swept-surface heat exchanger for crystallizing. The mass emerging from the cool-
ing tubes is partially solidified, and it is further crystallized in the working unit. The
texture of the product is further modified in the resting tube before the margarine is
packaged.
Margarine and shortening have fat crystal networks in which liquid oil is
entrained. As a result, they exhibit a yield stress that must be exceeded before
the product begins to flow as a viscous fluid. The yield stress is related to spread-
ability. The rheological properties of margarine have been discussed by Segura
et al. (244).
In North America, margarines may be composed of blends of hydrogenated soy-
bean oil and palm oil, partially hydrogenated soybean oil and cottonseed oil, liquid
soybean oil and partially hydrogenated soybean oil, liquid corn oil and hydroge-
nated corn oil, or simply hydrogenated soybean oil. Most oil blends contain high
levels of soybean oil to keep costs competitive. Table 18 shows some typical com-
positions and properties of margarine.
The most important functional properties of margarines and spreads are spread-
ability and hardness, oiliness, and melting characteristics. These properties relate to
fat level, proportion of solid fat, fat melting point, and crystal form. Diverse tex-
tures and functionalities can be achieved by varying the extent of hydrogenation.
Consistency and emulsion stability depend on the amount and type of crystallized
fat. Spreadability and hardness can be predicted by the solid fat index and penetra-
tion measurements. A cone penetrometer is typically used to determine margarine
hardness (245). Typical margarines should be spreadable at refrigeration tempera-
tures, remain semisolid at ambient temperatures, and melt at less than body
temperature. Oil-off refers to the separation of liquid fat when the fat crystals no
longer form a network able to hold the liquid oil.
TABLE 18. Compositions and Properties of Hydrogenated and Interesterified Soybean

Margarine Oils (187).
Melting Point
Trans IV
Soybean Oil Type 10 C 21.1 C 33.3 C ( C) (%) (calc)
Hydrogenated Stick margarine 28.6 18.9 5.3 46 31.0 92.1

Hydrogenated Tub margarine 15.6 8.8 1.3 46 23.2 108.0
Hydrogenated Tub margarine 7.1 4.5 2.0 46 12.9 121.8
Interesterified 90:10a 1.7 1.3 0.2 40 1.7 123.8
624 SOYBEAN OIL
Fats exhibit polymorphism in which they can exist in different crystalline forms
depending on how the triacylglycerols pack in the crystal and a, b0 , and b poly-
morphs are known. The preferred polymorphic form for margarine is b0 , which
gives a smooth, pleasing mouthfeel and proper spreadability. Despite hydrogenated
soybean oils tendency to form b crystals, it is used in over 90% of all margarines
and table spreads in the United States. The less heterogeneous the fatty acid com-
position of the hydrogenated fats, the more it is b tending. Hydrogenated fats richer
in trans-isomers are less b tending and tend to produce margarines with smoother
textures. Blending small amounts of b0 -tending base fats (palm and cottonseed
oils) or different soybean base oils increase fatty acid heterogeneity favoring b0
crystal stability. Blending unmodified oils with oils that have been hydrogenated
to various degrees allows the production of margarines with desirable texture.
The greater the number of base stocks available, the greater the flexibility to pro-
duce a wide range of products and the higher the tolerance to processing conditions.
Different procedures for designing good margarine from various base stocks were
evaluated by Cho et al. (246).
Base oils for margarine must be hydrogenated to achieve the desired solid-fat
content with the consequential isomerization of some fatty acids. The new regula-
tions requiring reporting of trans-fats content on labels may dissuade some consu-
mers from using traditional margarine. Emken (247) reported that some traditional
margarines may have as much as 21% trans-fatty acids while Kellens (187) found
as much as 31%, and DSouza et al. (248) reported that the high-melting acylgly-
cerols contained in hydrogenated base stocks used for formulating North American
margarines have 33.145.0% trans-fatty acid content in stick margarine and 22.4
30.1% trans-fatty acid content in soft margarine. Trans-acyl groups contribute to
the firmness of margarine. A recent comprehensive review concluded that consum-
ing more than 4% of total calories as trans-fatty acids may raise plasma lipid levels
(249) and may cause heart disease (250, 251).
Some companies are producing low-trans- or zero trans-margarines by random
(252) or directed interesterification of mixtures of unhydrogenated and fully hydro-
genated soybean oils and other fats (253). To produce these products, a liquid oil
and completely hydrogenated hardstock are interesterified, so that proper plasticity
can be obtained. Oils that contain considerable amounts of palmitic acid favorably
influence crystallization and polymorphic form of the interesterified fat blends
(254).
Chemical interesterification is conveniently achieved by using alkali metal
methylates as a catalyst. Microbial lipases are also used as biocatalysts in enzy-
matic interesterification. In contrast to the chemical process, the enzymatic process
can be more selective if an enzyme with positional specificity is used, but this
reaction is usually much slower and more sensitive to reaction conditions. Recent
developments in lipase-catalyzed interesterification have resulted in new industrial
applications of this process (255). Nevertheless, the high costs of enzymes and pro-
cess equipment may limit widespread adoption of this process.
In developing trans-free fat, various methods for laboratory-scale, pilot plant,
and commercial batch reaction were described by Erickson (256). List et al.
TABLE 19. Example of Combined Hydrogenation, Interesterification, and Fractionation

to Produce Low Trans-Margarine Fat (187).
Solid Fat Content (%, at C)

Melting
Iodine Value Point ( C) 10 20 30 40
Soybean oil (SBO) feedstock 134 7 0

Fully hydrogenated SBO (FHSBO) 1 71 95 94 94 93
Blending SBO and FHSBO (60:40) 81 63 44 42 39 35
Random interesterification of SBO 81 53 38 33 20 11
and FHSBO (60:40)
Fractionation of the interesterified oil
Soft fraction 91 24 25 1 0 0
Hard fraction 63 58 60 58 45 32
(252) developed a zero-trans margarine by interesterifying 80% refined, bleached,

and deodorized (RBD) soybean oil with 20% fully hydrogenated soybean oil. The
resulting product has a solid fat index comparable with that of conventional pro-
ducts. The randomly interesterified low- [zero-] trans-soybean margarines crystal-
lize in the more favorable b0 crystal form (252) but tend to crystallize slowly after
chilling and result in a product that is harder than desired (257). Addition of 20%
liquid soybean oil to the interesterified oil yielded a softer, more desirable product.
Table 19 presents a typical example of the combined use of hydrogenation, inter-
esterification, and fractionation to produce low-trans fats with physical properties
comparable with partially hydrogenated soybean oil with high trans content.
Alternatively, recent research has focused on soybeans bred for high contents of
saturated fatty acids, some with as much as 43% saturates, 23% palmitate, and
20% stearate compared with the normal 15 % saturates, 11% palmitate, and
4% stearate. Soybeans only produce cis-fatty acids and, thus, there are no sources
of trans-fatty acids in the blends. List et al. (258, 259) showed that soybean oil from
soybeans bred to produce 3040% saturates was not sufficiently solid to make good
margarine, but soybean oil with elevated saturated fatty acid contents (1738%)
could be blended with high-melting oils, such as palm oil, interesterified palm
oil, interesterified palm and soybean oils, and cottonseed and soybean hardstocks,
to make a good margarine. Kok et al. (260) used blends (50:50) of traditional soy-
bean oil and oil from soybeans bred to produce oil high in saturated fat (43% satu-
rates, 23% palmitate, 20% stearate). The blend was then interesterified to produce
oil that was made into soft tub margarines. The small differences in sensory proper-
ties observed in comparisons with other tub margarines indicated the interesterified
product should be quite acceptable to most consumers. List et al. (259) also report
randomly interesterifying (randomizing) neat soybean oil high in saturated fatty
acids (10% palmitate, 18% stearate) gave good margarine without graininess
(SFI values of 58 at 10 C, 23 at 21.1 C, and 12 at 33.3 C).
626 SOYBEAN OIL
11.7. Shortenings
Shortenings are fats of vegetable or animal origin used in baking, but the term
shortening also has been accepted as a term to describe semisolid fats for frying
and cooking. Just as in margarine, the solid fat exists as a tight network of small
crystals, which trap liquid oil. Plastic shortenings differ from margarine in that
shortening is not an emulsion; it is all lipid material and may contain emulsifiers.
Prior to the development of hydrogenation, lard and tallow were the principle short-
ening fats, but these fats lack the diversity of texture and functionality required for
many products. Today, most shortenings contain at least some soybean oil, largely
because it is the least expensive oil that can confer adequate functionality. Short-
ening is available in many forms: plastic and semisolid (cubed, sheeted, and
printed), pourable fluid (with suspended solids), encapsulated powder, and flaked.
Most plastic shortenings are produced by blending oils with hydrogenated fats and
often emulsifiers and solidifying or crystallizing and plasticizing the blend. The
shortening is packaged and tempered by holding it in a quiescent state for several
days at 30 C. During solidification, 1025% air is often incorporated to improve the
color and texture. Pourable and fluid shortenings are produced by blending appro-
priate oils and emulsifiers. They are crystallized by cooling the fluid mass and stir-
ring the suspended crystals for 46 hr at precise temperatures so that large crystals
do not develop, and the fluid becomes stabilized.
Shortenings are added to baked goods to shorten or tenderize them by interrupt-
ing the gluten structure. Shortenings improve mouthfeel and eating qualities, add
lubricity, improve dough-handling properties, contribute flavor and structure, and
promote desirable crumb grain and texture (261). Shortening and tenderizing
effects are especially important in cakes, piecrusts, pastries, cookies, and crackers.
Generally, solid fat indices that change little with temperature are desired for most
shortening applications. Table 20 shows plasticity and melting properties of differ-
ent commercial shortenings. Typical shortening levels are 25% in bread, 525% in
cake, 2030% in sweet goods, 3040% in puff pastry, and 2035% in piecrusts.
Many plastic shortenings are packaged in 50-lb polyethylene-lined boxes, pri-
marily for use in retail bakeries, e.g., in grocery stores. These are difficult to handle
in large, automated wholesale bakeries. Sometimes, 190-kg drums are used, but are
still difficult to manage and use in the bakery where large amounts are needed.
Pourable and pumpable fluid shortenings were developed to avoid these problems
and are based on soybean oil. However, liquid oils do not cream and aerate well.
The addition of small amounts of hardfats, known as stearine, and various emulsi-
fiers can impart good functional properties to the liquid shortening.
Although adequate quality bread and rolls can be produced without shortening
by using the sponge-and-dough or straight-dough methods, the inclusion of short-
ening increases volume by as much as 25% compared with breads with no short-
ening. This volume increase often is referred to as oven spring, and it reduces
firmness throughout the products storage life. The largest volume of bread is
made by the continuous-mixing method in the United States and shortening is cri-
tical to good quality bread manufactured when using this method. Shortening
TABLE 20. Typical Compositions and Properties of Baking Shortenings.
Solid Fat Index Melting Point

Type Composition 10 C 21.1 C 26.7 C 33.3 C 37.8 C ( C)
Cookie and pie dough shortening Partially hydrogenated soybean and palm oils 2630 1822 1620 1215 9.513 4648
(unemulsified)
Cake and icing shortening Partially hydrogenated soybean and cottonseed oils 2327 1619 1518 1215 912 4850
(mono and diglycerides)
Yeast-raised sweet goods Partially hydrogenated soybean and palm oils (mono and 2429 1418 912 4447
diglycerides)
Fluid cake shortening Partially hydrogenated soybean oil (mono and diglycer-
ides, triglycerol monostearate, sodium stearoyl 2-
lactylate)
High volume cream filling and Partially hydrogenated soybean and palm oils (mono and 2528 1922 1821 1417 1114 4749
icing diglycerides, polysorbate 60)
Biscuit shortening Partially hydrogenated soybean and palm oils 2530 1620 7.5 4447
(unemulsified) 11.5
Roll-in margarine for yeast-raised Partially hydrogenated soybean and palm oils (mono and 2530 1519 69 4142
sweet goods diglycerides)
Fluid bread shortening Partially hydrogenated soybean oil (mono and
diglycerides)
Fluid bread shortening Partially hydrogenated soybean oil (mono and
diglycerides, sodium stearoyl 2-lactylate, ethoxylated
mono and diglycerides)
628 SOYBEAN OIL
delays starch gelatinization and allows the dough to expand more before the struc-
ture is set. Maximum loaf volume, which is a desirable trait in the United States, is
achieved with 6% of emulsified shortening, based on flour weight, but, in practice,
35% is normally used. Hardfats in bread shortenings are important in reducing col-
lapse of the loafs sidewall. At least 4% hydrogenated lard stearine is desired in
many bread shortenings. Refined, bleached, and deodorized soybean oil is used
in most commercial white pan breads.
Bread shortenings should crystallize in the b form. The base fat of a typical plas-
tic bread shortening is comprised of 90% partially hydrogenated soybean oil (70
IV) and 10% lard stearine (<5 IV); whereas the base fat of a typical fluid bread
shortening is comprised of 95% partially hydrogenated soybean oil (95 IV) and
5% lard stearine. Mono-and diglycerols, are added to reduce staling rate and
more functional emulsifiers, such as sodium steroyl-2-lactylate or ethoxylated or
succinylated mono-and diglycerols, are added as dough conditioners to impart
greater mixing tolerance to enable the bread to withstand abuse without loss of
loaf volume (262).
Using emulsified shortening in layer cakes, cake doughnuts, and muffins
increases volume and reduces air cell size and produces a fine internal grain.
Creaming is defined as the mixing of the shortening over wheat flour particles
and incorporating of air nuclei into the fat. The air nuclei can become sites for
gas bubble formation, which is important in cakemaking. The large number of min-
ute air bubbles incorporated into shortening improves the leavening in baked goods.
For the shortening used in cakes and icings, small (1 mm) needle-like b0 crystals
are preferred to the larger (515 mm) b crystals because the b0 shortenings appear
smooth, provide good aeration, and have better creaming properties (263).
Typically, partially hydrogenated soybean oil is blended with cottonseed or palm
oil hardstock to obtain b0 crystals. Most cake shortenings contain mono- and digly-
cerides to decrease the size of entrained air cells during creaming, to produce finer
air cells and grain in the cake crumb, and obtain a larger volume per unit weight of
batter (specific volume). To achieve proper aeration of fluid cake shortenings, how-
ever, partially hydrogenated soybean oil with b-tending soybean hardstock is
balanced with a-tending emulsifiers, which are typically mono- and di-glycerides
and glyceryl-lacto fatty esters.
Generally, plastic baking shortenings should be firm and plastic, but not brittle or
too soft and oily. Hardfat is added to soybean oil to achieve proper texture, plasti-
city, and creaming properties. Plastic shortenings should be soft and plastic at low
temperatures and still remain semisolid at body temperature.
Soybean oil is excellent for preparing hydrogenated base stocks from which a
wide array of shortenings is made. Up to 50% soybean hardfats are blended with
partially hydrogenated soybean oil in some shortenings. Soybean hardfats, however,
crystallize in the b polymorph unless blended with an equal or greater amount of b0
hardfat, such as hydrogenated palm or cottonseed oil. Partial hydrogenation of the
base soybean oil improves the oxidative stability of the shortening. The amount of
hardstock is varied to achieve the desired texture for the specific product applica-
tion. Various kinds of baked goods need varied shortening functionalities and
OXIDATIVE QUALITY OF SOYBEAN OIL 629
plasticities to produce optimum quality. Plasticity is controlled by achieving the

proper solid fat content or solid fat index. Typical plastic shortenings should
have a relatively flat solid fat index, with solids content in the range of 15% to
30% over the temperature range of 15 C to 32 C (264). One means of getting these
properties is blending 10% hardstock from two sources to get the proper crystal
structure with 90% partially hydrogenated soybean oil (IV 6580).
11.8. Confectionery and Imitation Dairy Products, and Low-Calorie Fat

Substitutes
Very little soybean oil is used to manufacture the hard butters used in confectionery
products or imitation dairy products. For imitation chocolate, enrobing fats, coffee
whiteners, whipped toppings, imitation cheese, frozen desserts, and filled milk,
coconut and palm kernel oils are preferred because of their sharp melting points.
It is important in these applications to have very low solids at body temperature
to prevent a waxy mouthfeel. A few fractionated specialty blends of hydrogenated
soybean oil and hydrogenated cottonseed oils (265) or soybean oil that has been
hydrogenated by using sulfur-treated nickel catalysts to achieve high selectivity
(266) occasionally may be used. These fats, however, are also high in trans-fatty
acids (>40%) and new trans-fat labeling requirements discourage their use. The
advantages of imitation dairy and chocolate products are improved functionality
compared with natural products. Thus, freeze-thaw stability in whipped toppings
and melting properties can be customized for specific applications (267).
As a result of widespread concern about weight control, the production of lipid
materials with reduced or zero calories has been of special interest recently. The
lipid-based fat replacers are esters that resist enzymatic hydrolysis, are poorly
absorbed, have relatively low-energy content, or have different modes of metabo-
lism. Many of these materials can be made from soybean oil or contain soybean oil
fatty acids. Sucrose polyester or other synthetic esters and diacylglycerol oils are
examples of these low-calorie fat substitutes (268274).
12. OXIDATIVE QUALITY OF SOYBEAN OIL
The oxidative stability of soybean oil is affected by its composition, handling of

beans prior to extraction, processing conditions, and additives. Important composi-
tional factors in soybean oil stability include its fatty acid composition and the pre-
sence of free fatty acids, phospholipids, natural antioxidants, and pigments (275).
Important handling and processing factors include excessive bean moisture,
damage, and temperature; exposure to oxygen; contamination by pro-oxidant
metals; and exposure to light (276).
12.1. Flavor Reversion

Soybean oil has poor oxidative stability, which is a major problem for the soybean
industry. Crude soybean oil has a characteristic green-beany flavor, which is
630 SOYBEAN OIL
eliminated during refining, bleaching, and deodorization, to produce a bland-tast-

ing, light-colored oil. During storage, however, refined soybean oil develops a char-
acteristic flavor that often is called reversion flavor (277). Prior to the 1940s,
some believed that soybean oil reverted to its unrefined flavor after being refined
and deodorized. Soybean oil was considered extremely light sensitive, and it was
believed to revert if one carried the freshly deodorized oil past the light of a north
window. This reversion was not considered an oxidative phenomenon (278). Actu-
ally, the term reversion is a misnomer, because (1) soybean oil does not revert to
its original crude-oil flavor, (2) the effect of light is real but was greatly exagger-
ated, and (3) the off-flavor development is indeed an oxidative reaction (278). Pro-
cedures available for following oxidation prior to the 1940s involved an iodometric
titration to obtain a peroxide value, but this method was too insensitive to measure
the low degree of oxidation that could be detected in soybean oil by sensory exam-
ination. With the support of more sensitive methods, we now know that upon oxi-
dation, soybean oil develops beany and grassy flavors at the early stages (i.e.,
peroxide value 10 or below), rancidity at higher levels of oxidation (peroxide value
of 10 or more), and fishy or painty flavors at the more advanced stages. These
flavor deterioration characteristics are common to all unsaturated oils containing
significant amounts of linolenate (279). It is now widely accepted that flavor dete-
rioration of soybean oil is an oxidative phenomenon, and that linolenate is the most
important precursor of flavor reversion of soybean oil.
The technology to handle soybean oils off-flavor was discovered by an interest-
ing set of circumstances. Near the end of World War II, Warren Goss, who was
commissioned to learn the secrets of the German oilseed industry, found that a
Dr. Tassusky and his daughter Ilona had patented a process involving multiple
washes of crude soybean oil with water or sodium silicate solution and the addition
of 0.01% citric acid to the deodorizer (278). This process worked, not because of
the washings, but because of the addition of citric acid. Now we know that trace
metals accelerate flavor deterioration and that treatment with citric acid or other
metal deactivators is a practical and effective means of improving flavor stability
(274).
12.2. Studies on Oil Oxidation

Extensive work has been done to clarify the mechanism of oil oxidation. It is a free-
radical chain reaction catalyzed by light, heat, and metals, in which molecular oxy-
gen reacts with unsaturated fatty acids to produce hydroperoxides. (280). An impor-
tant factor in initiating the oxidation of unsaturated fats is by exposure to light in
the presence of oxygen and a sensitizer. The activation of ordinary triplet oxygen in
this way forms singlet oxygen, which reacts readily with unsaturated fatty acids
(281). Oxygen is quite soluble in soybean oils (282), which frequently contain nat-
ural photosensitizers, such as chlorophylls or pheophytins. Singlet oxygen readily
reacts with the double bonds of unsaturated fatty acids; for example, singlet oxygen
reacts with methyl linoleate at a rate of at least 1500 times faster than normal triplet
oxygen (282). Once oxidation is initiated by singlet oxygen, the hydroperoxides
that result can decompose to yield free radicals, and the reaction mode quickly
becomes autocatalytic in the presence of triplet oxygen. A study by Carlsson et
al. (283) found that the photo-oxidation of various unsaturated vegetable oils was
not retarded by known free-radical scavengers, but was retarded by compounds
known to quench singlet oxygen. Furthermore, the degree of retardation apparently
paralleled the singlet oxygen-quenching ability of these compounds.
Commonly, the fatty acids in food lipids are exposed to heat during oil
processing and food manufacture. Once peroxides are formed, they can decompose
and generate free radicals, and the rate of peroxide decomposition increases with
temperature. Such reactions are of extreme importance to both consumers and
processors, because of their flavor significance, and under frying conditions
they can affect the physical, nutritional, and toxological properties of the fried
food.
Enzymes native to plants and animals can initiate oxidation reactions. The most
important and best known of these enzymes is lipoxygenase (linoleate:oxygen oxi-
doreductase, E.C. 1.13.11.12) (LOX) (284, 285). Enzymatic oxidations in plant sys-
tems are mediated by lipoxygenases that use molecular oxygen to catalyze the
oxidation of lipids containing a cis, cis-1,4-pentadiene moiety, such as linoleate
and linolenate. The reaction leads to the formation of hydroperoxides, giving the
same isomers as those formed during autoxidation of linoleate and linolenate. Soy-
beans are a rich source of lipoxygenase isozymes known as LOX-1, LOX-2, and
LOX-3, and their activity is associated with the development of off-flavors, espe-
cially green-beany flavors, in soybean products (285).
Monohydroperoxides are the primary products of lipid oxidation. A variety of
hydroperoxides with positional and geometrical isomers are formed depending
on the position and number of double bonds of the unsaturated fatty acids and
the oxidation mechanism. A number of reviews have been published on the com-
position of isomeric hydroperoxides formed from oxidation of oleate, linoleate, and
linolenate (286, 287291). The hydroperoxides formed are odorless, but they are
relatively unstable and are the precursors of a variety of volatile and nonvolatile
scission products that are important to the oxidized flavor.
Secondary volatile scission products from primary hydroperoxide decomposition
include aldehydes, alkanes, alkenes, alkynes, alcohols, and hydrocarbons. There are
considerable differences, however, in the flavor significance of these volatile com-
pounds. When estimating the impact of volatile oxidation products on flavor, it is
necessary to know not only their relative concentration, but also their relative
threshold values. One way of evaluating flavor impact is to divide the concentration
by the threshold concentration, although the relative flavor impact may change with
absolute concentration (292). Also, interactions among flavor compounds in the
olfactory response may be important. The relative volatility also may play a role
if a compound must be in the gas phase to reach the olfactory organ. Lee et al.
(293) created equations to relate the flavor impact of individual volatiles, dispersed
in an oil-water emulsion, to a specific concentration of 2-heptanone (Table 21). By
this method, in a fresh and oxidized soybean oil, nonanal contributed the greatest
individual effect on the flavor intensity, followed by trans, trans- and trans,
632 SOYBEAN OIL
TABLE 21. Concentrations (ppb in emulsion) of 2-Heptanone Perceived to Have

the Same Flavor Intensity as the Components Isolated from Commercial Soybean
Oil Oxidized at 35 C Under Fluorescent Light for up to 11 Days (293).
Day

Component 0 4 7 11
1-Penten-3-one 0.21 0.46 1.14 1.00

Pentanal 0.27 0.27 0.34 0.35
t-2-pentenal 0.27 0.18 0.24 0.35
Toluene 0.80 0.44 0.37 1.24
Hexanal 2.57 4.67 6.21 6.42
Heptanal 12.35 15.50 17.47 16.19
t-2-Heptenal 5.37 16.02 28.55 38.86
1-Octen-3-one 1.58 1.86 2.43 2.47
1-Octen-3-ol 1.41 3.98 8.74 12.60
t,c-2,4-Heptadienal 17.09 29.30 42.66 48.05
2-Pentylfuran 3.54 4.51 5.26 5.52
t,t-2,4-Heptadienal 20.80 34.27 45.10 48.18
2-Octenal 5.14 8.07 10.82 12.09
Nonanal 76.58 108.18 113.70 101.90
t,c-2,4-Decadienal none none 15.62 26.24
t,t-2,4-Decadienal none none 25.09 43.30
Total 148.0 227.8 323.8 364.8
cis-2,4-heptadienal, and 2-heptenal. Hexanal produced a large GC peak, but its

effect on flavor intensity was relatively small. More recently, Kao et al. (294)
suggested that particles formed in the oral cavity could transport entrained triacyl-
glycerols to the olfactory epithelium, allowing the triacylglycerols themselves to
impart flavor, thus implying that compounds in oxidized soybean oil do not need
to be volatile to contribute to flavor. They noted that the nutty flavor of fresh
soybean oil could only be observed when the lips were parted or the tongue drawn
away from the palate, both being conditions that generated particles. Liu and
Hammond (295) did further work to support the hypothesis that oral particles
strongly influence flavor perception of ketones typically found in oxidized soybean
oils and of flavor compounds in other foods.
Numerous studies have shown that the off-flavor intensity of soybean oil is cor-
related with its concentration of linolenate. Although the concentrations of both
linoleate and linolenate, which can reach 6065% in typical soybean oil, undoubt-
edly contribute to soybean oils instability, it is not clear why the much smaller
amount of linolenate has such a strong effect on soybean oil flavor. Linolenate is
expected to oxidize about twice as fast as linoleate, but there is seven to eight times
more linoleate than linolenate in typical soybean oil. The flavor compounds pro-
duced by linolenate do not seem to have much lower thresholds than those produced
from linoleate. Possibly flavor interactions in olfaction may account for these
effects.
12.3. Control/Stabilization Measures

Selective hydrogenation to lower the concentrations of linolenate or linolenate and
linoleate has been practiced to improve the oxidative stability of soybean oil. The
linolenate concentration of soybean oil also can be altered by mutation breeding
and genetic engineering (296).
Autoxidation can be inhibited or retarded by adding low concentrations of chain-
breaking antioxidants that interfere with either chain propagation or initiation
(286). Chain-breaking antioxidants include phenolic and aromatic compounds hin-
dered with bulky alkyl substituents. Common synthetic chain-breaking antioxidants
used in food lipids include butylated hydroxyanisole (BHA), butylated hydroxyto-
luene (BHT), tert-butylhydroquinone (TBHQ), and propyl gallate (PG). This class
of antioxidants react with peroxy free radicals to terminate reaction chains. The
antioxidant radical (A) formed in Equation 5 should be relatively stable and unable
to initiate or propagate the oxidation chain reaction.
ROO AH ! ROOH A 5
The phenolic antioxidants achieve stability by forming resonance hybrids (Figure 10)
(297). A radical intermediate, such as semiquinone, can undergo a variety of
reactions, including dismutation, to form a stable quinone and can regenerate the
original hydroquinone (Figure 11). However, these antioxidants generally lose their
efficiency at elevated temperatures, and they are most effective during the induction
period. Once the antioxidant is consumed, oxidation accelerates (297).
Preventive antioxidants reduce the rate of the chain initiation. The most impor-
tant initiation suppressors are metal deactivators that chelate metal ions. Metal
deactivators used for stabilizing edible fat and lipid-containing foods include citric,
phosphoric, tartaric acids, and phospholipids. Peroxide destroyers also are preventive
OH O O O
ROO + ROOH +
OH O OH OH
Figure 10. The formation of resonance hybrids by phenolic antioxidants.
O O OH O
+ +
OH OH OH O
Figure 11. This dismutation of a semiquinone radical intermediate.

634 SOYBEAN OIL
antioxidants; for example, the sulfur compounds, phosphates, and phosphines

reduce hydroperoxides to more stable alcohols (286).
Ultraviolet light deactivators can prevent oxidation by absorbing irradiation
without the formation of radicals. Examples include pigments such as carbon black,
phenyl salicylate, and a-hydroxybenzophenone. A significant synergistic antioxida-
tive effect can be achieved when chain-breaking and preventive antioxidants are
used together, because they suppress both initiation and propagation. The synergis-
tic effect of common antioxidants in combination with metal inactivators in foods
has been known for some time (33). Loliger (298) showed that the tertiary antiox-
idant system of Vitamin E, Vitamin C, and phospholipids provided the best protec-
tion against oxidative degradation when compared with the two antioxidants used
alone or in combination.
Light deterioration is also an important factor in the storage stability of soybean
oils. Refining and bleaching remove not only natural photosensitizers, but also sing-
let oxygen quenchers such as carotenoids. The restoration of the removed carote-
noids may protect lipids effectively against singlet oxygen deterioration, but the
resulting yellow coloration may be objectionable to consumers. Another approach
to protecting stored oils from light is the use of a package or container that absorbs
the light necessary for photosensitization or that prevents light from reaching the
oil.
Avoiding metal contamination is also very important, as metals such as copper
and iron are strong pro-oxidants for soybean oil. Copper or iron-containing alloys,
except stainless steel, should never be used for equipment involved in direct contact
with soybean oil. Soybean oil may be stored in containers made from carbon steel
that is coated on the interior with an epoxy or polyurethane lacquer, in stainless
steel, or in fiberglass-reinforced polyester.
Displacement of oxygen in container headspaces by nitrogen or carbon dioxide
to 2% has been shown to reduce oxidation effectively in vegetable oil (299).
Therefore, nitrogen or other inert gas protection should be considered whenever
the oil is to be stored for an extended period or held in the hot, liquid state.
12.4. Evaluation of Finished Oil Quality

Regardless of the official specifications for soybean oil and its products, the ulti-
mate proof of the pudding is in the eating; that is, sensory evaluation of the odors
and flavors of soybean oil and its products is the ultimate method to assess oil qual-
ity and stability. Sensory evaluation cannot be replaced fully by any chemical or
instrumental analysis, although some methods can correlate fairly well with sensory
results. Sensory evaluation of oils usually is done by a panel of experts or a trained
panel, and often the method recommended by the American Oil Chemists Society
(300) is used. During the evaluation, the panel is asked to score the overall flavor
quality, as well as the intensity of many individual off-flavors. Although chemical
and physical tests are more reproducible and less time consuming than sensory eva-
luations, oxidative rancidity and off-flavor evaluation of soybean oils are best done
by sensory tests. Correlations established between sensory evaluation scores and
various chemical tests, however, can be used to predict the sensory quality of fin-
ished oil products.
Peroxide value, expressed as milliequivalents of peroxide per kilogram of oil,
measures the primary oxidation products of oilsthe hydroperoxides. The peroxide
value has shown a particularly good correlation with sensory flavor scores of soy-
bean oil, and its use during storage is quite common. The peroxide value is an index
to the oxidative state of an oil. Soybean oil is considered fresh with a peroxide
value <1.0 mEq/kg, to have low oxidation with 1.05.0 mEq/kg, to have moderate
oxidation at 5.010.0 mEq/kg, to have high oxidation at >10.0 mEq/kg, and to have
poor flavor quality at >20 mEq/kg (6). Several methods (300303) can be used to
measure the peroxide value of an oil depending on the specific circumstance.
One of the first steps in the oxidation of polyunsaturated fatty acids is a shift in
the position of double bonds, resulting in the formation of conjugated hydroperox-
ides. The conjugated structure absorbs strongly at a wavelength of 232234 nm.
The conjugated diene value (300) is expressed as the percentage of conjugated die-
noic acid in the oil and is an indication of initial or primary oxidation products.
Conjugated diene value can be used as a comparative method only when the oils
have the same initial fatty acid composition, because the greater the amount of
polyenoates in an oil, the greater the potential rise in the conjugated diene value.
As a result, this method should be used as a relative measurement of oxidation in an
oil only if the fatty acid composition is known (303).
As aldehydes and some ketones have long been identified as oxidation and
breakdown products of fats, their determination also has been common in soybean
oil quality control. The p-anisidine value (300) measures light absorbance of alde-
hydes, primarily 2-alkenals, and 2,4-dienals at 350 nm. However, this measure is
not entirely specific, because the color intensity developed depends not only on
the concentration but also on the structure of the aldehyde. Therefore, the results
are comparable only within oils of similar type and treatment (304).
Free fatty acid (305), polar compounds (300), viscosity, and color analyses are
often performed to determine the degree of abuse that oils receive during heating or
frying. They are important indicators of frying oil quality, because these compo-
nents affect the quality of the fried food. The free fatty acid increase during frying
indicates released from triacylglyceride ester linkages via hydrolysis (233). Thus, it
is an important marker for oil quality. Abused frying oil should be discarded if it
contains >27% total polar compounds, according to a German standard of frying
oil quality (306). Changes in viscosity and color of the frying oil also are used as
indicators of the extent of frying oil degradation.
There are many other methods for measuring lipid oxidation and quality by che-
mical means. Among the best-known procedures are the thiobarbituric acid (TBA)
test, carbonyl value, and headspace oxygen analysis. These methods have been
reviewed and discussed elsewhere (287, 307).
The volatile carbonyl compounds formed during oxidation of fats and oils are
major contributors to off-flavor development. Therefore, there have been significant
efforts at identifying and quantifying these compounds. It is difficult to analyze
these compounds in fats and oils for several reasons. First, it is difficult to remove
636 SOYBEAN OIL
them quantitatively from the fats and oils. Second, widespread contamination by
carbonyls in solvents, glassware, and other laboratory materials may cause artifacts.
Finally, hundreds of volatile compounds may be formed in fats and oils during oxi-
dation causing difficulties in the interpretation. Today, the use of efficient gas chro-
matography (GC) columns and proper means of identification has made reliable
volatile compound analysis become possible.
Three basic GC procedures are generally employed (300), including static head-
space, dynamic headspace, and direct injection. Static headspace involves equili-
bration of gases from the area above a liquid sample; a set volume of the
headspace gas from the sample is then injected directly into the GC for separation
and quantification. The dynamic headspace method, also known as purge and trap,
employs a sorbent, such as Tenax GC, Chromosorb, or Porapak Q, to collect vola-
tile compounds that are swept from a heated sample with an inert gas such as
helium or nitrogen. After trapping, the sorbent may be extracted with solvent, or
transferred directly to the GC inlet port. In direct injection, an oil sample may be
injected directly into the port of the GC through a silanized glass wool plug. Each
of these methods has their own advantages and disadvantages (287).
Recently, the method of gas chromatographic solid-phase microextraction (GC-
SPME) has been developed (308310). This method uses fibers coated with various
polymers to extract volatile compounds from a food system. The method can be
used in solid, liquid, and gaseous systems. It is fairly easy to evaluate volatile com-
pounds by this analysis and to maintain consistent conditions.
Evans et al. (311) and Scholz and Ptak (312) used GC analysis of n-pentane as a
measurement of rancidity of vegetable oils. Dupuy et al. (313, 314) determined the
volatile carbonyl compounds from soybean oil using a modified gas chromato-
graphic inlet tube and found good correlations between the volatile profile analysis
and sensory scores. The Flavor Quality and Stability Committee of the AOCS eval-
uated GC volatile profiling as a standard method of flavor evaluation (275). As a
result, they wrote two Recommended Practices, entitled Volatiles in Fats and
Oils by Gas-Liquid Chromatography Cg 4-94, 1997 (300) and Correlation of
Oil Volatiles with Flavor Scores of Edible Oils AOCS method Cg 1-83, 1997
(300). These AOCS methods were validated in an AOCS collaborative study on
sensory and volatile analyses, in which three methods of volatile compound ana-
lyses were compared with sensory analyses by using the AOCS flavor scales
(315). Despite agreement on the usefulness of these methods, the committee
stressed that only humans can measure flavor, thus these volatile GC methods mea-
sured features such as oxidative stability and compound breakdownnot sensory
perceptions per se.
Not surprisingly, heat treatment, such as commercial and household frying,
accelerates autoxidation. In addition to undergoing autoxidation, when fats are
heated in the presence of moisture, as often is the case in food applications, fatty
acids are released via hydrolysis of the ester linkages (233). The free fatty acids can
accelerate oxidation of the oil. During heat treatment, the formation of dimeric and
cyclic compounds seems to be the predominant thermolytic reaction of unsaturated
fatty acids. In the presence of oxygen during heat treatment, however, oxidative
polymerization also can occur (233). Obviously, temperature, heating time, avail-
ability of oxygen, etc. can largely influence the extent to which these thermal and
oxidative polymerization reactions occur.
Decomposition and condensation of hydroperoxides also produces a multitude
of nonvolatile monomeric products, including di- and tri-oxygenated esters, and
dimeric and polymeric materials, especially at elevated temperature. Many of these
dimers and polymers are known to be rich sources of volatile carbonyl compounds
and to decrease the flavor and oxidative stability of soybean oil (316). These high-
molecular-weight materials also can produce a series of physical and chemical
changes to the oil and food products, including increased viscosity, polarity, free
acid content, development of dark color, and an increased tendency of the oil to
foam (233).
12.5. Storage and Handling

Production of good quality soybean oil requires close control from harvesting of the
soybeans, during bean storage, during and after oil processing, through consump-
tion of the finished oil products to guard against oxidative, enzymatic, and micro-
biological deterioration. Good processing measures include careful control of
refining temperature, vacuum bleaching, and inert gas blanketing. Heat accelerates
the reaction of atmospheric oxygen with edible oils, therefore, localized overheat-
ing is detrimental to final oil quality. After processing, soybean oil should be stored
at as low a temperature as possible and practical, and with protection from light.
Vacuum conditions are very important during bleaching, because oxidation can
readily occur by exposure of a large surface area to air at elevated temperatures.
During storage, a package containing the maximum amount of oil is preferable,
because oxygen availability is lower with a lower headspace-to-oil ratio. Peroxide
formation also is a linear function of surface-to-volume ratio (275). According to
List (317), in field storage tanks, the oil is also subjected to conditions that cause
development of sizable temperature gradients that can produce considerable inter-
nal oil movement. Such movement would be expected to increase the quantity of oil
available at the surface and to accelerate oxygen diffusion. Therefore, soybean oil
stored in filled tanks should be at as low a temperature as possible to avoid such
conditions.
12.6. Special Processing for Off-Specification Oil

Oils from field-, frost-, moisture-, and storage-damaged beans usually have higher
levels of free fatty acids and iron, lower levels of phosphorous, darker colors, and
poorer flavor and oxidative stability in the finished products than do oils from unda-
maged beans. Such beans are difficult to process, and standard processing methods
usually do not produce finished oils that can meet soybean oil specifications for
trading or domestic consumption.
The National Soybean Processors Association (318) trading rules specify that
prime crude oils, after refining and bleaching by an official method (300), must
638 SOYBEAN OIL
meet a Lovibond color level of 6.0. Frost-damaged oils often will not meet this
requirement. Oils from frost-damaged beans tend to have an undesirable green col-
or in the crude oils caused by compounds related to, but not identical to, chlorophyll
(includes pheophytin) or some of its derivatives, according to Stern and Grossman
(319). When bleaching such oils, acid-activated clays are more efficient than neutral
clays and increased amounts of bleaching earth make the removal of the green color
more effective. According to Stern and Grossman (319), pretreatment with charcoal
(0.41.0%) at 90 C or treatment of a cold hexane-oil mixture with charcoal is effec-
tive in partly removing the green pigment. When charcoal pretreatment is combined
with additional treatment from sugars and activated bleaching clays, complete
removal of green pigments is possible. Hydrogenation can also be used to remove
green color from soybean oil. According to Beal et al. (175), a green oil (IV 132)
hydrogenated to IV 110 in the presence of 1% copper chromate catalyst was no
longer green after cooling and filtration. However, the use of copper chromate is
not a common practice.
When soybeans are exposed to rain or humid weather in the field, the beans tend
to sprout and decay, and the oil from these beans develops a dark-brown color and
chalky texture (312). Drought stress affects protein and oil content of soybeans but
seldom damages oil quality significantly. According to List (317), off-specification
oils from field-, frost-, heat-, and moisture-damaged soybeans result in high refining
losses during processing, poor refined-bleached color, and lowered flavor and oxi-
dative stability. High refining losses may be partly overcome by use of phosphoric
acid or acetic anhydride degumming. Color problems of oils from damaged beans
may be alleviated, in part, by use of acidic bleaching earths, increased amounts of
bleaching earths, and higher bleaching temperatures. Overall, however, the best
practice for producing high-quality oil is to segregate the bad beans and not include
them in the processing.
13. DIETARY FATTY ACIDS AND THEIR HEALTH EFFECTS
13.1. Cholesterol and Heart Disease

Heart disease is still the number one cause of death for both men and women in the
United States. High-blood-cholesterol levels increase the risk of getting heart dis-
ease (319), so, generally, serum (blood) cholesterol is measured to determine a per-
sons risk of developing heart disease. Although some cholesterol is essential in
forming the bodys cell membranes and synthesizing hormones and bile acids,
too much cholesterol is associated with heart disease. The fat eaten can affect
the blood-cholesterol level. In addition to monitoring total blood cholesterol, the
ratio of high-density lipoproteins (HDL) to low-density lipoproteins (LDL) of the
blood also is important in predicting heart disease. As cholesterol, a waxy sub-
stance, does not mix with water, it needs help circulating through blood, which
is mostly water. Lipoproteins transport cholesterol throughout the body. Low-den-
sity lipoproteins carry cholesterol from the liver to the body and leave deposits on
artery walls. High-density lipoproteins carry cholesterol back to the liver for
DIETARY FATTY ACIDS AND THEIR HEALTH EFFECTS 639
elimination. If the ratio of low/high-density lipoproteins becomes too large, it is

likely that more cholesterol will be deposited in the arteries than is removed. So
the low/high-density lipoprotein ratio also may be used to predict a persons
chances of developing heart disease. A ratio greater than 3 can indicate above aver-
age risk. The most important dietary influences on blood cholesterol levels are satu-
rated fat, total fat, and dietary cholesterol.
13.2. Saturated Fat and Health Effects

Saturated fat has more impact on raising blood cholesterol levels than anything else
in the diet. The most effective way to reduce the blood cholesterol level is to reduce
the amount of saturated fat in the diet. Animal products are a major source of satu-
rated fat in the average American diet. A very few vegetable oils, including coco-
nut, palm kernel, and palm oils, are rich in saturated fat. Other vegetable oils,
including soybean oil, can become saturated by hydrogenation. Consumption of
too much saturated fat has been associated with the development of heart disease,
some cancers, and other health problems. As soybean oil is the major edible oil
consumed in the United States, lowering its saturated fat could help reduce heart
disease in this country, even though its total saturated fatty acid composition is
only about 15% to 16%. As noted, the major saturates in soybean oil are palmitate
and stearate. Palmitate is responsible for about 70% of the total saturated fat in soy-
bean oil. Substitution of palmitate for carbohydrates or monounsaturates in the diet
increased levels of serum low-density lipoproteins and total cholesterol (320). Stea-
rate has been found to be relatively neutral in its effects on blood lipids, and some
researchers (321, 322) showed that dietary stearate actually lowered serum low-
density lipoproteins and total cholesterol levels; thus, many people recommend
that this saturate not be included in the category of hypercholesterolemic acyl
groups. It was for these reasons that Iowa State University scientists developed
LoSatSoyTM, a soybean oil with half the saturated fat of conventional soybean
oil, with reduction of palmitate to <3%.
13.3. Unsaturated Fat and Health Effects

Unsaturated fats, classified as either monounsaturated or polyunsaturated, can help
lower the cholesterol levels in blood when substituted for saturated fats. Sources of
monounsaturated fat include nuts, olive oil, and canola oil. Sources of polyunsatu-
rated fat include corn, safflower, sesame, soybean, and sunflower oils.
Soybean oil contains about 21% of the monounsaturate oleate. Studies have
shown that the oxidation rate of oleate is much slower than that of the polyunsatu-
rates, linoleate and linolenate, which oxidize quickly and are the major contributors
to the poor stability of soybean oil (287, 323). A diet high in monounsaturates may
help to reduce elevated levels of total plasma cholesterol without reducing the high-
density lipoprotein-cholesterol level (324). Therefore, high-oleate soybean oil is not
only more stable than conventional soybean oil (275), but also has enhanced nutri-
tive value.
640 SOYBEAN OIL
In both clinical trials and population studies, polyunsaturated fats in the diet
have been shown to actively lower serum cholesterol levels. Soybean oil is consid-
ered to have good nutritive value mainly because of its high concentration of essen-
tial polyunsaturates. As noted previously, it contains about 55% linoleate and 8%
linolenate, both recognized as essential fatty acids. Ingestion of approximately
12% of daily calories as linoleate is widely accepted as the amount needed to
meet the essential fatty acid requirement of rodent species and humans (325).
The physiological effects of linoleate have been well characterized. Various
deficiency symptoms include depressed growth, scaly dermatoses, increased skin
permeability, fatty liver, kidney damage, and impaired reproduction. The 8%
linolenate of soybean oil, makes it not only an excellent source of essential fatty
acids, but also a member of the n-3 fatty acid group (the third carbon atom from
the terminal end of the hydrocarbon chain is involved in a double bond). A number
of health benefits have been associated with the consumption of foods or oils that
contain n-3 fatty acids. These associations originally derived from epidemiological
studies of Eskimos who consumed high levels of n-3 fatty acid from seals and
coldwater fish (326). Compared with Danish counterparts, these Eskimos were
found to have a low incidence of heart disease and immune system diseases,
although a somewhat higher level of hemorrhagic stroke. Still today, large-scale
epidemiological studies suggest that individuals at risk for CHD benefit from the
consumption of plant- and marine-derived n-3 fatty acids (327).
13.4. Trans-Fatty Acids and Their Health Effects

The process of catalytic hydrogenation of vegetable oils was discovered in 1897 to
reduce the polyunsaturates and to improve flavor stability, versatility, and perfor-
mance of vegetable oils in salad dressings, during cooking, in deep-fat-frying,
and for the manufacture of margarines, shortenings, and other baking and snack
food applications (328). A side reaction that occurs during hydrogenation is the for-
mation of positional and geometrical isomers of the unsaturated sites that are left
unsaturated. Formation of trans-isomers is rapid and extensive (320). Although
hydrogenation can improve soybean oil oxidative stability and performance versa-
tility, the presence of the trans-fatty acids may make hydrogenated oils nutritionally
undesirable. In particular, the role of partially hydrogenated soybean oil in nutrition
has been under scrutiny because of the health concerns over the presence of trans-
acyl groups in our diets (329); however, the biological significance of these trans-
acyl groups is unclear. The formation of trans-acyl groups in vegetable oils also can
occur, to a small extent, during deodorization (330, 331) and during frying (332,
333). The 9-cis,12-trans-linoleate is present in most vegetable shortenings in
much greater quantities than the 9-trans,12-trans-linoleic acid (334). In heated
vegetable oils, the isomers just mentioned have been reported, plus trans-, cis-iso-
mers of linolenate (330, 332, 335). Trans-isomers are essential fatty acid antago-
nists, especially when the linoleate and linolenate are limited in the diet. For
example, the cis, cis, trans-isomer of 18:3 is elongated and desaturated to form
n-3 trans-isomers of 20:5n-3 and 22:6n-3 in rats (336); isomers that also have
REFERENCES 641
been found in human platelets (337). The 9-cis,12-trans-linoleic acid can be con-
verted to 20:4n-6 containing a trans-double bond. Unfortunately, this trans-isomer
of 20:4n-6 inhibited the formation of prostaglandins from all-cis-20:4n-6 (338).
Mensink and Katan (250) reported that a diet high in trans-acyl groups raised total
and low/high-density lipoprotein cholesterol ratio compared with a diet high in cis-
acyl groups may be more cholesterolemic than saturates (339), and were linked to
an increased risk of breast cancer development (340).
The estimated trans-acyl group intake by typical U.S. consumers is 11.127.6 g/
person/day (341). A comprehensive review concluded that trans-acyl groups con-
sumed at 4.0% or more of total calories may raise plasma lipid levels (342). As a
result of health concerns over the presence of trans-acyl groups in our diet, mod-
ifying fatty acid composition of soybean oil to improve its oxidative and flavor
stability in ways similar to that obtained by hydrogenation, but without trans-
formation, has become an objective of plant breeders.
13.5. Total Fat and Its Health Effects

Excessive intake of any fat is not healthy. According to Klurfeld and Kritchevsky
(342), the enhancement of tumor growth by dietary fat may result, in part, from the
caloric contribution of this nutrient. Significant reduction of tumor incidence with
consumption of 25% less energy was seen consistently in rat tumor systems induced
by chemicals. Currently, most American children get about 34% of their calories
from fat (318). It is recommended, however, that healthy childrens intake of fat
average no more than 30% of calories. Experts also suggest lowering childrens
saturated fat intake to less than 10% of calories. Similar recommendations have
been made for adults (343).
REFERENCES
1. Economic Research Service of United States Department of Agriculture. (October,

2002). Oil Crop Situation and Outlook Yearbook. Available: http://usda.mannlib.
cornell.edu/reports/erssor/field/ocs-bby/ocs2002.pdf.
2. 2004 Soya & Oilseed Bluebook, Soyatech, Inc., Bar Harbor, Maine, 2003.
3. T. J. Brumm and C. R. Hurburgh, Jr., Quality of the 2002 Soybean Crop from the
United States, Report to the American Soybean Association.
4. D. M. Maestrl, J. M. Mariles, and C. A. Guzman, Grassas Aceites, 49, 395399 (1998).
5. C. R. Hurburgh, Jr., J. Amer. Oil Chem. Soc., 71, 14251427 (1994).
6. D. M. Maestri, D. O. Labuckas, J. M. Meriles, A. L. Lamarque, J. A. Zygadlo, and C. A.
Guzman, J. Sci. Food Agric., 77, 494498 (1998).
7. R. F. Wilson, Agro-Food Industry Hi-Tech., 12(6), 1722 (2002).
8. R. F. Wilson, Lipid Technol., 11, 107110 (2000).
9. J. A. Hoeck, W. R. Fehr, P. A. Murphy, and G. A. Welke, Crop Sci., 40, 4851 (2000).
10. J.-S. Choi, T.-W. Kwon, and J.-S. Kim, Food Biotechnol., 5, 167169 (1996).
642 SOYBEAN OIL
11. C. Tsukamoto, S. Shimada, K. Igita, S. Kudou, M. Kokubun, K. Okubo, and

K. Kitamura, J. Agric. Food Chem., 43, 11841192 (1995).
12. T. Wang, Phospholipid fatty acid composition of modified soybeans and the effect of
saturated fatty acid content on seed performance, Ph.D. Dissertation, Iowa State
University, Ames, Iowa, 1998.
13. T. Gutfinger and A. Letan, Lipids, 9, 658663 (1974).
14. M. Bastic, L. J. Bastic, J. A. Javanovic, and G. Spiteller, J. Amer. Oil Chem. Soc., 55,
886891 (1978).
15. J. D. Weete and R. C. Manley, J. Ala. Acad. Sci., 50, 3546 (1979).
16. T. L. Mounts, S. L. Abidi, and K. A. Rennick, J. Amer. Oil Chem. Soc., 73, 581586
(1996).
17. J. L. Weihrauch and J. M. Gardner, J. Amer. Diet. Assoc., 73, 39 47 (1978).
18. C. D. Evans, G. R. List, R. E. Beal, and L. T. Black, J. Amer. Oil Chem. Soc., 51, 444448
(1974).
19. G. R. List, A. J. Heakin, C. D. Evans, L. T. Black, and T. L. Mounts, J. Amer. Oil Chem.
Soc., 55, 521522 (1978).
20. G. O. Almonor, G. P. Fenner, and R. F. Wilson, J. Amer. Oil Chem. Soc., 75, 591596
(1998).
21. C. Vlahakis and J. Hazebroek, J. Amer. Oil Chem. Soc., 77, 4953 (2000).
22. K. L. McCord, W. R. Fehr, T. Wang, G. A. Welke, S. R. Cianzio, and S. R. Schnebly,
Crop Sci., 44, 772776 (2004).
23. S.-N. Ko, C.-J. Kim, H. Kim, C.-T. Kim, S.-H. Chung, B.-S. Tae, and I.-H. Kim, J. Amer.
Oil Chem. Soc., 80, 585589 (2003).
24. L. H. Wiedermann, J. Amer. Oil Chem. Soc., 58, 159166 (1981).
25. S. Ramamurthi, A. R. McCurdy, and R. T. Tyler, in S. S. Koseoglu, K. C. Rhee, and
R. F. Wilson, eds., Proc. World Conf. Oilseed Edible Oils Process, vol. 1, AOCS Press,
26. Yearbook and Trading Rules, National Oilseed Processors Association, Washington,
D.C., 1993.
27. H. Klein, Ernahrung/Nutrition, 23, 452460 (1999).
28. D. M. Maestri, D. O. Labuckas, C. A. Guzman, and L. M. Glorda, Grasas Aceites, 49,
450453 (1998).
29. J. Siebrecht and J. Hazebroek, J. Amer. Oil Chem. Soc., 74, 989998 (1997).
30. X. Wu, E. G. Hammond, P. J. White, and W. Fehr, J. Amer. Oil Chem. Soc., 74, 1099
1103 (1997).
31. F. G. Sietz, Fette Scifen Anstrichm., 67, 411412 (1965).
32. M. Wesolowski, Food Sci.Technol., 95, 377383 (1993).
33. X. Ndzana, W. R. Fehr, G. A. Welke, E. G. Hammond, D. N. Duvick, and S. R. Cianzio,
Crop Sci., 34, 646649 (1994).
34. S. R. Schnebly, W. R. Fehr, G. A. Welke, E. G. Hammond, and D. N. Duvick, Crop Sci.,
36, 14621466 (1996).
35. G. L. Graef, L. A. Miller, W. R. Fehr, and E. G. Hammond, J. Amer. Oil Chem. Soc., 62,
773775 (1985).
36. T. Wang, E. G. Hammond, and W. R. Fehr, J. Amer. Oil Chem. Soc., 78, 11391144 (2001).
REFERENCES 643
37. A. J. Kinney, J. Food Lipids, 3, 273292 (1996).

38. W. R. Fehr and E. G. Hammond (to Iowa State University Research Foundation), U.S.
Patent 5,534,425, July 9, 1996.
39. R. F. Wilson, J. W. Burton, V. R. Pantalone, and R. E. Dewey, Lipid Biotechnology,
Marcel Dekker, New York, 2002, pp. 95113.
40. T. F. Horejsi, W. R. Fehr, G. A. Welke, D. N. Duvick, E. G. Hammond, and S. R. Cianzio,
Crop Sci., 34, 331334 (1994).
41. R. B. Hartmann, W. R. Fehr, G. A. Welke, E. G. Hammond, D. N. Duvick, and S. R.
Cianzio, Crop Sci., 36, 14661470 (1996).
42. T. Wang, T. Harp, E. G. Hammond, J. S. Burris, and W. R. Fehr, Seed Sci. Res., 11, 9397
(2001).
43. W. R. Fehr, J. C. Thorne, and E. G. Hammond, Crop Sci., 11, 211213 (1971).
44. G. Ishikawa, H. Hasegawa, Y. Takagi, and T. Tanisaka, Plant Breeding, 120, 417423
(2001).
45. T. K. Harp and E. G. Hammond, Lipids, 33, 209216 (1998).
46. T. Wang, unpublished results, September 2003.
47. C. Litchfield, Analysis of Triglycerides, Academic Press, New York, 1972, pp. 233264.
48. W. E., Neff, G. R. List, and W. C. Byrdwell, J. Liq. Chromatogr. Related. Technol., 22,
16491662 (1999).
49. W. E. Neff and W. C. Byrdwell, J. Amer. Oil Chem. Soc., 72, 11851191 (1995).
50. T. Wang, E. G. Hammond, and W. R. Fehr, J. Amer. Oil Chem. Soc., 74, 15871594
(1997).
51. T. Wang, E. G. Hammond, J. L. Cornette, and W. R. Fehr, J. Amer. Oil Chem. Soc., 76,
13131321 (1999).
52. T. L. Mounts and A. M. Nash, J. Amer. Oil Chem. Soc., 67, 757760 (1990).
53. G. R. List, T. L. Mounts, and A. C. Lanser, J. Amer. Oil Chem. Soc., 69, 443446 (1992).
54. E. Gutierrez, Method Development for Sphingolipid Quantification in Soybeans and Soy
Products, M. S. Thesis, Iowa State University, Ames, Iowa 2003.
55. V. Piironen, T. Koivu, O. Tammisalo, and P. Mattila, Food Chem., 59, 473480 (1997).
56. T. Yamagami, M. Aoyama, T. Tsutsumi, S. Tokairin, H. Ehara, T. Maruyama, and
I. Niiya, Yukagakkaishi, 48, 12711274 (1999).
57. T. Yamagami, M. Aoyama, T. Tsutsumi, S. Tokairin, H. Ehara, T. Maruyama, and
I. Niiya, Chem. Abstr., 132, 77768 (2000).
58. T. Koivu, V. Piironen, A.-M. Lampi, and P. Mattila, J. Food Chem., 64, 411414
(1999).
59. R. F. Wilson, J. W. Burton, W. P. Novitzky, and R. E. Dewey, J. Oleo Sci., 50, 353358
(2001).
60. K. Warner, P. Orr, and M. Glynn, J. Amer. Oil Chem. Soc., 74, 347356 (1997).
61. H. Guth and W. Grosch, Fat Sci. Technol., 93, 249255 (1991).
62. J.-W. Kao, X. Wu, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 75, 831
835 (1998).
63. J. deD. Alvarado, Grasas Aceites, 46, 264269 (1995).
64. J. deD. Alvarado, Chem. Abstr., 124, 200581 (1996).
644 SOYBEAN OIL
65. V. V. Kravchenko, A. K. Pugach, and V. Z. Geller, Izv. Vysshikh, Uchebnykh, Zavedenii,

Pishchevaya Tekhnol., 1987, 7981 (1987).
66. V. V. Kravchenko, A. K. Pugach, and V. Z. Geller, Chem. Abstr., 108, 54638 (1987).
(1992).
68. Y. Tochitani and M. Fujimoto, Netsu Bussei, 15, 230236 (2001).
69. Y. Ali, M. A. Hanna, and S. L. Cuppett, J. Amer. Oil Chem. Soc., 72, 15571564 (1995).
70. R. Sh. Kuliev, F. R. Shirinov, and F. A. Kuliev, Chem. Technol. Fuels Oils., 35, 235237
(1999).
71. H. Koseki, Y. Natsume, and Y. Iwata, J. Fire Sci., 19, 3144 (2001).
72. E. W. Washburn, ed. International Critical Tables, vol. 5, McGraw-Hill, New York,
1929, pp. 162169.
73. K. S. Miller, B. E. Farkas, and R. P. Singh, J. Food Process. Preservation., 18, 461472
(1994).
74. E. S. Perry, W. H. Weber, and B. F. Daubert, J. Amer. Chem. Soc., 71, 37203726 (1949).
75. A. Tekin and E. G. Hammond, J. Amer. Oil Chem. Soc., 75, 737740 (1998).
76. S. B. Detwiler, Jr. and K. S. Markley, Oil & Soap, 17, 3940 (1940).
77. K. Yokota and M. Tachimori, Kogyo Kagaku Zasshi, 40 (Suppl. binding) 426429 (1937).
78. K. Yokota and M. Tachimori, Chem. Abstr., 32, 13680 (1937).
79. J. D. Halvorsen, W. C. Mammel, Jr., and L. D. Clements, J. Amer. Oil Chem. Soc., 70,
875880 (1993).
80. C. M. Rodenbush, F. H. Hsieh, and D. S. Viswanath, J. Amer. Oil Chem. Soc., 76, 1415
1419 (1999).
81. H. F. Johnstone and I. H. Spoor, Ind. Eng. Chem., 32, 832835 (1940).
82. F. C. Magne and E. L. Skau, Ind. Eng. Chem., 37, 10971102 (1945).
83. L. K. Arnold and L. J. Breuklander, Iowa Acad. Sci., 57, 157160 (1950).
84. F. C. Magne, E. J. Hughes, and E. L. Skau, J. Amer. Oil Chem. Soc., 27, 552555 (1950).
85. P. E. Clark, C. R. Waldeland, and R. P. Cross, J. Ind. Eng. Chem., 38, 350353 (1946).
86. T. Kasprzycka-Guttman and D. Odzeniak, Thermochimica Acta, 191, 4145 (1991).
87. T. Wang and J. L. Briggs, J. Amer. Oil Chem. Soc., 79, 831836 (2002).
88. N. A. Morad, A. A. Mustafa Kamal, F. Panau, and T. W. Yew, J. Amer. Oil Chem. Soc.,
77, 10011105 (2000).
89. Y. Hagura and K. Suzuki, Nippon Shokuhin Kagaku Kogaku Kaishi, 49, 272276 (2002).
90. Y. Hagura and K. Suzuki, Chem. Abstr., 138, 38393 (2003).
91. V. Chioffi, Chim. Provinciali, 12, 123136 (1961).
92. G. J. Arissen, Verfkroniek, 6, 911 (1996).
93. D. B. Dahlberg, S. M. Lee, S. J. Wenger, and J. A. Vargo, Applied Spectroscopy, 51,
11181124 (1997).
94. I. Boyaci, A. Tekin, M. Cizmeci, and I. Javidpour, J. Food Lipids, 9, 175183 (2002).
95. J. A. Ibemesi and I. O. Igwe, J. Amer. Oil Chem. Soc., 68, 147152 (1991).
96. S. Z. Erhan, S. Asaduauskas, and A. Adhvaryu, J. Amer. Oil Chem. Soc., 79, 11571161
(2002).
97. E. G. Hammond and W. O. Lundberg, J. Amer. Oil Chem. Soc., 31, 427432 (1954).
REFERENCES 645
98. K. R. Majors and R. T. Milner, Oil & Soap, 16, 228231 (1939).
99. Y. Maruta and K. Tenurama, Kogyo Kagaku Zasshi, 40 (Suppl), 299 (1937).
100. Y. Maruta and K. Tenurama, Chem. Abstr., 31, 65303 (1937).
101. J. R. Dafler, J. Amer. Oil Chem. Soc., 52, 425426 (1975).
102. C. F. Cole, G. M. Hill, and A. J. Adams, J. Amer. Oil Chem. Soc., 71, 13391342 (1994).
103. Y. H. Chu, J. Amer. Oil Chem. Soc., 68, 379384 (1991).
104. C. Inoue, Y. Hagura, M. Ishikawa, and K. Suzuki, J. Food Sci., 67, 11261129 (2002).
105. N. Tomoto and K. Kusano, Yukagaku, 16, 108113 (1967).
106. N. Tomoto and K. Kusano, Chem. Abstr., 66, 106141 (1967).
107. M. Loncin, Fette Seifen Anstrichm., 57, 413415 (1955).
108. A. S. Smith and F. J. Wechter, J. Amer. Oil Chem. Soc., 27, 381383 (1950).
109. A. S. Smith, J. Amer. Oil Chem. Soc., 28, 360361 (1951).
110. A. S. Smith, J. Amer. Oil Chem. Soc., 28, 356359 (1951).
111. K. Kusano, Yukagaku, 8, 814 (1959).
112. K. Kusano, Chem. Abstr., 54, 2002060 (1960).
113. A. I. Anikin, G. N. Danilova, and V. V. Klyuchkin, Izvestiaya Vysshikh Ucbebnykh,
Zavedenii Pishchevaya Tekhnologiya, 4345 (1987).
114. A. I. Anikin, G. N. Danilova, and V. V. Klyuchkin, Chem. Abstr., 108, 54636 (1988).
115. C. Aberhard and Z. Spekuljak, Revista Facultad Ingenieria Quimica, 48, 712 (1988).
116. C. Aberhard and Z. Spekuljak, Chem. Abstr., 112, 156952 (1990).
117. W. H. Dickhart, Amer. J. Pharm., 104, 284 (1932).
118. G.-C. Yen, C. H. Shao, C. J. Chen, and P. D. Duh, Food Sci. Technol., 30, 648652
(1997).
119. B. Kowalski, Thermochimica Acta, 173, 117127 (1990).
120. V. Barquin, Tecnicas Laboratorio, 11, 345349 (1987).
121. M. Wakakura, Kenkyu Hokoku-Kanagawa-ken Kogyo Shikensho, 55, 7677 (1984).
122. M. Wakakura, Chem. Abstr., 103, 217718 (1985).
123. Federal Grain Inspection Service, Federal Grain Inspection Handbook, Washington,
D.C., 1990.
124. U.S. Department of Agriculture. Available: http://www.usda.gov/gipsa/reference-
library/brochures/soyinsepction.pfd.
125. T. J. Brumm and C. R. Hurburgh, J. Amer. Oil Chem. Soc., 67, 302307 (1990).
126. University of Illinois, Urbana-Champaign, Illinois. Available: http://www.stratsoy.uiuc.
edu/epv/.
127. T. L. Mounts, G. R. List, and A. J. Heakin, J. Amer. Oil Chem. Soc., 56, 883885 (1979).
128. G. R. List C. D. Evans, K. Warner, R. E. Beal, W. F. Kwiolek, L. T. Black, and K. J.
Moulton, J. Amer. Oil Chem. Soc., 54, 814 (1977).
129. C. R. Hurburgh, T. J. Brumm, J. M. Quin, and R. A. Hartwig, J. Amer. Oil Chem. Soc., 67,
966973 (1990).
130. E. G. Hammond, L. A. Johnson, and P. A. Murphy, in C. Wrigley, C. Walker, and
H. Corke, eds., Encyclopaedia of Grain Sciences, Academic Press, London, 2003.
131. P. Golbitz, 2003 Soya & Oilseed Bluebook, Soyatech, Inc., Bar Harbor, Maine,
2003.
646 SOYBEAN OIL
132. D. Erickson, ed., Handbook of Soybean Processing and Utilization, AOCS Press,
133. L. A. Johnson, in P. J. Wan and W. Farr, eds., Introduction to Fats and Oils, AOCS Press,
134. L. A. Johnson, in C. Akoh and D. Min., eds., Food Lipids, 2nd ed., Marcel Dekker,
New York, 2002, pp. 223273.
135. T. J. Brumm, C. R. Hurburgh, and L. A. Johnson, J. Amer. Oil Chem. Soc., 67, 750756
(1990).
136. J. M. Guinn, American Soybean Association, St. Louis, MO. (2002). Available: http://
www.asasoya.org/Import/Guinn_Quality_Standards_Trading_Rules2002.htm.
137. L. A. Johnson and E. W. Lusas, J. Amer. Oil Chem. Soc., 60, 181A193A (1983).
138. R. J. Hron, J. Amer. Oil Chem. Soc., 59, 674A684A (1982).
139. P. J. Wan and P. J. Wakelyn, eds., Technology and Solvents for Extracting Oilseeds and
Nonpetroleum Oils, AOCS Press, Champaign, Illinois, 1997.
140. R. D. Good, Oil Mill Gaz., 75(3), 1417 (1970).
141. F. McDonald, Oil Mill Gaz., 84(12), 25 (1980).
142. M. A. Williams, Oil Mill Gaz., 91(5), 2429 (1986).
143. M. A. Williams, INFORM, 6(3), 289293 (1995).
144. E. W. Lusas and L. R. Watkins, J. Amer. Oil Chem. Soc., 65, 11071114 (1988).
145. National Oilseed Processors Association, Washington, D.C. (2002). Available: http://
www.asasoya.org/Import/Guinn_Quality_Standards_Trading_Rules2002.htm.
146. A. I. Nelson, W. B. Wijeratne, S. W. Yeh, T. M. Wei, and L. S. Wei, J. Amer. Oil Chem.
Soc., 64, 13411347 (1987).
147. N. W. Said, INFORM, 9(2), 139144 (1997).
148. T. Wang, L. A. Johnson, and W. Wijeratne, in N. T. Dunford and H. B. Dunford, eds.,
Nutritionally Enhanced Edible Oil Processing, AOCS Press, Champaign, Illinois, 2004,
p. 219238.
149. T. Wang and L. A. Johnson, J. Amer. Oil Chem. Soc., 78, 311318 (2001).
150. T. Wang and L. A. Johnson, in Proceedings of 2003 International Conference on
Environmental Systems, Vancouver, British Columbia, Canada, 2003.
151. L. A. Johnson, D. J. Myers, and D. J. Burden, INFORM, 3, 282290 (1992).
152. L. A. Johnson, D. J. Myers, and D. J. Burden, INFORM, 3, 429444 (1992).
153. J. C. Endres, Soy Protein Products: Characteristics, Nutritional Aspects, and Utilization.
AOCS Press, Champaign, Illinois, 2001.
154. C. T. Egger and R. E. Olson (assigned to Grain Processing Corporation, Muscatine,
Iowa), U.S. Patent 3,849,391, November 19, 1974.
155. C. Y. Wang and L. A. Johnson, J. Amer. Oil Chem. Soc., 78, 189195 (2001).
156. P. A. Howard, M. Campbell, and D. T. Zollinger (assigned to Central Soya Co., Ft.
Wayne, Indiana), U.S. Patent Re. 32,725, August 2, 1988.
157. S. Wu, P. A. Murphy, L. A. Johnson, A. R. Fratzke, and M. A. Reuber, J. Amer. Oil Chem.
Soc., 76, 285293 (1999).
158. S. Wu, P. A. Murphy, L. A. Johnson, M. A. Fratzke, and M. A. Reuber, J. Agric. Food
Chem., 48, 27022708 (2000).
159. K. H. Lee, H. S. Ryu, and K. C. Rhee, J. Amer. Oil Chem. Soc., 80, 8590 (2003).
REFERENCES 647
160. National Archives and Records Administration, Federal Register, 64, (206), 57699
57733 (1999).
161. Food and Drug Administration. Available: http://vm.cfsan.fda.gov/%7Elrd/fr991026.html.
162. M. Kock, J. Amer. Oil Chem. Soc., 60, 198202 (1983).
163. C. D. Evans, P. M. Cooney, C. R. Scholfield, and H. J. Dutton, J. Amer. Oil Chem. Soc.,
31, 295297 (1954).
164. G. R. List, C. D. Evans, L. T. Black, and T. L. Mounts, J. Amer. Oil Chem. Soc., 55, 275
276 (1978).
165. Yearbook and Trading Rules (19931994), National Oilseed Processors Association,
Washington, D.C., p. 86.
166. J. C. Seger and R. L. K. M. van de Sande, in D. R. Erickson, ed., World Conference
Proceedings Edible Fats and Oils Processing: Basic Principles and Modern Practices,
167. A. J. Dijkstra, in T. A. Applewhite, ed., Proceedings of the World Conference on Oilseed
Technology and Utilization, AOCS Press, Champaign, Illinois, 1992, pp. 138151.
168. C. Pagliero, N. Ochoa, J. Marchese, and M. Mattea, J. Amer. Oil Chem. Soc., 78, 793
796 (2001).
169. L. P. Raman, M. Cheryan, and N. Rajagopalan, J. Amer. Oil Chem. Soc., 73, 219224 (1996).
170. S. Subramanian, M. Nakajima, A. Yasui, H. Nabetani, T. Kimura, and T. Maekawa,
171. G. R. List, J. W. King, J. H. Johnson, K. Warner, and T. L. Mounts, J. Amer. Oil Chem.
Soc., 70, 473476 (1993).
172. K. J. Moulton and T. L. Mounts, J. Amer. Oil Chem. Soc., 67, 3338 (1990).
173. D. R. Erickson, in D. R. Erickson, ed., Practical Handbook of Soybean Processing and
Utilization, AOCS Press, Champaign, Illinois, 1995, pp. 184202.
175. R. E. Beal, E. G. Lancaster, and O. L. Brekke, J. Amer. Oil Chem. Soc., 33, 619624 (1956).
176. E. Hernandez, Abstracts: 92nd AOCS Annual Meetings and Expo, Minneapolis, 2001,
p. S98.
178. M. Y. Jung, S. H. Yoon, and D. B. Min, J. Amer. Oil Chem. Soc., 66, 118120 (1989).
179. D. R. Erickson, in D. R. Erickson, ed., Practical Handbook of Soybean Processing and
180. D. R. Erickson and M. D. Erickson, in D. R. Erickson, ed., Practical Handbook of
Soybean Processing and Utilization, AOCS Press, Champaign, Illinois, 1995, pp. 218
238.
181. G. R. List, W. E. Neff, R. L. Holliday, J. W. King, and R. Holser, J. Amer. Oil Chem. Soc.,
77, 311314 (2000).
182. Y. Chu and L. Lin, J. Amer. Oil Chem. Soc., 69, 880883 (1992).
183. K. J. Moulton, R. E. Beal, and E. L. Griffin, J. Amer. Oil Chem. Soc., 50, 450454 (1973).
184. S. Koritala, J. Amer. Oil Chem. Soc., 54, 267268 (1977).
185. T. L. Mounts, in D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and R. A. Falb,
eds., Handbook of Soybean Oil Processing and Utilization, AOCS Press, Champaign,
Illinois, 1980, pp. 131144.
648 SOYBEAN OIL
186. R. C. Hastert, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products, vol. 4: Edible
Oil and Fat Products: Processing Technology, 5th ed., John Wiley & Sons, Inc.,
New York, 1996, pp. 213300.
187. M. Kellens, in W. Hamm and R. J. Hamilton, eds., Edible Oil Processing, Sheffield
Academic Press, Sheffied, U.K., 2000, pp. 129173.
188. T. Okamoto, T. Tsutsumi, S. Tokairin, H. Ehara, T. Maruyama, I. Niiya, and M. Sugano,
Jpn. Inst. Oil & Fat, 48, 877883 (1999).
189. Z. Kemeny, K. Recseg, G. Henon, K. Kovari, and F. Zwobada, J. Amer. Oil Chem. Soc.,
78, 973979 (2001).
190. W. F. De Greyt, M. J. Kellens, and A. D. Huyghebaert, Fett/Lipid, 101, 428432 (1999).
191. W. De Greyt and M. Kellens, in W. Hamm and R. J. Hamilton, eds., Edible Oil
Processing, Sheffield Academic Press, Sheffied, U.K., 2000, pp. 79128.
192. T. W. Kwon, H. E. Snyder, and H. G. Brown, J. Amer. Oil Chem. Soc., 61, 18431846
(1984).
193. T. Verleyen, U. Sosinska, S. Ioannidou, R. Verhe, K. Dewettinck, A. Huyghebaert, and
W. De Greyt, J. Amer. Oil Chem. Soc., 79, 947953 (2002).
194. R. Krishnamurthy and M. Kellens, in Y. H. Hui, ed., Baileys Industrial Oil and Fat
Products, vol. 3: Edible Oil and Fat Products: Products and Application Technology, 5th
ed., John Wiley & Sons, Inc., New York, 1996, pp. 193223.
195. A. Wendel, in M. Howe-Grant, ed., Kirk-Othmer Encyclopedia of Chemical Technology,
vol. 15, 4th ed., John Wiley & Sons, New York, 1995, pp. 192210.
196. O. L. Brekke, in D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and R. A. Falb,
eds., Handbook of Soy Oil Processing and Utilization, AOCS Press, Champaign, Illinois,
1980, pp. 7188.
197. L. Teberikler, S. Koseoglu, and A. Akgerman, J. Amer. Oil Chem. Soc., 78, 115119
(2001).
198. Y. Z. Wu and T. Wang, J. Amer. Oil Chem. Soc., 80, 127132 (2003).
199. Y. Wu and T. Wang, J. Amer. Oil Chem. Soc., 80, 319326 (2003).
200. G. R. List, in B. F. Szuhai, ed., Lecithins: Source, Manufacture & Uses, AOCS Press,
201. M. D. Pickard, T. J. Jones, and R. T. Tyler, in Y. H. Hui, ed., Baileys Industrial Oil and
Fat Products, vol. 4: Edible Oil and Fat Products: Processing Technology, 5th ed.,
John Wiley & Sons, Inc., New York, 1996, pp. 603630.
202. R. L. Winters, in D. R. Erickson, ed., Proceedings: World Conference Edible Fats and
Oils Processing, Basic Principles and Modern Practices, AOCS Press, Champaign,
Illinois, 1990, pp. 402405.
203. J. P. Clark and S. S. Frandsen, in S. S. Koseoglu, K. C. Rhee, and R. F. Wilson, eds., Proc.
World Conf. Oilseed Edible Oils Process, vol. 1, AOCS Press, Champaign, Illinois, 1998,
pp. 135138.
204. M. R. Law, Western Journal of Medicine, 173, 43 47 (2000).
205. L. Normen, P. Dutta, A. Lia, and H. Andersson, Amer. J. Clin. Nutr., 71, 908913 (2000).
206. P. Hollingsworth, Food Tech., 55, 5962 (2001).
207. K. B. Hicks and R. A. Moreau, Food Techol., 55, 6367 (2001).
208. H. Wu, X. Weng, O. Qiu, and L. Li, Shanghai Daxue Xuebao-Ziran Kexueban, 7, 331
333 (2001).
REFERENCES 649
209. Y. Shimada, S. Nakai, M. Suenaga, A. Sugihara, M. Kitano, and Y. Tominaga, J. Amer.

Oil Chem. Soc., 77, 10091013 (2000).
210. C. J. Chang, Y. Chang, H. Lee, J. Lin, and P. Yang, Ind. Eng. Chem. Res., 39, 45214525
(2000).
211. C. E. Sumner, Jr., S. D. Barnicki, and M. D. Dolfi (to Eastman Chemical Co.), U.S. Patent
5,424,457, June 12, 1995.
212. P. Golbitz, Soya & Oilseed Blue Book, Soyatech, Inc., Bar Harbor, Maine, 2000.
213. R. S. Daniels, Can. Patent 1,256,449, June 27, 1989.
214. R. Stern, G. Hillion, P. Gateau, and J. C. Guibet, in A. R. Baldwin, ed., Proceeding:
World Conference on Emerging Technologies in the Fats and Oils Industry, AOCS Press,
215. H. N. Basu, and M. E. Norris (to Agricultural Utilization Research Institute), U.S. Patent
5,525,126, June 11, 1996.
216. M. J. Haas and K. M. Scott, J. Amer. Oil Chem. Soc., 73, 13931401 (1996).
217. M. J. Haas, S. Bloomer, and K. Scott, J. Amer. Oil Chem. Soc., 77, 373379 (2000).
218. M. J. Hass, K. M. Scott, P. J. Michalski, and S. Runyon (to Runyon Industries), U.S.
Patent 2,003,158,074, August 21, 2003.
219. P. Golbitz, 2004 Soya & Oilseed Bluebook, Soyatech, Inc., Bar Harbor, Maine, 2004.
220. D. R. Kodali, INFORM, 14, 121123 (2003).
221. H. Eierdanz, in T. Applewhite, ed., Proceedings of the World Conference on Oilseed
222. L. Crandall, INFORM, 13, 626631 (2002).
223. Annonymous, Oils and Fats, 17(6), 1819 (2001).
224. F. Sliney, INFORM, 12, 1018 (2001).
225. G. Knothe, INFORM, 13, 900903 (2001).
226. R. Ohlson, in T. Applewhite, ed., Proceedings of the World Conference on Oilseed
227. American Soybean Association, St. Louis, MO. (2003). http://www.soystats.com/2003/
page_24.htm.
228. Yearbook and Trading Rules 20032004, National Oilseed Processors Association,
Washington, D.C., 2003.
229. E. F. Sipos and B. F. Szuhaj, in Y. H. Hui, ed., Baileys Industrial Oil and Fat Products,
vol. 2, 5th ed., John Willey & Sons, Inc., New York, 1996, pp. 497601.
230. E. R. Sherwin, J. Amer. Oil Chem. Soc., 53, 430436 (1976).
231. Q. Chen, H. Shi, and C. Ho, J. Amer. Oil Chem. Soc., 69, 9991002 (1992).
232. National Archives and Records Administration, Federal Register, 68, (133), (2003).
Available: http://fda.gov/OHRMS/DOCKETS/98fr/03 17525.htm.
233. W. W. Nawar, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of Fats and Oils,
234. G. Paquette and D. B. Kupranycz, Can. Inst. Food Sci. Technol. J., 18, 197206 (1985).
235. I. P. Freeman, F. B. Padley, and W. L. Sheppard, J. Amer. Oil Chem. Soc., 50, 101103
(1973).
236. M. A. Augustin and S. K. Berry, J. Amer. Oil. Chem. Soc., 60, 15201523 (1983).
650 SOYBEAN OIL
237. M. H. Gordon and L. Kourimska, Food Chem., 52, 175 (1994).

238. P. J. White, C. P. Su, and M. Gupta, in American Oil Chemists Society annual meeting
abstract book, pp. 4546 (2003).
239. K. Warner and S. Knowlton, J. Amer. Oil Chem. Soc., 74, 13171322 (1997).
240. C. P. Su and P. J. White, in American Oil Chemists Society annual meeting abstract
book, p. 132 (2003).
241. Code of Federal Regulations, Food and Drug Administration, Washington, D.C.
242. H. J. Basset, Cereal Foods World, 24(1), 79, 24, 25, 29, 30 (1979).
243. A. Moustafa, in D. Erickson, ed., Practical Handbook of Soybean Processing and
244. J. A. Segura, M. L. Herra, and M. C. Anon, J. Amer. Oil Chem. Soc., 72, 375378 (1995).
245. A. J. Haighton, J. Amer. Oil Chem. Soc., 36, 345348 (1959).
246. F. Cho, J. M. deMan, and O. B. Allen, J. Food Lipids, 1, 2552 (1993).
247. E. A. Emken, Ann. Rev. Nutr., 4, 339376 (1984).
248. V. DSouza, L. deMan, and J. M. deMan, J. Amer. Oil Chem. Soc., 68, 153162 (1991).
249. R. Wood, in C. K. Chow, ed., Fatty Acids in Foods and Their Health Implications,
Marcell Dekker, New York, 1992, pp. 663688.
250. R. P. Mensink and M. B. Katan, N. Engl. J. Med., 323, 439445 (1990).
251. H. B. Hu, M. J. Stampfer, J. E. Manson, E. Rimm, G. A. Colditz, B. A. Rosner, C. H.
Hennekens, and C. W. Willet, N. Engl. J. Med., 337, 14911499 (1997).
252. G. R. List, E. A. Emken, W. F. Kwolek, T. D. Simpson, and H. J. Dutton, J. Amer. Oil
Chem. Soc., 54, 408413 (1997).
253. H. H. Hustedt, J. Amer. Oil Chem. Soc., 53, 390392 (1976).
254. M. A. M. Zeitoun, W. E. Neff, G. R. List, and T. L. Mounts, J. Amer. Oil Chem. Soc., 70,
467471 (1993).
255. C. Hoy and X. Xu, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel
256. M. D. Erickson, in D. R. Erickson, ed., Practical Handbook of Soybean Processing and
257. G. R. List, T. Pelloso, F. Orthoefer, M. Chrysam, and T. L. Mounts, J. Amer. Oil Chem.
Soc., 72, 383384 (1995).
258. G. R. List, T. L. Mounts, F. Orthoefer, and W. E. Neff, J. Amer. Oil Chem. Soc., 73, 729
732 (1996).
259. G. R. List, T. Pelloso, F. Orthoefer, K. Warner, and W. E. Neff, J. Amer. Oil Chem. Soc.,
78, 103104 (2001).
260. L. L. Kok, W. R. Fehr, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 76,
11751181 (1999).
261. C. E. Stauffer, Fats and Oils, Eagan Press, American Association of Cereal Chemists,
St. Paul, Minnesota, 1996.
262. W. M. Smith, Bakers Dig., 53(8), 810 (1979).
263. H. Birnbaum, Bakers Dig., 52(2), 2836 (1978).
264. A. E. Thomas, J. Amer. Oil Chem. Soc., 55, 830833 (1978).
265. F. R. Paulicka, J. Amer. Oil Chem. Soc., 53, 421424 (1976).
REFERENCES 651
266. R. D. OBrien, in D. Erickson, ed., Practical Handbook of Soybean Processing and

267. J. H. Hetrick, J. Amer. Oil Chem. Soc., 46, 58A, 60A, 62A (1969).
268. S. A. Schlicker and C. Regan, Topics Clin. Nutr., 6, 5060 (1990).
269. T. Nagao and T. Teramoto, Rinsho Eiyo, 99, 414418 (2001).
270. M. G. Soni, H. Kimura, and G. A. Burdock, Food Chem. Toxicol., 39, 317329 (2001).
271. N. Matsuo and I. Tokimitsu, INFORM, 12, 10981102 (2001).
272. T. Yasukawa and K. Yasunaga, J. Oleo Sci., 50, 427432 (2001).
273. C. Watkins, INFORM, 14, 70 (2003).
274. Anonymous, Oils and Fats, 17(1), 1819 (2001).
275. C. Su, Fatty Acid Composition of Oils, Their Oxidative, Flavor, and Heat Stabilities and
the Resultant Quality in Foods, Ph. D. Dissertation, Parks Library, Iowa State University,
Ames, Iowa, (2003).
276. Y. H. Hui, in Y. H. Hui, ed., Baileys Industrial Oil & Fat Products, vol. 2, 5th ed., John
Wiley & Sons, Inc., New York, 1996, pp. 497601.
277. T. H. Smouse, J. Amer. Oil Chem. Soc., 56, 747A750A (1979).
278. H. J. Dutton, in K. Warner and N. A. M. Eskin, eds., Methods to Assess Quality and
Stability of Oils and Fat-Containing Foods, American Oil Chemists Society, Cham-
paign, Illinois, 1995, pp. 116.
279. E. N. Frankel, in D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and R. A. Falb,
1980, pp. 229244.
280. E. H. Farmer, G. F. Bloomfield, A. Sundralingam, and D. A. Sutton, Trans Faraday Soc.,
38, 348356 (1942).
281. M. B. Korycka-Dahl and T. Richardson, CRC Critical Reviews in Food Science and
Nutrition, 10, 209241 (1978).
282. P. J. Ke and R. G. Ackman, J. Amer. Oil Chem. Soc., 50, 429435 (1973).
283. D. J. Carlsson, T. Suprunchuk, and D. M. Wiles, J. Amer. Oil Chem. Soc., 53, 656660
(1976).
284. J. M. deMan, in C. K. Chow, ed., Fatty Acids in Foods and Their Health Implications,
285. R. J. Hsieh, in C.-T. Ho and T. G. Hartman, eds., Lipids in Food Flavors, ACS
Symposium Series 558, American Chemical Society, Washington, D.C., 1994, pp.
3048.
287. P. J. White, in R. D. OBrien, W. E. Farr, and P. J. Wan, eds, Fats and Oils Technology,
2nd ed., AOCS Press, Champaign, Illinois, 2000, pp. 341353.
288. H. W. Gardner, in H. W.-S. Chan, ed., Autoxidation of Unsaturated Lipids, Academic
Press, Inc., London, 1987, pp. 5194.
289. E. N. Frankel, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of Fats and Oils,
292. M. D. Dixon and E. G. Hammond, J. Amer. Oil Chem. Soc., 61, 14521456 (1984).
652 SOYBEAN OIL
293. I. Lee, S. H. Fatemi, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 72, 539
546 (1995).
294. J.-L. Kao, E. G. Hammond, and P. J. White, J. Amer. Oil Chem. Soc., 75, 1103 (1998).
295. L. Liu, and E. G. Hammond, J. Food Quality, 23, 521 (2000).
296. K. Liu, Inform, 10, 868878 (1999).
297. M. H. Gordon, in B. J. F. Hudson, ed., Food Antioxidants, Elsevier Applied Science,
New York, 1990, pp. 1318.
298. J. Loliger, in D. B. Min and T. H. Smouse, eds., Flavor Chemistry of Lipid Foods, AOCS
299. S. J. Bishov, A. S. Henick, J. W. Giffee, I. T. Nil, P. A. Prell, and M. Wolf, J. Food Sci., 36,
532535 (1971).
300. D. Firestone, ed., Official Methods and Recommended Practices of the American Oil
Chemists Society, 4th ed., AOCS Press, Champaign, Illinois, 1993.
301. D. L. Hamm, E. G. Hammond, V. Parvanah, and H. E. Snyder, J. Amer. Oil Chem. Soc.,
42, 920922 (1965).
302. C. Paqout, ed., Standard Methods for the Analysis of Oils, Fats and Derivatives, 7th ed.,
Pergamon Press, New York, 1987.
303. G. Yildiz, R. L. Wehling, and S. L. Cuppett, J. Amer. Oil Chem. Soc., 80, 103107
(2003).
304. P. J. White, in K. Warner and N. A. M. Eskin, eds., Methods to Assess Quality and
Stability of Oils and Fat-Containing Foods, AOCS Press, Champaign, Illinois, 1995,
pp. 159178.
305. I. H. Rukunudin, P. J. White, C. J. Bern, and T. B. Bailey, J. Amer. Oil Chem. Soc., 75,
563568 (1998).
306. G. Billek, G. Guhr, and J. Waibel, J. Amer. Oil Chem. Soc., 55, 728732 (1978).
307. J. I. Gray, J. Amer. Oil Chem. Soc., 55, 539546 (1978).
308. J. Pawliszyn, in J. Pawliszyn, ed., Solid Phase Microextraction: Theory and Practice,
Wiley-VCH, Inc., New York, 1997, pp. 97140.
309. H. H. Jelen, M. Obuchowska, R. Zawirska-Wojtasiak, and E. Wasowicz, J. Agric. Food
Chem., 48, 23602367 (2000).
310. D. D. Roberts, P. Pollien, and C. Milo, J. Agric. Food Chem., 48, 24302437
(2000).
311. C. D. Evans, G. R. List, R. L. Hoffman, and H. A. Moser, J. Amer. Oil Chem. Soc., 46,
501504 (1969).
312. R. G. Scholz and L. R. Ptak, J. Amer. Oil Chem. Soc., 43, 596599 (1966).
313. H. P. Dupuy, S. P., Fore, and L. A. Goldblatt, J. Amer. Oil Chem. Soc., 50, 340342
(1973).
314. H. P. Dupuy, E. T., Rayner, and J. I. Wadsworth, J. Amer. Oil Chem. Soc., 53, 628631
(1976).
315. K. Warner and T. J. Nelsen, J. Amer. Oil Chem. Soc., 73, 157166 (1996).
316. G. Paquette and D. B. Kupranycz, Can. Inst. Food Sci. Technol. J., 18, 197206 (1985).
317. G. R. List, in D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and R. A. Falb,
1980, pp. 267354.
REFERENCES 653
318. Iowa State University Extension. (2003). Available: http://www.extension.iastate.edu/

Publications/4H956.pdf.
319. K. G. Stern, and Grossman H. S., (to H. Sobotka), U.S. Pat. 2,842,577 (1958).
320. J. M. DeMan, in C. K. Chow, ed., Fatty Acids in Foods and Their Health Implications,
321. A. Bonanome and S. M. Grundy, N. Engl. J. Med., 318, 12441248 (1988).
322. S. M. Grundy and M. A. Denke, J. Lipid Res., 31, 11491172 (1990).
323. S. H. Fatemi and E. G. Hammond, Lipids, 15, 379385 (1980).
324. S. M. Grundy, L. Florentin, D. Nix, and M. F. Whelan, Amer. J. Clin. Nutr., 47, 965969
(1988).
325. R. T. Holman, S. B. Johnson, and P. L. Ogburn, Proc. Natl. Acad. Sci. USA, 88, 4835
4839 (1991).
326. W. E. M. Lands, in Fish and Human Health, Academic Press, Inc., Orlando, Florida,
1986.
327. P. M. Kris-Etherton, W. S. Harris, and L. J. Appel, Circulation, 106, 27472757
(2002).
328. J. M. King and P. J. White, in L. Jackson, ed., Impact of Processing on Food Safety,
Kluwer Academic/Plenum Publishers, New York, 1999, pp. 5165.
329. K. Almendingen, O. Jordal, P. Kierulf, B. Sandstad, and J. I. Pedersen, J. Lipid Res., 36,
13701384 (1995).
332. J.-L. Sebedio, A. Grandgiard, C. Septier, and J. Prevost, Rev. Francas Corps Gras, 34,
1518 (1987).
334. K. Almendingen, O. Jordal, P. Kierulf, B. Sandstad, and J. I. Pedersen, J. Lipid Res., 36,
13701384 (1995).
335. A. Grandgirard and F. Juilliard, Rev. Francais Corps Gras, 34, 213219 (1987).
336. A. Grandgirard, A. Piconneaux, J. L. Sebedio, S. F. OKeefe, E. Semon, and J. L.
Lequere, Lipids, 24, 799804 (1989).
337. J. M. Chardigny, J. L. Sebedio, and A. Granddirard, in A. Sinclair and R. Gibson, eds.,
Essential Fatty Acids and Eicosanoids, AOCS Press, Champaign, Illinois, 1992,
pp. 148152.
338. L. Bretillon, J. M. Chardigny, J. P. Noel, and J. L. Sebedio, J. Lipid Res., 39, 22282236
(1998).
339. J. R. Judd, B. A. Clevidence, R. A. Muesing, J. Wittes, M. E. Sunkin, and J. J. Podczasy,
Amer. J. Clin. Nutr., 59, 861868 (1994).
340. L. Kohlmeier and M. Mendez, Proc. Nutri. Soc., 56, 369383 (1997).
341. M. G. Enig, S. Atal, M. Keeney, and J. Sampugna, J. Amer. College Nutr., 9, 471486
(1990).
342. D. M. Klurfeld and Kritchevsky, INFORM, 3, 515516 (1992).
343. U.S. Department of Agriculture and U.S. Department of Health and Human Services,
Nutrition and Your Health: Dietary Guidelines for Americans, 3rd ed., Home and
Garden Bulletin No. 232, Washington D.C., 1990.
14
Sunflower Oil
Maria A. Grompone
1. HISTORICAL REVIEW
Sunflower (Helianthus annuus L.), one of the most ancient oilseed species in North
America, belongs to the family Compositae (Asteraceae) and the genus Helianthus.
Cultivation of sunflower dates from times earlier than 3000 B.C., as indicated by
archeological evidence obtained in sites once inhabited by the Hopi Indians, in
the north of Arizona.
According to other archaeological findings and traditional tales, sunflower was
cultivated by indigenous people throughout central North America (from New
Mexico to the Dakotas) and eastward (Pennsylvania and Ontario). It appears that
sunflower was domesticated in America even before corn was. Single-head plants
were preferred by the indigenous people, who differentiated them from multiple-
head plants growing wildly. Some tribes ate fruits directly, or ground into a meal
that was baked in the form of bread. Other tribal practices included boiling of
heads, and crushed roots, for the treatment of disease and bites. The oil extracted
from seeds was used as body and hair ointment. Seeds, petals, and pollen were used
in the preparation of facial and body makeup, and for dying cloth and utensils (12).
Sunflower was introduced into Europe by the Spanish explorers returning to the
continent at the beginning of the 1500s A.D. The first scientific review of American
plants was made by Sevillan doctor Nicolas Bautista Monardes (15081588), who
wrote Historia medicinal de las cosas que se traen de nuestras Indias Occidentales
(A medical review of things brought from the West Indies, published in Seville in
655
656 SUNFLOWER OIL
three volumes, in 1565, 1571, and 1574). It was in this study that sunflower was first
mentioned. Monardes, who never traveled to America, described those plants that
reached him and which he grew in a Botanic Garden designed for this purpose. He
also gathered information that he obtained from navy captains, missionaries, and
travelers.
There is also a description of sunflowercontemporaneous with the work of
Monardesin a herbarium made by Rembert Dodoens in 1568. Even though this
and other later herbaria attribute the origin of sunflower to Peru and Central
America, it is now believed to have originated in North America.
Starting from Spain, sunflower crops spread rapidly through France and Italy,
and toward the north and east of Europe. In several regions, it was a source of
smoking leaves, flowers for consumption in salads, or for the manufacture of paint,
edible, and medicinal seed, and cooking oil. But it was, perhaps, the beauty in the
inflorescence of sunflowers that interested the first growers, large and bright yellow,
always facing the sun. Hence, the name of the genus, Helianthus, derived from the
Greek helios meaning sun and anthos meaning flower; and its Spanish, English,
French, and German words: girasol, sunflower, tournesol and Sonnenblumen.
It was not until the eighteenth century that sunflower seeds were used as oilseed.
According to records, the first patent of oil extraction for industrial use was granted
to Arthur Bunyan in 1716 in England, where reference is made to an English seed
that could be pressed yielding sweet oil of great value for those interested in the
manufacture of wool, paint, leather, and so on. Seeds of double- and single-head
flowers, known as sunflowers, were indicated for oil extraction (2).
In Russia, it was introduced by Peter I the Great, Czar between 1682 and 1725,
who, having seen sunflowers in the Netherlands, took seeds to Russia. It was in
Russia where the most important development took place in the use of sunflower
as both food and oil source. The Russian Orthodox Church banned the consumption
of several foods during Lent and Advent (periods of the religious calendar dedi-
cated to fasting and penitence), including several sources of oil. As the ban did
not include sunflower seeds, they were adopted as an oil source.
Rapidly, sunflower spread through Russia, the earliest records of cultivation dat-
ing from 1770. The extraction of oil from sunflower seeds was first suggested in
1779, according to Russian Academy proceedings. The cultivated area increased
rapidly as a result of the development of the sunflower oil extraction industry,
Russia being the worlds first and largest sunflower producer until current times.
Once the value of the crop had been recognized, commercial production was started
in 1880 over 150,000 hectares, a figure that reached one million hectares toward
1910. Pioneering and fundamental research work has been carried out in Russia
since 1860 concerning the improvement of seed for oil content.
Toward the end of the nineteenth century, improved Russian cultivars were intro-
duced in the Balkans leading to the expansion of sunflower crops. The crop did not
reach northern Europe owing to the lack of cultivars adapted to cold climates.
Sunflower was the main Russian crop already at the beginning of the twentieth
century. In 1912, scientist V. S. Pustovoit started research work in the fields of the
Kuban region. Krasnodar was Russias experimental oilseed selection center, since
HISTORICAL REVIEW 657
1924. The Pustovoit All Union Research Institute was founded in 1932 and named
after V. S. Pustovoit for his valuable contributions (Pustovoit was in charge of the
Breeding Department until his death in 1972). Pustovoits work led to an improve-
ment in the oil content and seed yield. The average oil content of a Russian cultivar
was 330-g oil per kilogram seed in 1940, reaching values as high as 550 g per kilo-
gram in strains developed by Pustovoit in 1965.
Hybridization of sunflower resulting from natural cross-breeding, performed in
seed-producing fields with parents planted in alternating lines, led to major
advances in research. It enabled improvements of yield in USSR cultivars and rapid
disease control, as well as increases of oil content and other issues of agronomical
interest. Most remarkable among these open-pollinated varieties was the Peredo-
vik, named after the Russian agronomist. Two major events in the 1960s had a
marked effect on the sunflower industry worldwide: the introduction of USSR cul-
tivars of high oil content, and the discovery of cytoplasmic male sterility and ferti-
lity-restoring genes. Male sterile sunflowers were obtained in 1968 by Leclercq
from the offspring of an interspecific hybrid between the cultivated sunflower
and wild sunflower Helianthus petiolaris. The identification of fertility-restoring
genes of several breeders led to hybrids of special characteristics. Open-pollinated
cultivars were rapidly replaced by hybrids of higher yield, uniformity, and disease
resistance. Currently, hybrid seeds are widely used for cheap and efficient produc-
tion throughout the world (3).
Sunflower crops cultivated in North America are derived from seeds introduced
by eastern European immigrants toward the end of the nineteenth century; hence,
the name Russian Peanuts. Russian emigrants in the United States and Canada
grew strains such as Giant or Mammoth Russian in gardens for the production of
edible seeds. These served as a base for the development of improved cultivars for
commercial production. The cultivated area in the United States reached 200,000
acres in 1968; most of which was destined to the production of seed for manufac-
ture of food for human consumption, and to the bird meal market (4).
An open-pollinated Russian-bred cultivar of high oil content (Peredovik variety,
4045%) was introduced in the United States in 1966 (3). Commercial production
of oilseed-type sunflower was started with the Peredovik variety among other cul-
tivars, and since 1966, several research programs in the United States have sought
to improve sunflower hybrids for oil yield.
Around 1960, the USSR interrupted the supply of sunflower oil to Europe,
because of the high internal demand, including satellites, thus leaving an unat-
tended sector in the European market, where consumption of tallow and butter
were then indicated as causes of coronary disease. The high content of polyunsatu-
rated fatty acids (PUFA) of sunflower oil naturally interested many American and
Canadian oil industries, with the consequent increase in sunflower production in the
late 1970s (1).
Russian immigrants carrying sunflower seeds introduced the crop into Argentina
in the nineteenth century, for human consumption of seeds. Cultivation of the crop
was performed at small-scale initially, and it was not until the world economic
crisis of 1930 that it was first sown intensively to supply the internal market, in
658 SUNFLOWER OIL
replacement of imported oils. Around 1500 metric tons (MT) of sunflower oil were
produced in Argentina in 1930, a figure that reached 5000 MT in 1945. New, disease-
resistant varieties were developed as a result of the work of Experimental Stations
of Argentinean National Institute of Agrarian Technology (INTA) and of private
seed breeders. The appearance of hybrids characterized the Argentinean market
in the period after 1975, although the first hybrid had been launched in 1972.
Almost 100% of the cultivated area is currently sown with hybrids. A higher
seed and oil production capacity, together with the introduction of specialized
upgraded technology, led to an increase in oil yield per hectare in Argentinean plan-
tations. Seed yield levels increased from 0.73 tons per hectare in 19771978
2,200,000 hectares of sunflower plantations producing 1,600,000 tons of seed
to 1.38 tons per hectare in 1987-19882,117,000 hectares (2,915,000 tons) (1,2).
2. SUNFLOWER CROPS
2.1. General Characteristics

The genus Helianthus comprises 68 known species divided into two major separate
groups: the North American and the South American species. North American sun-
flower species spread throughout the United States, reaching Canada and Mexico.
Both groups do not seem to relate to each other; in South America, they appear to
have originated by parallel evolution of the genus Viguiera (1).
Sunflower is a highly cross-pollinated crop. Wild sunflowers have several flowers
or heads and depend on the work of insects for pollination. Wild sunflowers are the
genetic base of current commercial sunflowers of a single flower or head per plant.
Sunflower is an annual crop. Plants reach 13 m in height. The head is composed
of a large number of tiny flowers that are tubular in shape (700 to 4000 single blos-
soms) forming a disk, those in the outer row having long strap-shaped corollas that
form the rays of the composite flower. Plants have a large number of flowers clus-
tered in a capitulum, inflorescence, or head. The back of the head is covered with
small green bracts. Radial structures in the shape of petals are displayed over the
bracts. These are known as ray flowers, and they do not have a reproductive func-
tion other than serving as a signal for bees and other pollinizer insects. Toward the
disk center, are a large number of complete tiny flowers known as disk florets. Each
of these flowers is capable of bearing an achene or seed (a fruit from a strict bota-
nical viewpoint).
Those flowers that form the rays are generally sterile, and although they have
vestigial styles and stigmas, they do not possess anthers. Flowers yielding seed
are complete, each with a tubular corolla and an anther. Sunflower heads will follow
the sun cycle until practically all flowers comprising the head have been pollinated.
After that, they remain in a fixed position, facing eastward. Around 70 days are
required from sowing to flowering of the crop. Seeds reach maturity at 130 days
and can be harvested 10 days later (45).
Sunflower grows in moderate climates (temperate to temperate-hot), especially
in America, Europe, and China, predominantly at temperatures between 20 C and
SUNFLOWER CROPS 659
25 C, with an optimum temperature of 2728 C. It grows well in dry, sunny

weather, in deep soils capable of supplying abundant water. The oil content of seeds
is lower in regions of extreme heat. The crop has high resistance to temperature
fluctuations between night and day varying between 8 C and 34 C (1, 5).
The highest yield in seed is achieved at temperatures between 18 C and 25 C
through the period from formation to filling of the seed. Humidity conditions are
critical during both 1520-day periods prior to and after flowering. Improved filling
of the seed takes place in periods without rain. Pollination is nearly all cross-type,
i.e., from one flower to another. Crossing within one head is scarce. Insects, in par-
ticular bees, are the main fertilization agent (2).
Climatic conditions in Argentina are ideal for cultivation of sunflower in view of
the varying degrees of influence of the Atlantic Ocean. The buffering of thermal
extremes between summer and winter (mean values in winter and summer of
8 C and 28 C), the circulation of east winds (allowing a rainfall range from 500
to 1100 mm annually), and the span of the period free of frost (over 6 months)
are major climatic factors contributing to ensure the establishment and success of
the cycle of sunflower (6). In Argentina, the sunflower is grown between latitudes
26 S (Chaco) and 39 S (southern Buenos Aires), over an area averaging 2.93 mil-
lion hectares for 19902000 summer seasons. The cultivation environments include
subtropical (northern Argentina) and temperate (central and southern) climate (7).
Sunflowers are ripe when the back of the head has turned from green to yellow
and the bracts are turning brown. Harvest is done when the seed reaches commer-
cial ripeness, that is, allowing time for the seed to dry from a 35% moisture content
of heads at physiological maturity down to 11%. The harvest of seed-loaded, heavy
fruits is advanced to prevent seeds falling off (2, 8).
Drying agents such as magnesium chlorate may be used as an aid to advance
harvesting. This practice readily reduces the moisture content of heads, stems,
and leaves, but seeds retain most of their moisture. Artificial drying of seeds is often
necessary prior to storage.
Seeds must be stored with moisture levels lower than 9.5% to avoid undesired
enzymatic reactions. Some of these reactions start within 12 hours after harvesting
for seeds with moisture higher than 20% (5). Seeding and harvesting periods
obviously differ according to hemisphere of producer country. For Argentina and
Uruguay, in the South Hemisphere, seeding is in October and harvesting between
February and April, and in Australia, harvesting is between January and May. For
countries in the Northern Hemisphere, such as ex-USSR, seeding is done in March
and harvesting is between August and October; in the United States seeding is in
April or May and harvesting is in September or October; in Spain and Italy, harvest-
ing is in August (1, 5).
2.2. Yield of Sunflower Crops

Seed yield varies according to region. Maximum yield values in 1994 were
obtained in Italy (2596 kg/ha), Greece (2577 kg/ha), and Austria (2544 kg/ha);
minimum values were obtained in Tanzania (370 kg/ha). These values have been
660 SUNFLOWER OIL
2000
1800
world
1600 Argentina
1400 USA
Seed yield (kg/ha)
exUSSR
1200
1000
800
600
400
200
0
1930 1940 1950 1960 1970 1980 1990 2000
Figure 1. Evolution of seed yield (kg/ha) in the most important producer countries, as well as
world average values (1, 5).
increased in the last years as a result of genetic improvements. Figure 1 shows the
evolution of yield levels in major producer countries, as well as world average
values (1, 5).
2.3. Structure of Sunflower Seeds

An achene, the seed of sunflower, is pointed at the base and rounded at the top. Seed
size ranges between 10 and 15 mm in length and between 4 and 12 mm in width,
appearing to be four-sided in cross-section. The outer layer, the pericarp or hull,
represents 1845% of the total achene weight. The white papery layer immediately
beneath the pericarp, the testa or seedcoat, is made up of three parenchyma layers,
the inner layer being spongy in texture. The endosperm comprises a single layer
of cells rich in protein, firmly attached to the hull, and an embryo, commonly referred
to as kernel. The embryo consists of two cotyledons attached to a protruding
radicle (9).
There are two basic types of sunflower: (1) oilseed type and (2) nonoil type, the
latter supplying the bird meal and confectionary markets. The first hybrid oilseed
types bore small black seeds with a thin hull (representing 2025% of total seed
weight) with a 40% oil content. The non-oilseed type is somewhat different; it
has a larger seed with a thicker black-and-white-striped hull (representing 40
45% of total seed weight), which is weakly attached to the kernel and can easily
be removed. These seeds contain 30% of oil.
The size of oil-type seeds varies according to cultivar and according to the seeds
position in the head, those on the periphery being larger. Besides affecting the oil
content, the position of seeds in the head influences the fatty acid composition.
SUNFLOWER CROPS 661
ash oil protein

3% 3% 3%
N-free extract
31%
fiber
61%
Figure 2. Hull composition (1, 5, 11).
Internal seeds contain less oil than those in intermediate or external zones. The con-
tent of linoleic and palmitic acids increases, and the content of oleic acid decreases
from the perimeter toward the flower head center (10).
One thousand seeds of most currently used hybrids weigh from 30 to 80 g. Hull
color ranges from completely white to black, with gray- or brown-striped inter-
mediates. Both hull thickness and structure, as well as other seed characteristics,
depend on variety and ambient growth conditions (9).
The hull is mainly composed of fibrous substances, lignin and cellulosic materi-
als in equal proportions. Kernels of oilseed-type sunflower contain nearly all of the
oil of seeds, besides proteinaceous substances and carbohydrates. The kernel repre-
sents 70% of the seed, with an oil content of approximately 55%, amounting to 40%
with respect to the whole seed. The protein content ranges between 20% and 35%,
amounting to up to 57% on a water-and-oil-free basis (1, 5, 11).
A commonly occurring hull and kernel composition is shown in Figures 2 and 3,
respectively, as well as of sunflower meal in Figure 4. Data correspond to a fully
dehulled meal, a condition difficult to obtain in practice (1, 5, 11).
2.4. The Influence of Ambient Factors on Sunflower Seed Oil

The oil content of sunflower seeds varies during the development of the seed:
increasing from the fourteenth to the thirty-fifth day after flowering, when the
seed is physiologically mature. The oil content remains steady after reaching matur-
ity. Oil composition also changes during the formation and ripening stages of the
seed. The linoleic acid content increases from the fourteenth day after flowering
while the oleic acid content decreases; also, saturated fatty acids decrease slightly
(12).
662 SUNFLOWER OIL
ash
N-free extract
3%
6%
fiber
3%
protein
21%
oil
67%
Figure 3. Kernel (dehulled seed) composition (1, 5, 11).
Ambient factors, such as temperature and light, affect the oil composition of sun-
flower seeds. Robertson and Russell (13) studied the effect of climatic conditions
(temperature difference between night and day in Canada, Minnesota, and California)
on the composition of sunflower oil, finding that linoleic acid increased propor-
tionally with increasing temperature difference. Robertson and Green (14) studied
the effect of sowing time on oil content and composition. Eleven different hybrids
fiber
9%
N-free extract
protein 19%
63 %
ash
8%
oil
1%
Figure 4. Composition of fully dehulled meal (1, 5, 11).

SUNFLOWER CROPS 663
were used. Sowing was performed in February and August in Florida. The results
show that the oleic acid content is lower (average: 19.4%) and that the linoleic acid
content is higher (average: 68.4%) for seeds sown in August (lowest mean tempera-
ture from flowering to maturity, 18 C). Seeds harvested in plantations sown at the
beginning of April (highest mean temperature from flowering to maturity, 27 C)
have a lower content of linoleic acid (average: 36.3%) and a higher average content
of oleic acid (54.6%). This indicates that adjustments in seeding time in Florida
may lead to sunflower oil of different compositions.
Average temperature from flowering to ripening appears to be a major factor
affecting the fatty acid composition of sunflower oil (15). The oleic/linoleic ratio
was correlated with average temperature for different sunflower plantations across
Spain, average temperature being related not only to latitude, but also to other geo-
graphic factors that determine a microclimate of the crops. An increase in tempera-
ture resulted in a decrease in the linoleic acid content and an increase in the oleic
acid content. The linoleic acid content ranged from 48.7% in seeds grown in war-
mer weather in the south of Spain to 70.2% for colder weather plantations. The
addition of the contents of linoleic and oleic acids is not constant with temperature,
the increase in oleic acid being greater than the decrease in linoleic acid, with a
partial compensation by a reduction in the stearic acid content, suggesting the con-
version of stearic acid to oleic acid. The above mechanism is related to the effect of
temperature on the activity of desaturase enzymes converting oleic into linoleic
acid.
Research carried out with sunflower crops grown in controlled temperature
chambers 20/10 C (day/night) and 30/20 C (day/night) with equal light intensity
and photoperiod, demonstrated a decrease in the linoleic/oleic ratio with increasing
temperature (16). An increase in temperature also leads to a slight decrease in the
stearic acid content. Similar results were obtained by authors in different countries,
like France, Italy, Japan, the United States, Canada, and so on (17). This is because
those enzymes involved in the sequence of steps leading to the formation of linoleic
acid are alike in all higher plants; that is, all oilseed crops have a common enzyme
base catalyzing the synthesis of fatty acids, producing varying amounts of 16:0,
18:0, 18:1, 18:2, and 18:3.
Among a number of studies carried out in different regions worldwide, the com-
position of sunflower seeds was determined for six different Argentinean regions
(18). For the latitudes considered, according to region, mean ambient temperature
during the development period of seed decreased from 28 C to 20 C southward.
The linoleic acid content was found to decrease, and the oleic acid content
increased with increasing temperature. A similar behavior was observed for sun-
flower seeds grown in Japan (19).
2.5. Sunflower Associations

The International Sunflower Association (ISA), with main offices in Paris aims to
enhance international cooperation toward the improvement of cultivation, growth,
and technical and nutritional levels, besides promoting and facilitating close
664 SUNFLOWER OIL
cooperation relations among researchers. ISA holds an International Conference on

Sunflower every four years.
In addition, there is a national association in different countries. The U.S.
National Sunflower Association (NSA) with main offices in Bismarck, North Dakota,
promotes the production and trade of sunflower products both at national and inter-
national levels. Founded in 1981, it currently has 20,000 members. The Australian
Sunflower Association was established in 1976; the membership of the ASA
consists of growers, researchers, and personnel from all facets of the industry. The
National Sunflower Association of Canada (NSAC) was founded in 1996, with
105 members in 1999. The Argentinean Sunflower Association (ASAGIR), created
in 1980, hosted the Eleventh International Conference on Sunflower in 1985.
3. CHEMICAL AND PHYSICAL PROPERTIES OF REGULAR

SUNFLOWER OIL
3.1. Composition of Regular Sunflower Oil

Sunflower oillike most vegetable oilsis composed mainly of triacylglycerols
(9899%), and a small fraction of phospholipids, tocopherols, sterols, and waxes
(all of the latter are commonly referred to as the unsaponifiable fraction).
3.1.1. Sunflower Fatty Acids Regular sunflower oil is characterized by a high

concentration of linoleic acid, followed by oleic acid. Saturated fatty acids (mainly
palmitic acid and stearic acid) do not amount to more than 15% of the fatty acid
content. Table 1 shows the variation range of major fatty acids in regular sunflower
oil (9, 20).
Two facts regarding the composition of regular sunflower oil are worth noting
from the nutritional viewpoint: It provides an essential fatty acid (linoleic acid),
and it has a low content of palmitic acid compared with other oils (palmitic acid
is believed to increase LDL-C in blood).
The reported composition of regular sunflower oil has changed with adjustments
of analytical methods and the samples considered. This is reflected in the variation
ranges approved successively by the Codex Alimentarius Comission. The values
approved in 1981 and 1993 (21) are compiled in Table 2, as well as the current
TABLE 1. Variation Range for Major Fatty Acids (%)

of Regular Sunflower Oil (9, 20).
Fatty Acid AOCS (20) Merrien (9)
16:0 58 57
18:0 2.57.0 46
18:1 1340 1525
18:2 4074 6270
18:3 <0.3 <0.2
CHEMICAL AND PHYSICAL PROPERTIES OF REGULAR SUNFLOWER OIL 665
TABLE 2. Variation Range for Fatty Acids (%) of Regular Sunflower Oil According
to Standards Approved by the Codex Alimentarius Commission in Different Years.
Fatty Acid 1981 1993 1999
12:0 ND0.1
14:0 <0.5 <0.2 ND0.2
16:0 310 5.67.6 5.07.6
16:1 <1.0 <0.3 ND0.3
17:0 ND0.2
17:1 ND0.1
18:0 110 2.76.5 2.76.5
18:1 1465 1439.4 14.039.4
18:2 2075 48.374.0 48.374.0
18:3 00.7 00.2 ND0.3
20:0 01.5 0.20.4 0.10.5
20:1 00.5 00.2 ND0.3
22:0 010 0.51.3 0.31.5
22:1 00.5 00.2 ND0.3
22:2 ND 00.3 ND0.3
24:0 00.5 0.20.3 ND0.5
24:1 <0.5 ND ND
ND nondetectable, defined as 0.05%.
value of 1999 (Codex-Stan 210-1999). The Codex Alimentarius Commission

(Codex) was established in 1962 by two United Nations organizations, the Food
and Agriculture Organization (FAO) and the World Health Organization (WHO).
Codex is the major international organization for encouraging fair international
trade in food and protecting the health and economic interests of consumers. The
Codex Committee on Fats and Oils (CCFO) was established to elaborate worldwide
standards for fats and oils and their products. The Codex Alimentarius is thus taken
as reference.
According to the composition indicated by the Codex Alimentarius (Codex-Stan
210-1999), the saturated fatty acid content of regular sunflower oil is lower than
that in corn (maximum 22%), cottonseed (maximum 32%), peanut (maximum
28%), and soybean (maximum 20%) oils, and higher than the saturated content
of safflower (maximum 12%) and rapeseed (maximum 12%) oils. The linolenic
acid content (18:3) of regular sunflower oil is fairly low (always lower than
0.3%), giving the oil a good oxidative stability.
The variation ranges of fatty acids in regular sunflower oil have also changed in
several countries. The Canola Council of Canada revised the table of composition
of edible oils prepared in 1979, based on a study carried out by the POS Pilot Plant
Corporation in Saskatoon, Saskatchewan, Canada. POS analyzed ten vegetable oils
and three animal fats supplied by food processing and manufacturing enterprises of
Canada and the United States, according to one issue of Canada Digest (22). Aver-
age regular sunflower oil compositions are shown in Table 3 for Canada and the
United States, as well as for Argentina (2, 23).
666 SUNFLOWER OIL
TABLE 3. Average Composition (%) of Regular Sunflower Oil for Canada/United States
(normalized to 100%) and Argentina, Elaborated in Different Years (2, 22, 23).
Canada/U.S. 1979 Canada/U.S. 1994 Argentina 1981 Argentina 1998

Fatty Acid (22) (22) (23) (2)
saturated 11 12 8.7 10.1

18:1 20 16 24.0 26.8
18:2 69 71 66.0 62.2
3.1.2. Triacylglycerol Composition Figure 5 shows the composition in major

triacylglycerols (above 1%) of regular sunflower oil [based on Prevot (17)]. As
expected from its high linoleic acid content, the main triacylglycerol is trilinolein
(36.3%), followed by oleo-dilinolein (29.1%); triolein being practically nonexistent
(0.6%). Thus, the percentage of triacylglycerols (TAG) with four or more double
bonds is higher than 80%. This TAG distribution is responsible for the low solidi-
fication point of regular sunflower oil (16 C to 19 C), allowing, for example,
storage of mayonnaise manufactured with regular sunflower oil in a refrigerator
without breakage of the emulsion (unlike the case of other oils such as peanut oil).
Rossell et al. (24) analyzed the triacylglycerols of 20 regular sunflower oil sam-
ples, regarding the total number of carbon atoms. The content of 54-carbon TAG
was 75.179.5%. The composition of these triacylglycerols is OOO, SOL, OOL,
SLL, OLL, and LLL. Grouping the data provided by Prevot (17) in the same man-
ner, the 54-carbon TAG content would be 82.1%. Thus, the results of both works
are to a large extent in agreement.
40
35
30
25
Percentage
20
15
10
0
SOL POL OOL SLL PLL OLL LLL
Triacylglycerol
Figure 5. Composition in major triacylglycerols (above 1%) of regular sunflower oil [based on
(17)]. (Key: P palmitic acid, S stearic acid, O oleic acid, L linoleic acid.)
40
35
experimental
30
random
25
Percentage
20
15
10
0
SatOL OOL SatLL LLO LLL
Figure 6. Triacylglycerol composition of regular sunflower oil as calculated from random

distribution and experimentally determined (17). (Key: Sat saturated acid, O oleic acid,
L linoleic acid.)
The 1,2,3-random hypotheses assumes that one pool of fatty acids is randomly
distributed to all three positions of the glycerol molecules in an oil. The fatty acid
compositions of the sn-1, sn-2, and sn-3-positions would thus be equivalent. Figure 6
shows the theoretical composition of regular sunflower oil as calculated by the
equations of random distribution. Calculation of the random distribution was based
on the following composition: 11% saturated fatty acids (Sat), 20% oleic acid (O),
and 69% linoleic acid (L). The TAG composition of a regular sunflower oil deter-
mined experimentally is also shown; there is no indication of the overall fatty acid
composition (17). Differences between both compositions are not great, in particu-
lar, taking into account the fact that the fatty acid composition may differ for the
oils considered.
Fatty acids are not randomly distributed in natural oils. Saturated fatty acids are
almost exclusively concentrated in the sn-1,3 positions and are practically nonexis-
tent in the sn-2 position (taxonomic pattern). Linoleic acid clearly has a higher
occurrence in the sn-2 position. For example, out of a total 16.239.3% linoleic
acid in peanut oil, the sn-2 position has 27.267.8%, clearly showing the concen-
tration of linoleic acid in this position (24).
Table 4 shows the fatty acid distribution in the sn-2 position with respect to the
composition of the sn-1 and sn-3 positions (25) or with respect to the overall com-
position of the sunflower oil samples analyzed (24). Occurrence of 18:2 is slightly
higher in the sn-2 position than it would be if distributed in equal proportions
among all three positions. The occurrence of linoleic acid is also slightly higher
in the sn-2 position than in the sn-1,3 positions by a ratio of 1.27. As the content
of linoleic acid is particularly high in regular sunflower oil, the preferential distri-
bution of linoleic acid is less apparent than for other vegetable oils. Saturated fatty
668 SUNFLOWER OIL
TABLE 4. Distribution of Fatty Acids in the sn-2 Position with Respect to the sn-1 and
sn-3 Positions (25) or with Respect to the Overall Sunflower Oil Composition (24).
Alvarez-Ortega et al. (25) Rossell et al. (24)
sn-1 sn-3 sn-2 overall sn-2

16:0 9.2 0.5 5.76.9 0.20.4
18:0 6.1 0.4 3.06.3 0.10.3
18:1 34.0 34.7 14.034.4 12.131.3
18:2 50.7 64.2 55.573.2 66.287.4
acids have a tendency to concentrate in the sn-1 and sn-3 positions, hardly ever
occurring in the sn-2 position. Oleic acid occurs equally among all three positions.
However, differences are small, resulting in an apparent agreement between the
TAG profile determined experimentally and the fatty acid distribution calculated
on a supposed random distribution (as indicated in Figure 6).
3.1.3. Nonacylglycerol Components of Regular Sunflower Oil

3.1.3.1. Phospholipids The phospholipid content of crude sunflower oil ranges
between 0.5% and 1.2%. Oils extracted by solvent generally have a higher content
of phosphlipids than those obtained by pressing. Major phospholipids are phospha-
tidylcholine, phosphatidylethanolamine, phosphatidylinositol, and phosphatidic
acid. Most are hydratable and may be removed from the crude oil through a
water-degumming process (See Section 5.3.1.)
3.1.3.2. Tocopherols Tocopherols are heterocyclic compounds with a phenolic

group and a substituted side chain of branched hydrocarbon. Their high solubility
in oil is caused by a long alkyl side chain. The Codex Alimentarius (Codex-Stan
210-1999) indicates levels of tocopherols and tocotrienols in crude regular sun-
flower oil (mg/kg), compiled in Table 5.
TABLE 5. Levels of Tocopherols

and Tocotrienols in Crude Regular
Sunflower Oil (ppm), According
to the Codex-Stan 210-1999.
Content (ppm)
Alpha-tocopherol 403935
Beta-tocopherol ND45
Gamma-tocopherol ND34
Delta-tocopherol ND7.0
Alpha-tocotrienol ND
Gamma-tocotrienol ND
Delta-tocotrienol ND
Total 4401520
ND nondetectable.
The biological value of tocopherols differs according to isomer. Their

importance as Vitamin E activity is the following: alpha > beta > gamma > delta.
Vitamin E functions primarily as an antioxidant, especially in preventing oxidation
and peroxidation of polyunsaturated fatty acid units of membrane phospholipid
(within and on the plasma membrane of cells). The value of regular sunflower
oil as a source of Vitamin E is enhanced by a high content of alpha-tocopherol.
Similar conclusions can be found in the literature for tocopherols of Argentinean
sunflower oil: 700 ppm of total tocopherols, 91% of which corresponds to alpha-
tocopherol (2).
Tocopherols also function as free radical scavengers. The alpha form has the
highest Vitamin E activity, and gamma-tocopherol has the highest antioxidant activ-
ity. In one study, sunflower tocopherols were added to stripped soybean oil, and
soybean tocopherols were added to stripped sunflower oil. The stability pattern
of sunflower oilgenerally less stable than soybean oilmimicked that of soybean
oil. With the added sunflower tocopherols, the stability pattern of soybean oil
resembled that of sunflower oil. The results suggest that gamma-tocopherol is a
better antioxidant than the alpha isomer (26). Other authors, however, attribute a
higher antioxidant activity to alpha-tocopherol. Disagreement with respect to the
relative antioxidant activity of tocopherol homologs may be because of differences
in the degree of unsaturation of the substrates used, the degree of oxidation
achieved in the measurement, and the method of oxidation analysis.
Figure 7 shows a comparison of maximum values for major tocopherols in crude
regular sunflower, peanut, soybean, corn, and cottonseed oils, according to the
Codex Alimentarius (Codex-Stan 210-1999). Among these, regular sunflower oil
has the highest tocopherol level. In general, gamma-tocopherol is the most widely
occurring isomer in vegetable oils.
4000
3500
total alpha
3000
gamma delta
Content (ppm)
2500
2000
1500
1000
500
0
corn soybean sunflower peanut cottonseed
Figure 7. Maximum values of major tocopherols (ppm) of crude sunflower, peanut, soybean,
corn, and cottonseed oils, according to the Codex Alimentarius (Codex-Stan 210 -1999).
670 SUNFLOWER OIL
TABLE 6. Levels of Desmethylsterols

in Crude Regular Sunflower Oil,
as a Percentage of Total Sterols and Total
Sterol Content (ppm), According to
Codex-Stan 210-1999.
Cholesterol (%) <0.7

Brassicasterol (%) ND0.2
Campesterol (%) 7.412.9
Stigmasterol (%) 7.011.5
-sitosterol (%) 56.265.0
-5-avenasterol (%) ND6.9
-7-stigmasterol (%) 7.024.0
-7-avenasterol (%) 3.16.5
Others (%) ND5.3
ND nondetectable, defined as <0.05%.
3.1.3.3. Sterols Sterols are polycyclic alcohols derived from sterane. Sterols con-
stitute most of the unsaponifiable fraction of an oil. The sterol profile is character-
istic of each oil. The Codex Alimentarius (Codex-Stan 210-1999) indicates the total
sterol content (ppm) and the percentages of each sterol type in regular sunflower oil,
as shown in Table 6.
Among vegetable oils, regular sunflower oil is characterized by a medium sterol
content. According to the Codex Alimentarius (Codex-Stan 210-1999), the oils with
the highest sterol content are rapeseed oil (low erucic acid), with 480011,300-ppm
sterols; corn oil with 800022,100 ppm, and sesame oil with 450019,000 ppm.
In regular sunflower oil, the main component is b-sitosterol, followed by -7-
stigmasterol. The latter may be used as a tracer for detection of adulterations in sun-
flower oil, as most vegetable oils (except safflower oil) have fairly low amounts of
-7-stigmasterol (less than 7%).
3.1.3.4. Other Components of the Unsaponifiable Matter The unsaponifiable

matter in a crude regular sunflower oil is usually in the range of 0.51.5% (9,
17), or lower than 15 g/kg according to the Codex-Stan 210-1999. In addition to
sterols (around 2.44.6 g/kg) and tocopherols and tocotrienols (0.41.5 g/kg), there
are minor components of sunflower oil. Aliphatic compounds and terpenoids occur
naturally in oils. Of the terpenoid family, squalene is the most widely occurring
compound. The occurrence of squalene in regular sunflower oil is fairly low:
0.0080.019% (5) or 1520 mg/100 g (9). The aliphatic alcohol content is 100-
mg/100-g oil (9).
Carotenoids and chlorophylls are the major lipochromes of vegetable oils. Crude
regular sunflower oil is not particularly rich in carotenoids (as palm oil is) or in
chlorophylls (like rice bran, rapeseed, olive, and avocado oils). This gives crude
regular sunflower oil its light-amber color, turning to pale yellow upon the bleach-
ing operation.
3.2. Chemical Characteristics of Regular Sunflower Oil

The Codex Alimentarius (Codex-Stan 210-1999) indicates the characteristics of
crude regular sunflower oil: (1) saponification value 188194-mg KOH/g oil;
(2) iodine value (calculated from the fatty acid composition) 118141. However,
Merrien (9) reports an iodine value of 120134, and Bockisch (5) reports a value in
the range of 110143 (Wijs method).
3.3. Physical Characteristics of Regular Sunflower Oil

3.3.1. Refractive Index The refractive index is a characteristic property of fats
and oils and may be used as a fast measurement of the advance of a hydrogenation
operation. The Codex Alimentarius (Codex-Stan 210-1999) indicates a refractive
index (nD) of regular sunflower oil in the range of 1.4611.468 at 40 C; Merrien
(9) reports the range 1.4741.476 at 20 C.
3.3.2. Density Determinations of the content of tanks or flow rates are usually
based on methods of volumetric dosing. These methods are used to facilitate equip-
ment automation. However, mass determinations based on volume measurements
will depend on the nature and temperature of an oil.
The Codex Alimentarius (Codex-Stan 210-1999) indicates a relative density of
regular sunflower oil in the range of 0.9180.923 (20 C/water at 20 C). The values
suggested by the Codex do not differ appreciably from the expected values for most
vegetable oils.
Figure 8 shows the temperature dependence of the density (g/mL) of an Indian
edible sunflower oil [based on Subrahmanyam et al. (27)].
0.96
0.95
0.94
Density (g/cc)
0.93
0.92
0.91
0.9
0.89
0.88
30 10 10 30 50 70 90
Temperature (C)
Figure 8. Temperature dependence of the density (g/cc) of an Indian edible sunflower oil [based
on (27)].
672 SUNFLOWER OIL
60
50
Dynamic Viscosity (cP)
40
30
20
10
0
20 25 30 35 40 45 50 55 60
Temperature (C)
Figure 9. Temperature dependence of the dynamic viscosity (cP) of refined sunflower oil [based
on (28)].
3.3.3. Viscosity The viscosity of an oil is a fundamental parameter when pump-

ing is required. The viscosity of a vegetable oil will depend on the fatty acid com-
position. Oils with hydroxylated fatty acids (like castor oil and lesquerelle oil) have
a particularly high viscosity.
Figure 9 shows the temperature dependence of the dynamic viscosity (cP) of
refined sunflower oil. The curve corresponding to the crude oil is along the same
curve [based on Abramovic and Klofutar (28)].
3.3.4. Specific Heat and Combustion Heat The specific heat of sunflower oil
at constant pressure is 2.197 J/kg C (29). The energy content or combustion heat of
an oil is a major parameter when used as an energy source. The gross heat contents
of all vegetable oils are fairly close to each other. Ali and Hanna (30) report a gross
heat content of regular sunflower oil of 39,575 kJ/kg, and Bhattacharyya and Reddy
(31) a value of 39,486 kJ/kg.
3.3.5. Smoke Point, Flash Point, and Fire Point The smoke point, flash point,
and fire point of an oil are relevant parameters in deep-fat frying processes. The
fatty acid composition of the oil is not relevant (unless the oil has short-chain fatty
acids, as is the case of butter or coconut oil). The most important effect is generally
that of free fatty acids (FFA) in the oil. The following values have been reported for
fully refined sunflower oil (with 0.10% free fatty acids): smoke point 209 C;
flash point 316 C; fire point 341 C (5).
The flash point is also an important parameter when considering the possibility
of using an oil as an alternative diesel fuel in ignition engines. The flash points of
all vegetable oils are far above that of diesel fuel, reflecting the nonvolatile nature
of vegetable oils. Ali and Hanna (30) report a value of 274 C for the flash point of
regular sunflower oil.
3.3.6. Melting Characteristics
3.3.6.1. Cloud Point and Pour Point The cloud point is the temperature at which
solids first become visible when an oil is cooled. The pour point is the temperature
at which the amount of solids out of solution is sufficient to gel the liquid; thus, it is
the lowest temperature at which the oil is fluid.
The above parameters are relevant when pumping oils at low temperatures or for
their use as alternative diesel fuel in ignition engines. The cloud points and pour
points of the vegetable oils are higher than for diesel fuel.
Widely varying values have been reported in the literature for regular sunflower
oil. For example, Bockisch (5) reports a cloud point of 10 C (and a solidification
point in the range 16 to 18 C). Ali and Hanna (30) report a cloud point of 7.2 C
and a pour point of 15.0 C. Differences between reported cloud points are possi-
bly caused by a varying degree of winterization of the oils considered.
3.3.6.2. Thermal Behavior of Regular Sunflower Oil Crude regular sunflower oil
is a liquid at room temperature. The refined oil resists refrigerator temperatures
without the appearance of turbidity. These characteristics make it suitable as salad
oil. The thermal behavior of an oil may be determined within wide temperature
ranges through methods of nuclear magnetic resonance (NMR) or through differen-
tial scanning calorimetry (DSC). Both methods allow the evaluation of indices
related to the solid content as a function of temperature.
Power
Sunflower oil
Soybean oil
-60 -40 -20 0 20 40

Temperature (C)
Figure 10. Thermograms of a refined sunflower oil and of a refined soybean oil (32).
674 SUNFLOWER OIL
100
80
Sunflower oil
60
Soybean oil
SFI
40
20
0
-60 -40 -20 0 20 40
Temperature (C)
Figure 11. Solid fat content (SFI) of a refined regular sunflower oil and a refined soybean oil
(32).
The thermogram of an oil determined by DSC allows the study of thermal beha-
vior, and the evaluation of the solid percentage from the area of peaks. This infor-
mation is characteristic of the fatty material considered. The thermograms of a
refined regular sunflower oil and a refined soybean oil (32) are compared in Fig-
ure 10. It is clear that melting of sunflower oil is practically complete above 15 C.
Soybean oil, however, has a second, smaller peak between 5 C and 5 C, which
corresponds to a higher content of saturated fatty acids. Figure 11 shows the solid
content as a function of temperature for both oils, as determined by partial integra-
tion of peaks in the above thermograms (32).
4. SUNFLOWER SEED OF MODIFIED FATTY ACID COMPOSITION
Until two decades ago, the fatty acid composition of vegetable oils was closely
related with their origin. The fatty acid profile of sunflower oil was thus defined
within natural variation ranges. Current practices, however, are widely based on
the production of oilseed of modified fatty acid composition. Several methods
have been developed to this end.
The genetic diversity of wild sunflower allowed researchers to obtain a number
of varieties of defined characteristics. The North Central Regional Plant Introduc-
tion Station (Iowa), which gathers dozens of species, has distributed materials to
researchers and companies interested in either the study or use of these materials.
The sunflower germplasm collection of the U.S. Department of Agriculture
(USDA) is the largest and most complete collection worldwide, including wild
materials from 48 U.S. states, as well as samples from Canada and Mexico. The
SUNFLOWER SEED OF MODIFIED FATTY ACID COMPOSITION 675
program of the U.S. Department of Agriculture/Agricultural Research Service

(USDA/ARS) evaluates wild material according to oil content and composition.
Most breeding programs are aimed at the development of hybrids, although other
projects are for the improvement of open-pollinated varieties and synthetic culti-
vars. In view of the improvements in yield, disease resistance, uniformity, and
self-compatibility achieved in some modifications, open-pollinated varieties have
been replaced by hybrids. Varieties have been produced with increased oil content
of seeds and/or improvements in oil composition.
Inbreeding has been used since 1920 for the improvement of sunflower, the most
common method consisting in the self-pollination of those phenotypically desirable
plants within the existing cultivars. The progeny of the best plants are sown in the
following season, and the selection procedure is continued among the resulting pro-
geny. A variation of recurrent selection, the method of reserves, developed by Pus-
tovoit in the USSR, consists of the evaluation of progeny and subsequent cross-
pollination among superior progenies.
In order to obtain hybrid seeds, self-pollination and pollination by a sibling plant
must be avoided. It is sought to develop a female parental line accepting pollen
from other lines (cross-pollination), without reproduction of its own line. This
can be achieved directly through emasculation, i.e., elimination of the pollen-
bearing organ, usually at the expense of large labor requirements. Other methods
can be used to induce male sterility.
Modifications in the characteristics of sunflower are obtained through crosses to
recombine genes from two sexually compatible parents. Both for self pollination
and for controlled crosses, it is important that heads be isolated before flowering
to avoid natural cross-pollination.
Emasculation of the female parent is used for the production of artificial hybrids.
Acid-induced male sterile plants pollinated without emasculation are also used.
These hybrids were produced by natural crossing in seed production fields (breed-
ing nursery), with the two parents planted in alternating groups of rows. The first
hybrid cultivars were introduced in Canada for commercial production in 1946 (4).
Cytoplasmic male sterility of sunflower was discovered in 1958, whereby one
factor in the cell cytoplasm leads to the male-sterility in all plants of the second
generation. However, as a result of the pollination of fertile ordinary lines, plants
will still bear some fertile progeny. Patrice Leclercq, in France, in 1969, reported on
cytoplasmic male sterility obtained in the progeny of a cross of Helianthus petio-
laris Nutt and Helianthus annuus. In 1970, fertility restorer lines were found. Two
years later, hybrid seeds produced by this system were made available to farmers. In
1976, 80% of the U.S. sunflower harvest had been produced on hybrid seed.
The production of single crosses or three-way hybrids using cytoplasmic male
sterility and nuclear fertility restorer system is widely used. Genic male sterility
was used for the production of hybrid seeds in the early 1970s in France and
Romania. The first hybrids produced in this way were introduced in the United
States in 1972 (4).
Various genetically different types of cultivar are commercially available. In
open-pollination varieties, each individual is genetically different from another
676 SUNFLOWER OIL
among the population. This results in a high level of adaptation to different envir-
onment conditions on the one hand, but it may also cause handling difficulties
because of the lack of phenotypical uniformity. Open-pollination varieties may
be recovered yearly by the producer, the regular purchase of new seed being neces-
sary only for maintenance of purity and type.
Commercial hybrid cultivars include single and three-way hybrids. The former
type is produced by crossing two lines, resulting in genetically identical individuals;
the latter are developed by crossing a single hybrid with a homozygote line. Seed
must be procured yearly by the producer in both cases.
The production of high-quality hybrid seeds does not only depend on the use of
parental lines of superior class, but also on the degree of isolation of the cultivation
field from other sunflower plantations, including wild varieties. The isolation con-
ditions cannot be accurately established, in view of the role of insects as pollination
agent, and the long viability period of pollen; yet, recommendations have been
made.
Genetic variation occurring naturally within crop species is scarce. Additional
genetic modification strategies are required to generate variants of a certain fatty
acid profile. Such variants may be included in crossing programs for the develop-
ment of cutivars or strains of interest. Modifications in the fatty acid profile of an oil
may be achieved through techniques of mutagenesis, Stable genetic mutations may
be induced by use of chemical mutagens, such as ethyl methyl sulfonate. This pro-
cedure was used in the USSR to develop sunflower seeds of improved oleic acid
content.
4.1. Modification of the Saturated Fatty Acid Content

The consumption of fat of high saturated fatty acid content has been associated with
increased risk of coronary heart disease. Traditional sunflower oil contains around
1112% saturated fatty acids, a considerably low value among vegetable oils.
Canola oil has 7% and safflower oil less than 10% of saturated fatty acids, both
being strong competitors of the edible oil market.
Since 1992, the National Sunflower Association (NSA) has supported a cultiva-
tion program developed by researchers of the USDA/ARS for a reduction of the
saturated content of sunflower oil. New germplasm stocks with reduced content
of palmitic and stearic acids were made available. They were developed through
continuous selection starting from a sunflower accession collected in Egypt around
1950. Several private companies have also carried out investigations aiming at
obtaining hybrids of low saturated fatty acid content (6% or lower). Cultivators
of Pioneer Hi-Bred International Incorporation managed to reduce the stearic
acid content to 1.5% for commercial products. SVO Enterprises and Triumph
Seed Company developed research lines aiming at reducing the saturated quantity
in high-oleic sunflower oil (3334).
With a view to finding new industrial uses of vegetable oils, the saturated fatty
acid content may also be increased. Three high stearic acid sunflower mutants, hav-
ing as much as 28%, 15%, and 14% of stearic acid in the seed lipids have been
biochemically characterized (35). An increased solid content of the oil could be

obtained in the oil without requiring hydrogenation, although such an increase
may not meet nutritional standpoints.
4.2. High-Oleic Sunflower

The typical sunflower oil composition is 6672% linoleic acid, 12% saturated acids
(palmitic and stearic), 1620% oleic acid, and less than 1% a-linolenic acid. An
increase in low-density lipoprotein cholesterol (LDL-C) and a decrease of high-
density lipoprotein cholesterol (HDL-C) are believed risk factors of coronary heart
disease (CHD). Diets rich in saturated fat increase plasma total and LDL-C. Tradi-
tional high-linoleic sunflower oil has always been regarded as healthy because of its
high content of polyunsaturated fatty acids (PUFA) and relatively low content in
saturated fatty acids.
The substitution of saturated fatty acids by PUFA in a diet leads to a reduction of
total cholesterol and LDL-C. It has been suggested that a high intake of polyunsa-
turated fatty acids leads to a decrease of HDL-C. Some authors have proved this
theory based on rather unrealistic high PUFA in the diet (PUFA/MUFA > 3). Others
have found statistically nonsignificant decrease values for HDL-C in more realistic
high PUFA diets. An increased intake of monounsaturated fatty acids (MUFA) also
leads to a decrease in total cholesterol and of LDL-C levels without reducing HDL-
C even with fairly high MUFA intake values. As oleic acid is more stable against
oxidation than linoleic acid, consumption of MUFA has further advantages over
PUFA. It is recommended to avoid foods containing peroxidized lipids, as these
might be initiators of pathologic processes. In view of the considerations above,
new genetic strategies were started toward a high-oleic sunflower oil (HOSO).
There is also controversy over the importance of MUFA over PUFA from the
metabolic viewpoint. Great emphasis is placed on the distinction between the n-3
PUFA and those of the n-6 family. An increased intake of n-3 and a reduced intake
of n-6 are recommended in light of the competitive metabolism of both families
of fatty acids. As linoleic acid is a n-6 parent, a reduction of its intake favors the
n-3/n-6 ratio; on the other hand, it is also an essential fatty acid.
High-oleic sunflower oil, with very low PUFA levels, may well suit the require-
ments of processors, but it does not support the work of nutritionists who recom-
mend n-6/n-3 ratios within the range 5 to 10. In addition, HOSO does not represent
an increased intake of family n-3 fatty acids as recommended by nutritionists, the
linolenic acid content being very low for all types of sunflower oil.
K. I. Soldatov, in Russia, developed high-oleic sunflower seeds through the treat-
ment of normal seed with a chemical mutagen (dimethyl sulfate). Through pro-
grams of selected breeding, a number of plants containing seed with as much as
8090% oleic acid were obtained. L. N. Kharachenko, also in Russia, studied the
standard Peredovik progeny and the Pervenets progenyobtained from treatment
of seeds of the former variety with a chemical mutagen. It seems that modifications
in the seed genotype of high-oleic Pervenets are responsible for an irreversible
blockage of the desaturating enzyme system. G. N. Fick developed progenies of
678 SUNFLOWER OIL
cultivar Pervenets in plantations in the United States, Argentina, and Chile and
incorporated the dominant genes into hybrids that were suitable for commercial
production. High-oleic seeds were first grown commercially in the United States
in 1984 (3637).
The Lubrizol Corporation obtained U.S. patents (granted to Sigco and inventor
Gerhardt N. Fick) for sunflower seeds and oils of oleic acid content of 80% or
higher and linoleic/oleic ratios lower than 0.09Patent 4,627,192 for seed granted
on December 9, 1986, and Patent 4,743,402 for oils granted on May 10, 1988. SVO
Enterprises, a division of Agrigenetics Company, which is a part of The Lubrizol
Corporation, has produced high-oleic sunflower oil trademarked under the name
Trisun in the United States (3334, 38).
High-oleic sunflower oil is sold in Australia under the name Sunolaa regis-
tered trademark of Meadow Lea Foods. The seed variety was bred by Australian
farmers through traditional selective breeding techniques. The first Sunola crop
was developed in Queensland. The oils fatty acid composition is 85% monounsa-
turated, 8% polyunsaturated, and 7% saturated. The composition of oil extracted
from Sunola seed in the first stages of ripening resembles that of regular sunflower
varieties. Only when the synthesis of oil has actually started (some three weeks
after flowering) does the oleic acid content start to increase considerably and the
linoleic content start to decrease rapidly (39).
A further approach to modified sunflower oils was made by the Agriculture
Canada Research Station in Saskatoon (Saskatchewan) from two different types
of dwarf early-ripening sunflower trademarked under the name Sunola (Western
Grower Seed Corp. was created for commercialization and further improvement
of Sunola). One of these types was regular high-linoleic Sunola, which was first
produced commercially in 1993. The hybrid was specially developed for farmers
in regions of the west of Canada where cultivation of sunflower was nonexistent
because of a short growing season. The fatty acid composition of this oil is 72
74% linoleic acid (owing to the colder growth conditions), 14% oleic acid, and
12% saturated acids. The other hybrid is high-oleic Sunola sunflower, whose pro-
duction started in 1995. The ripening time of high-oleic Sunola is about 100 days
(three weeks shorter than for most sunflower crops). The fatty acid composition is
87% oleic acid, 5% linoleic acid, and 78% total saturated acids. Seeds of this type
are smaller than regular sunflower seeds, the hull being slightly lighter and bearing
a narrow stripe (3334, 40).
As both Sunola crops are special, care must be taken against contamination with
traditional sunflower or canola. However, this is rarely the case, as Sunola is grown
in northern areas of the United States, where regular sunflower is not grown and in
areas of southern Canada that are too hot and dry for the development of canola
crops.
It is worth noting that the name Sunola for modified oils is used in Australia for
high-oleic sunflower oil, whereas, in Canada, it is a registered trademark of two oils
of different composition: one of higher linoleic acid content than traditional sun-
flower and another of high-oleic type. Care must be taken that this should not
lead to confusion. Canadian Western Grower Seed Corporation has also developed
TABLE 7. Fatty Acid Composition (%) of High-Oleic Sunflower Oil (37, 41, 43).
Fatty Acid Vermeersch (41) Krawczyk (43) Purdy (37)
16:0 3 6.7 2.74.2

18:0 5 3.45.0
18:1 83 80.0 80.586.7
18:2 9 12.0 4.08.5
lines of Sunola with higher oleic acid content to be introduced in the cosmetics mar-
ket. One of these has as much as 88% oleic acid.
Purdy (37) reported on the fatty acid composition and other analytical character-
istics of high-oleic sunflower oil of the Pervenets variety cultivated in three regions
of the United States. The oil content of seeds ranged between 43.3% and 47.9%
(dry basis), the hull accounting for 2433% of total seed weight. Table 7 shows
the fatty acid composition of high-oleic sunflower oil (37, 41 43).
Table 8 shows the chemical and physical characteristics of crude high-oleic sun-
flower oil according to the Proposed Draft Amendment to the Codex Standard for
Named Vegetable Oils (Alinorm 01/17). The Active Oxygen Method (AOM) value
for refined oil extracted from seed of high-oleic Pervenets variety is 5156 hours,
compared with 13 hours for regular oils (37). In another study, Purdy (36) extracted
oil from high-oleic Pervenets seed and from high-linoleic seed. The saturated fatty
acid content of these oils varies only slightly (811%); the main variation occurs in
the oleic/linoleic ratio. Table 9 shows AOM values for refined sunflower oils as a
TABLE 8. Chemical and Physical Characteristics

of Crude High-Oleic Sunflower Oil (Codex Alimen-
tarius, Alinorm 01/17).
Relative density (25 C/water at 20 C) 0.9090.915

Refractive index (ND 25 C) 1.4671.471
Saponification value (mg KOH/g oil) 182194
Iodine value 7890
Unsaponifiable matter (g/kg) <15
TABLE 9. AOM Time (hr) for Refined

Sunflower Oils of Different Oleic Acid
Content [based on (36)].
Oleic Acid AOM (hr)
26 % (regular) 11
51 % (regular) 18
79 % (high-oleic) 38
680 SUNFLOWER OIL
TABLE 10. Fatty Acid and Triacylglycerol Composition

(%) of Regular Sunflower Oil and of High-Oleic Sunflower
Oil [based on (25)].
Regular Oil High-Oleic Oil
Fatty Acid: % %
16:0 6.8 5.3
18:0 5.0 3.8
18:1 31.4 88.3
18:2 55.4 1.4
Triacylglycerol: % %
sn-1 sn-3
16:0 9.2 5.1
18:0 6.1 5.8
18:1 34.0 87.4
18:2 50.7 1.6
sn-2
16:0 0.5 0.3
18:0 0.4
18:1 34.7 98.6
18:2 64.2 1.1
function of the oleic acid content. Data presented show that the oxidative stability
of this oil is enhanced considerably with an increase in the oleic acid content. Table 10
shows the triacylglycerol composition of high-oleic oil (derived from Pervenets)
and of regular sunflower oil, with practically no occurrence of saturated fatty acids
in the sn-2 position for either oil (25).
Although the content of oleic acid is high in both high-oleic sunflower oil and
olive oil, there is a higher content of saturated acids in olive oil. Their MUFA con-
tents are similar, but the composition of triacylglycerols differs widely (Table 11).
Whereas the triacyglycerol OOO (O oleic) is the main species in both oils, HOSO
has a higher content, olive oil, in contrast, having a higher proportion of POO
(P palmitic). Further differences are in the fatty acids occupying those positions
other than sn-2, which is occupied by oleic acid in both oils. In addition, HOSO has
a higher proportion of linoleic acid in position sn-2, whereas olive oil has more a-
linolenic acid (44)
Another major difference between olive oil and HOSO is a most distinct flavor
of olive oil that characterizes it from HOSO and other oils. Extra virgin olive oil,
TABLE 11. Composition in Major Triacylglycerols (%) of Olive

Oil and High-Oleic Sunflower Oil (HOSO) [Based on (44)].
Major Triacylglycerols (%) Olive Oil HOSO
POO 30.5 12.1

OOO 49.9 65.1
OLL 0.3 3.1
O oleic, P palmitic, L linoleic.

TABLE 12. Sterol Composition (as Percentage of Total Sterols)

of High-oleic Sunflower Oil (According to Codex Alimentarius,
Alinorm 01/17) and Olive Oil (20).
Sterol Composition High-Oleic Sunflower Oil (Codex) Olive Oil (20)
Campesterol (%) 5.013.0 4.0

Stigmasterol (%) 4.513.0 <4.0
b-Sitosterol (%) 42.070 75.0
5-Avenasterol (%) 1.56.9 414
7- Stigmasterol (%) 6.524.0 0.5
7-Avenasterol (%) ND9.0
Others (%) 3.59.5
ND nondetectable, defined as <0.05%.
being the preference of so many gourmets worldwide, is considered the finest

choice oil.
The similarity in the fatty acid composition of HOSO and olive oil may lead
to cases of adulteration or fraud, in view of the price difference between the
two oils. These adulterations may be difficult to detect through conventional analy-
tical methods. The nature of an oil can be traced through a study of its sterol
composition.
The sterol composition of both oils is compared in Table 12. Clearly, for HOSO,
b-sitosterol is the sterol with the highest occurrence (4270% of total sterols), fol-
lowed by 7-stigmasterol (6.524%), campesterol (513%), and stigmasterol (4.5
13%). Although there are differences in the sterol composition of both oils, they are
not large enough to enable easy analysis.
Several studies have been aimed at the detection of a fraudulent addition of
vegetable oils to olive oil. In particular, different analytical methods can be applied
to determine blends of regular and high-oleic sunflower oil with olive oil. The mini-
mum sunflower oil detection level depends on the analytical method used. For
example, a minimum detectable level of 0.7% of regular or high-oleic sunflower
oil may be achieved through methods of sterol analysis, and analysis of the fatty
acids will not enable detection of additions below 20% (45).
High-oleic sunflower oil is widely used as salad oil and cooking oil, because of
its composition, light flavor. A high content of oleic acid provides enhanced oxida-
tive stability in frying processes. In addition, it does not require partial hydrogena-
tion for an increase in product shelf-life, with the additional nutritional advantages.
The effect of trans-fatty acids generated as byproducts of hydrogenation processes
on the plasma lipoprotein profile is as adverse as that of saturated fatty acids, both
increasing the concentration of LDL-C and reducing that of HDL-C (44).
High-oleic sunflower oil is sprayed on cereals, crackers, and cookies to retain
freshness and crispness. It is also used in the manufacture of non-dairy creamers,
snack foods, and frozen desserts. Special properties of oleic acid make high-oleic
sunflower oil a choice ingredient for cosmetic formulations. The AOM value of
682 SUNFLOWER OIL
Florasun-90 (of International Flora Technologies Ltd.) is higher than 90 hoursa

high value compared with less than 40 hours for high-oleic rapeseed oil and about
20 for sesame oil. Research has indicated that the oil is not skin-irritating or sensi-
tizing. It may be used in suntanning products and cosmetics with a high content of
natural lipids, such as bath oils, massaging oils, skin-care products, lipstick, and
cosmetic cream bases (33, 34).
High-oleic sunflower oil is currently used in the manufacture of a lubricant for
diesel and gas motors. The product, denominated Helianthe, is commercialized by
the Tecnol Society (France). It is composed of 7080% high-oleic sunflower oil and
2030% additives. Helianthe is a formulation type 5W40, with a high viscosity
index and high fluidity at ignition (41).
4.3. Mid-Oleic Sunflower

The production of high-oleic sunflower oil with 80% or higher oleic acid content
was protected under patents. However, the patent holder agreed to license breeding
material for the development of mid-oleic sunflower seed. In July 1995, the NSA
decided to redirect efforts toward an increase in oleic acid. It was established that
mid-oleic sunflower should contain 65% oleic acid, no higher than 10% saturated
acids, and the rest being linoleic acid, a balance that, according to research, pro-
vides in-process functionality in frying.
Breeding a mid-oleic sunflower requires at least one oleic parent. The USDA/
ARS Northern Crop Science Laboratory in Fargo, North Dakota, provided private
companies with crossing lines of mid-oleic sunflower. Hybrid seeds were developed
by traditional crossing methods; no hybrids of transgenic sunflower were used. The
mid-oleic concentration appears to be controlled by a partially dominant major
gene and one or more dominant minor modifier genes (46, 47).
In a market accustomed to HOSO and traditional high-linoleic sunflower, the
name NuSun seemed suitable and was trademarked by the NSA. Seed and other
companies using the name NuSun in their commercial products should have author-
ization of the NSA. NuSun contains less than 10% saturated, 5075% monounsa-
turated, and 3032% polyunsaturated fatty acids, with less than 1% linolenic acid (46).
Harvests of NuSun were first commercialized in 1999. In 2000, Procter & Gam-
ble chose NuSun for the manufacture of Pringles chips in North America, part of
Europe, and Asia, finding a low rate of formation of polar compounds as compared
with other oils, an important factor for extending product shelf-life (46, 48).
Figure 12 shows the fatty acid composition (%) of regular, mid-oleic, and high-
oleic sunflower oils according to the Proposed Draft Amendments to the Standard
for Named Vegetable Oils (Report of the Eighteenth Session of the Codex Commit-
tee on Fats and Oils, London, 2003). Table 13 shows the composition in major tria-
cylglycerols of mid-oleic sunflower oil, compared with the composition of regular
sunflower oil (49). Clearly, there is a difference in the unsaturated triacylglycerol
composition of both oils: mid-oleic sunflower oil has a higher content of OOO, and
regular sunflower oil is richer in LLL and LLO (the addition of both contents
amounting to 60.3%).
100
90
16:0 18:0
80
70 18:1 18:2
60
Percentage
50
40
30
20
10
0
minimum maximum minimum maximum minimum maximum
regular mid-oleic high-oleic

Figure 12. Fatty acid composition (%) for regular, mid-oleic, and high-oleic sunflower oil (based
on the Proposed Draft Amendments to the Standard for Named Vegetable Oils Committee on
Fats and Oils, 2003).
Figure 13 shows the composition in major triacylglycerols of mid-oleic sun-

flower oil (49), as compared with the calculated composition from a random distri-
bution. The triacylglycerol distribution does not fit the random model, the main
differences being in the levels of OOL and OOO, the main triacylglycerols. In con-
trast, as shown in Figure 6, the fatty acid distribution in regular sunflower oil TAG
differs only slightly from the random distribution. Table 14 shows the chemical and
physical characteristics of crude mid-oleic sunflower oil according to the Proposed
TABLE 13. Composition in Major Triacylglycerides (%) of

Mid-Oleic Sunflower Oil, as Compared with the Composition
of Regular Sunflower Oil [Based on (49)].
Triacylglycerols Mid-Oleic Regular
LLL 11.5 32.4

LLO 12.1 27.9
LLP 4.1 10.7
LOO 8.3 6.7
LLS 2.7 7.4
LOP 2.6 4.8
OOO 40.2 1.7
LOS 1.6 2.2
POO 5.7 0.6
SOO 5.4 0.4
L linoleic, O oleic, P palmitic, S stearic.

684 SUNFLOWER OIL
45
40
35 experimental
30 random
Percentage
25
20
15
10
0
SatOL SatLL Sat OO OOL LLO LLL OOO
Figure 13. Triacylglycerol composition of mid-oleic sunflower oil as calculated from random
distribution and experimentally determined (49). (Key: Sat saturated acid, O oleic acid,
L linoleic acid.)
Draft Amendments to the Standard for Named Vegetable Oils (Report of the Eight-
eenth Session of the Codex Committee on Fats and Oils, London, 2003).
Table 15 shows the sterol composition of mid-oleic sunflower oil according to
the Proposed Draft Amendments to the Standard for Named Vegetable Oils (Report
of the Eighteenth Session of the Codex Committee on Fats and Oils, London,
2003). Clearly, b-sitosterol is the sterol with the highest occurrence (5658% of
total sterols), followed by campesterol (9.19.6%) and stigmasterol (9.09.3%).
b-sitosterol is the main sterol in all three types of sunflower oil (regular, mid-oleic,
and high-oleic).
4.4. Semi-Dwarf and Dwarf Sunflower

The search for sunflower varieties of defined characteristics was also aimed at a
reduction of plant size and adaptation to other climates. Short-stature cultivars in
TABLE 14. Chemical and Physical Characteristics

of Crude Mid-oleic Sunflower Oil (According to the
Proposed Draft Amendments to the Standard for
Named Vegetable Oils, 2003).
Relative density (25 C/water at 20 C) 0.9140.916

Refractive index (ND 25 C) 1.4611.471
Saponification value (mg KOH/ g oil) 190191
Iodine value 94122
Unsaponifiable matter (g/kg) <15
EXTRACTION AND PROCESSING OF SUNFLOWER OIL 685
TABLE 15. Sterol Composition (as Percentage

of Total Sterols) of Mid-Oleic Sunflower Oil
(According to the Proposed Draft Amendments
to the Standard for Named Vegetable Oils, 2003).
Sterol Composition Mid-Oleic Sunflower Oil (%)
Campesterol 9.19.6
Stigmasterol 9.09.3
b-Sitosterol 5658
5-Avenasterol 4.85.3
7- Stigmasterol 7.77.9
7-Avenasterol 4.34.4
Others 5.45.8
sunflower are classified as semi-dwarf (typical 1.20 m to 1.50 m high) and dwarf
(typical 0.80 m to 1.20 m high) types. Dwarf cultivars were developed more
recently and include dwarf hybrids and dwarf open-pollinated cultivars.
Two types of hybrid denominated Sunola and Sunwheat were developed in
Canada to address the handling problems caused by traditional hybrids, requiring
special machinery adapted only to the long growing season areas in southeastern
Saskatchewan. These new hybrids are 2535% shorter than regular sunflower
(hence the denomination miniature or dwarf), allowing use of the same machinery
as is used for cereal or canola production. Both early maturing types offer producers
in short-growing-season areas the opportunity to diversify rotations.
Sunola is a miniature type of sunflower developed by the Agriculture Canada
Research Station in Saskatoon as a sowing alternative for areas where growth of
traditional sunflower is not viable. It is the result of persistent selection of open-
pollinated varieties. Plant height is small (6090 cm), and heads are 813 cm in
diameter. Ripening time is 99103 daysthree weeks shorter than for most sun-
flower varieties. Sunola has a high oil content (similar to that of the best hybrids)
and a higher content of linoleic acid (7274%) than any other commercial
sunflower.
Sunwheat is a dwarf hybrid of sunflower, having leaves and heads of similar size
to other hybrids, but short (96120 cm). Ripening time is 100110 days, and the oil
content is slightly lower than that of Sunola. It is appropriate for cultivation in bar-
ren areas and has a higher resistance to extreme-heat summer periods.
5. EXTRACTION AND PROCESSING OF SUNFLOWER OIL
The procedures used for the extraction and processing of sunflower oil are broadly
the same as for other seed oils. Focus will be made on those operations or details
specific of the production of sunflower oil. Sunflower oil is usually extracted
through pressing of seed and later extraction by solvent. The crude oil is usually
subjected to traditional refining stages. Otherwise, cold-pressed sunflower oil is cur-
rently valued as a new extra virgin oil.
686 SUNFLOWER OIL
Sunflower seed
Cleaning
Drying
Storage
Cracking
Hull Dehulling
Cracking
Conditioning
Pre-treated seed to oil

extraction
Figure 14. Preparatory treatment of sunflower seed for extraction.
5.1. Preparation of Sunflower Seeds for Extraction

Figure 14 shows normal stages in the preparation of sunflower seed. Once har-
vested, sunflower seeds are cleaned, dried, and stored. Seeds must be dehulled prior
to pressing and oil extraction stages. Depending on processing plant, seeds may be
cracked before the dehulling stage to reduce seed size and help remove the hull.
Kernels may be further broken and subjected to two conditioning stages: cooking
and flaking. Thermal conditioning or cooking is directed to an adjustment of the
moisture content (generally to 34.5%) and the temperature (generally 100 C) of
meats. The last stage in the preparation of seed is the conversion of the cracked,
dehulled, and conditioned meat into a flake. As these stages are common to a num-
ber of oilseeds, only those aspects specific of sunflower seeds are considered,
namely, in the drying and dehulling stages.
5.1.1. Drying of Sunflower Seeds The moisture content of sunflower seed

must be reduced to around 89% prior to storage. Different authors indicate slightly
varying levels: 15% (50), 910% (51), 8.5% (52), and 10.5% (5). In order to prevent
losses caused by unfavorable climatic conditions, seeds are usually harvested with
moisture contents above the recommended levels for storage.
When seed moisture is higher than the recommended value, enormous spoilage
of the seeds by microbiological attack is possible. Fungi may also grow explosively
over the surface of seeds, with the consequent increase in temperature caused by
biological activity. Such temperature increase leads to ideal life conditions for ther-
mophilic bacteria; their metabolism contributes further to a temperature increase.
Enzyme and mold activity reduces the quality and the yield of the extracted oil.
Seed moisture is normally expressed as weight percentage of the whole seed. As
water is insoluble in seed lipids, the moisture content is concentrated in the nonfatty
parts of the seed. The water content calculated on a nonfatty basis is defined as cri-
tical moisture. The critical moisture of sunflower seeds is 16%, although a max-
imum 15% is recommended for storage.
As the content of nonfatty materials in sunflower seeds decreases as the oil con-
tent increases, the moisture content corresponding to one critical moisture value is
inversely related to the oil content, as shown in Table 16 [based on Muller (50)]. For
critical moisture levels above 15%, the rate of respiration of seeds increases.
Respiration is accompanied by an exothermic transformation of organic substances
of seeds, creating conditions that may lead to spontaneous combustion (50).
Drying of seed may be performed at room temperature with no additional equip-
ment, or with hot-air dryer. The first stage of drying consists of the removal of
external moisture from the fresh seeds. Internal water diffuses outward, evaporating
in the external part of seeds. After a certain time, seeds reach a hygroscopic equili-
brium state at which the moisture content remains constant. The equilibrium
depends on ambient temperature and relative humidity of the surrounding air.
Figure 15 is a representation of these values for different temperatures: 10 C,
25 C, and 40 C [based on Mazza and Jayas (51)].
The equilibrium moisture of seeds is modified upon dehulling. Equilibrium
moisture values for undehulled sunflower seeds, hulls, and kernels are compared
as a function of the relative humidity of the surrounding air at 25 C (Figure 16).
The initial moisture content of all seeds was 5% (dry basis). Those samples stored
at a relative humidity below 33% reached the equilibrium by desorption, and those
at a relative humidity above 33% reached the equilibrium by adsorption (51).
TABLE 16. Critical Moisture of Sunflower Seeds with Different Oil and Moisture Content
[Based on (50)].
Oil Content (%) Nonfat Content (%) Moisture (%) Critical Moisture (%)
35 65 9.75 15
40 60 9.00 15
45 55 8.25 15
48 52 7.80 15
688 SUNFLOWER OIL
25
equilibrium moisture content (%)
20 10C
25C
15
40C
10
0
0 20 40 60 80 100
relative humidity (%)
Figure 15. Equilibrium moisture (%) of sunflower seed with hull as a function of air relative
humidity (%) for three temperatures [based on (51)].
5.1.2. Dehulling of Sunflower Seeds With approximately 30% of hull, sun-

flower seeds must be dehulled prior to processing. The high wax content of hulls,
which would otherwise be transferred to the oil during extraction, is one major
reason for dehulling. The wax content of an oil extracted from undehulled seed
is approximately five times higher than for oils extracted from dehulled seed.
However, a small fraction of hull (less than 15%) is left with the seed for easy
percolation during the process of extraction by solvent. Seed moisture is usually
reduced to values below 8% for hulls to turn more brittle and be easily removed.
30
seeds
equilibrium moisture content (%)
25
kernels
20
hulls
15
10
0
0 20 40 60 80 100
relative humidity (%)
Figure 16. Equilibrium moisture (%) of sunflower seeds with hull, hulls, and kernels as a function
of air relative humidity (%) at 25 C [based on (51)].
Genetic improvements of sunflower seeds have been aimed at increasing oil con-
tent, but also leading to a decrease in the amount of hull. The seed pericarp is thin-
ner and more firmly attached to the kernel in improved varieties. The hullability
of improved seeds, i.e., the ease with which hulls can be cracked and removed from
the seed, is lower than that of seeds with hull of higher thickness.
Dehulling consists of mechanical removal of the pericarp (hull) of seeds. The
most widely used method consists of colliding of seeds at high speed against a
hard surface by centrifugal effect, leading to the cracking of seeds. Loose hull
bits are separated from partially dehulled seed. In addition to the size and shape
of seeds, the moisture content is a most relevant parameter in the dehulling process.
A decrease in moisture content facilitates hull removal, the effect being greater for
hybrids of higher oil content. However, a decrease in moisture also leads to an
increase in the amount and composition of fines. Therefore, it is necessary to deter-
mine the optimum value of seed moisture for maximum hullability and a reduction
in the amount of fines.
5.2. Sunflower Oil Extraction

Partially dehulled sunflower seed is generally used for oil extraction, with 8%
moisture and 10% residual hull content, approximately. The process employs
mechanical pressing followed by hexane extraction. Figure 17 represents a diagram
of the unit operations involved.
Pre-treated seed
Hot pressing
Pressed oil Expander
Solvent extraction
Extracted oil Desolventization

Pelleting
SUNFLOWER
CRUDE SUNFLOWER OIL MEAL
Figure 17. Production of crude sunflower oil.
690 SUNFLOWER OIL
Extraction of sunflower oil is generally carried out in two stages. The first stage
consists of mechanical extraction using screw-presses (expellers). The meal
obtained in the pressing stage, containing 1520% of oil, is subjected to extraction
by solvent (normally hexane). The solvent must then be eliminated from both meal
and oil. Oils obtained through pressing are of better quality than those obtained by
solvent extraction. However, both are blended before storage. Pressed oils are
sometimes commercialized separately from solvent extracted oils. The solvent-
extracted meal is obtained as a byproduct of this stage.
5.3. Treatment of Crude Oil

Figure 18 shows a diagram of alkali refining and physical (steam) refining of sun-
flower oil. The traditional method, alkali refining, involves degumming, neutraliza-
tion with alkali, bleaching, dewaxing, and deodorization. A pre-dewaxing stage
may be performed after neutralization to reduce the wax content to 100150
ppm, in addition to a stage of winterization after bleaching for removal of the
remaining waxes. Physical refining includes the following stages: degumming,
bleaching, dewaxing, and deodorization. Also for physical refining, a combined
stage of predewaxing and degumming makes post-dewaxing easier and less costly.
Crude sunflower oil
Degumming
Predewaxing
Alkali refining or Bleaching

neutralization
Predewaxing
Dewaxing
Bleaching
Dewaxing Deodorization/
Deacidification
Deodorization
FULLY REFINED SUNFLOWER OIL
Figure 18. Refining of crude sunflower oil.

Sunflower oil contains moderate quantities of carothenoids and xanthophylls, but

it does not contain chlorophylls. It may be easily bleached with less than 1%
bleaching earth. In physical refining, bleaching is carried out for the removal of
phosphatides and metals in addition to colored materials. The refining of crude sun-
flower oil is performed along the same stage sequence as for other oilseeds. Inclu-
sions of a degumming stage and a dewaxing stage are both worth considering
separately in detail. Details of physical refining of sunflower oil are also given
below.
5.3.1. Degumming Major phosphoacylglycerols of sunflower oil are phosphati-

dylcholine (PC), phosphatidylethanolamine (PE), phosphatidylinositol (PI), and
phosphatidic acid (PA). The phospholipid content of solvent-extracted oils is higher
than that of hot-pressed oils. Cold-pressed oils contain hardly any phospholipids.
Most of the phospholipids in sunflower oil are hydratable and may be removed
by water degumming. Figure 19 shows a comparison of the variation in the phos-
pholipid content of two crude sunflower oils, obtained by pressing and by solvent
extraction, respectively, and a degummed sunflower oil (53).
Nonhydratable phospholipids (mainly Ca and Mg salts of PA and lysoPA, gly-
cerophosphates, and inorganic phosphates) remain in the oil after water degum-
ming. As a result, a degummed oil may have a significantly higher percentage of
PA and a lower percentage of other phospholipids (particularly PC) than the
original crude oilPC being almost fully hydratable and PA is nonhydratable
when complexed with Ca or Mg, PC and PI hydrating considerably faster than
1.4
1.2 minimum
Phospholipid content (%)
1 maximum
0.8
0.6
0.4
0.2
0
pressed extracted degummed
Figure 19. Minimum and maximum values of the total phospholipid content in two crude
sunflower oils, obtained by pressing and by solvent extraction respectively, and in a degummed
oil [based on (53)].
692 SUNFLOWER OIL
100
90
pressed
80
70 extracted
60 degummed
Percentage
50
40
30
20
10
0
PE (min) PE (max) PA (min) PA (max) PI (min) PI (max) PC (min) PC (max)
Figure 20. Minimum and maximum values of the content of each phospholipid in two crude
sunflower oils, obtained by pressing and by solvent extraction respectively, and in a degummed
oil [based on (53)].
PE and PA (54). Figure 20 shows a comparison of the variation range of each phos-
pholipid in two crude sunflower oils, obtained by pressing and by solvent extrac-
tion, respectively, and a degummed sunflower oil (53). The hydratability of these
compounds may be increased with the addition of either phosphoric or citric acid,
leading to a more efficient degumming process. Nonhydratable phospholipids may
also be removed by enzymatic treatment through special biochemical reactions,
such as enzyme-catalyzed hydrolytic cleavage of the phospholipid molecules (53).
Smiles et al. (55) studied the effectiveness of six different degumming reagents
for crude sunflower oil: water, citric acid, phosphoric acid, oxalic acid, acetic anhy-
dride, and maleic anhydride. All reagents were significantly more effective than
water in removing phospholipids, with maleic anhydride and oxalic acid removing
95% and 90% of total phosphorous, respectively. No significant changes were found
in the composition of the phospholipids remaining in the oil.
Pan et al. (56) studied the influence of different operation variables (temperature,
contact time, type and concentration of degumming reagent, pH, calcium and
magnesium content) on both water and acid degumming efficiency for sunflower
oil. All degumming solutions (phosphoric acid, citric acid, and a 50:50 blend of
both solutions) had a 2.5% concentration. Optimum degumming conditions with
phosphoric acid were 6070 C and addition of 10% of its solution. Optimum
degumming conditions with citric acid were 70 C and addition of 10% of its solu-
tion, whereas for the blend of both reagents, treatment at 60 C and 8% of the blend
was optimum.
5.3.2. Dewaxing Winterization is achieved by cooling an oil with the consequent

crystallization of high melting point fractions (waxes and/or triacylglycerols).
These are responsible for the appearance of turbidity in some edible oils during sto-
rage at low temperature or even at room temperature.
Winterization of sunflower oil is usually referred to as dewaxing. Improved
techniques have emerged in recent years with the appearance of seed varieties of
high oil and additional wax contents. As the improvements in oil yield (higher
than 40% for these seeds) have been obtained at the expense of a reduction in
hull thickness, the concentration of seed protection substances (waxes) in the
hull has also been increased. The concentration of wax in hulls of improved hybrids
may be as high as 34%, compared with 1% in hulls of the traditional seeds.
Around 83% of the wax content is in the seed hull, 17% in the seedcoat and traces
are in the seed.
In order to facilitate oil extraction, either through pressing or by solvent extrac-
tion, it is necessary to leave a certain amount of hulls in the seed. The wax content
is thus considerably higher in these oils than in oils of traditional seed varieties.
Crude sunflower oils may contain 20003000 ppm of wax, depending on the
seed type and the oil extraction method employed.
5.3.2.1. Cold Stability of Refined Sunflower Oil Sunflower oil waxes are fatty
alcohol esters of fatty acids. Their melting point is around 75 C, and their solubility
in the oil is low, leading to the appearance of turbidity in the refined oil with
decreasing temperature. An oils cold stability is usually assessed by means of
the cold test (method AOCS Cc 7-25). Oils passing the cold test will remain
clearwithout the appearance of turbidityafter 5.5 hours permanence at 0 C.
The solubility of waxes in sunflower oil is shown in Figure 21 as a function of
14
12
Wax solubility (ppm)
10
0
0 2 4 6 8 10 12 14 16 18 20 22
Temperature (C)
Figure 21. Solubility of sunflower oil waxes (ppm) as a function of temperature ( C) [based on
(57)].
694 SUNFLOWER OIL
10
9
8
Turbidity appearance (days)
7
6
5
4
3
2
1
0
6 8 10 20 30 40 50 60 70 80
wax content (ppm)
Figure 22. Time required for turbidity appearance in a sunflower oil stored at 0 C as a function
of the wax content [based on (58, 59)].
temperature [based on Bloch (57)]. The solubility of waxes is extremely low (in
the order of 0.1 ppm) at 0 C and increases to 12 ppm at 20 C. It is possible that
oils that are clear at room temperature may develop some cloudiness if stored in a
refrigerator.
Another problem concerning refined sunflower oil is that the wax precipitate
may appear several days upon elaboration, even for oils successfully passing quality
control checks carried out by means of the cold test. The precipitate, although not
affecting the nutritional or organoleptic properties of the oil, will be considered as
an impurity by the consumer and should be avoided for oils commercialized in
transparent bottles.
Consequently, other factors than the wax content influence the time required for
the appearance of turbidity. Figure 22 shows the time required for turbidity appear-
ance in sunflower oil stored at 0 C as a function of the wax content. All oils passed
the cold test, i.e., remained clear after 5.5 hours at 0 C, except the sample contain-
ing 80 ppm of wax (58, 59). Turbidity develops in an oil containing 6 ppm of
wax after 10 days, that is, a longer time period than that considered by the cold
test.
The phenomenon of turbidity appearance in sunflower oil is complex. The time
necessary for turbidity appearance for a given wax content depends on the temper-
ing temperature. Both the time necessary for the appearance of turbidity at tempera-
tures above 0 C and the minimum concentration causing turbidity may be expected
to increase with temperature. However, the wax crystallization rate is reported to be
the highest at 13 C, i.e., the time necessary for the appearance of visible turbidity in
an oil is the shortest at this temperature (58, 59). In view of the above difficulties,
25
cold pressed
20
hot pressed
hexane extracted
Percentage
15
10
0
36 37 40 41 42 44 46 48
Wax (carbon number)
Figure 23. Influence of type of extraction on wax composition in sunflower oils [based on (53)].
several analytical methods have been developed for predicting the appearance of
turbidity (6062) or for determining the wax content in sunflower oil (53, 63, 64).
5.3.2.2. Content and Composition of Sunflower Oil Wax Both the wax content
and composition of sunflower oil depend on the method of oil extraction. According
to work carried out by Carelli et al. (53) with sunflower oils extracted from the same
seed lot, the crude industrial oil obtained by hexane extraction contained 1073 ppm,
the crude industrial oil obtained by hot pressing contained 947 ppm, and the cold-
pressed oil obtained in the laboratory had 771 ppm of wax. That is, the oil extracted
by hexane had the highest wax content, the hot-pressed oil having a higher content
than the cold-pressed oil.
Figure 23 shows the composition of wax esters according to total number of car-
bons, for the oils obtained by the three above methods. Waxes in the cold-pressed
oil were composed predominantly of esters below 42 carbons. Both hot-extracted
oils had a similar wax profile. Wax extractability appears to depend largely on tem-
perature, in particular for those waxes containing over 42 carbons, where reductions
of 70% can be achieved. Degumming did not lead to any significant reduction in the
wax content for either type of industrial oil. In short, the content and composition of
sunflower waxes are affected by the extraction method, although degumming does
not have a significant effect on the total wax content.
Carelli et al. (53) studied the wax content of crude sunflower oil (995 ppm) and
of three commercially refined edible oils (366624 ppm), finding a high degree of
dependence of the wax content on the refining process conditions. The wax com-
position in the crude oil and in one of the refined oils is shown in Figure 24. Waxes
remaining after refining are richer in 40 and 41 carbon esters (less rich in esters of
696 SUNFLOWER OIL
40
35 crude oil
30 refined oil
25
Percentage
20
15
10
0
36 37 40 41 42 44 46 48
WAX (carbon number)
Figure 24. Wax composition in a crude oil and in a commercial refined oil [based on (53)].
over 42 carbons) than the original waxes, indicating a clear tendency of waxes of
higher molecular weight (higher melting point) to crystallize during refining; i.e.,
the cooling stage produces a fractionation of the waxes. Figure 25 shows the com-
position of waxes precipitating in the dewaxing process of sunflower oil [based on
40
35
30
25
Percentage
20
15
10
0
42 44 46 48 >48
Wax (carbon number)
Figure 25. Composition of precipitated waxes in the dewaxing process of sunflower oil [based
on (53)].
50
45
Carelli et al. (53)
40
35 Liu et al. (65)

Percentage
30
25
20
15
10
0
16:0 18:0 18:1 18:2 19:0 20:0 22:0 24:0 26:0 28:0 30:0
Fatty acid (carbon number)
Figure 26. Fatty acid composition of sunflower wax (53, 65).
Carelli et al. (53)]. The predominance of esters of higher number of carbons (over
44) is clear, in accordance with the above.
Because the composition of the wax remaining in an oil differs markedly from
that precipitating from it, care must be taken against differences in the sunflower
wax profile as reported in the literature. The fatty acid composition and the fatty
alcohol composition of sunflower wax (53, 65) are compared in Figures 26 and 27.
35
Carelli et al. (53)

30
Liu et al. (65)
25
Percentage
20
15
10
0
16 18 19 20 22 24 26 28 30 32
Fatty alcohol (carbon number)
Figure 27. Fatty alcohol composition of sunflower wax (53, 65).
698 SUNFLOWER OIL
It is worth noting that Liu et al. (65) studied the wax removed from an oil that had
been stored at 0 C for 1 week, allowing enough time for crystallization and frac-
tionation to take place. The wax contained preferentially esters of a higher number
of carbons. Carelli et al. (53) extracted quantitatively all the waxes contained in a
crude sunflower oil.
5.3.2.3. Sunflower Dewaxing Procedures Several methods are used for wax
removal. Most widespread are those associated to the refining process, whereas
cold degumming and superdegumming, both related to the treatment of crude
oils, are scarcely used (57, 66).
Cold neutralization was developed for high-capacity plants and for the proces-
sing of oils of high wax content. The crude oil is neutralized at a low temperature,
and the wax is removed simultaneously with the soapstock. The processed oil is of
good quality, although losses may be considerable, especially with oils containing
over 1.5% free fatty acids.
Water dewaxing is the latest method of sunflower oil dewaxing. As the crystal-
lization process of waxes is inhibited or delayed by the phosphatides in the oil,
dewaxing is carried out after full degumming has been completed. The process
of dewaxing through hot neutralization followed by cold washing has gradually
replaced the traditional methods. After conventional hot neutralization is per-
formed, and the soapstock removed, the oil is cooled to 68 C and left to settle
for 810 hours. A small percentage of soda (NaOH) is added; time is allowed
for maturation; it is heated to 2025 C and centrifuged. One final filtration step
must be performed on the cold oil to remove the remaining waxes.
Cold filtration may be performed before or after deodorization. Oil from the
bleacher or the deodorizer is cooled to 1215 C, and after settling for 12 hours,
it is cold-filtrated with the addition of filter aid (perlite or diatoma) to prevent clog-
ging caused by the wax.
5.3.3. Physical Refining The success of the physical refining stage depends lar-
gely on the pretreatment of crude oil. Among other compounds, phosphatides must
be efficiently removed. The elimination of nonhydratable phosphatides (NHP),
mainly Ca and Mg salts of phosphatidic acid and lysophosphatidic acid, poses
one major difficulty to the pretreatment of oil for physical refining. NHP may be
removed from the oil with the addition of an acid, generally phosphoric or citric
acid, or complexation agents for Ca/Mg (preventing the precipitation of insoluble
salts). Treatment with these acids is the basis for several oil pretreatment processes
prior to physical refining.
The content and ratio of NHP in an oil differs significantly depending on the
method applied (Table 17). The NHP content is lower in extracted and mixed
oils (67). The NHP content of sunflower oil is low, and degumming may readily
be accomplished. Dimic et al. (67) studied a simplified process for pretreatment
of sunflower oil with the application of multiple acid degumming stages.
Part of the unsaponifiable matter (such as tocopherols, sterols, and sterolesters)
is distilled together with the free fatty acids during deodorization/deacidification, as
TABLE 17. Content of Total Phosphatides (TP) and Nonhydratable Phosphatides (NHP)
of Crude Sunflower Oil [Based on (67)].
Pressed Oil Extracted Oil Mixed Oil
TP (g/100 g oil) 0.24 1.32 0.70

NHP (g/100 g oil) 0.14 0.04 0.09
NHP*100/TP 59% 3% 13%
well as volatile oxidation byproducts, and flavor components. Ideally, natural oil
components should remain in the oil in the highest possible amount. As much as
85% of tocopherols remain in a finished oil upon physical refining operations car-
ried out at temperatures below 240 C.
Sunflower oil processing byproducts depend on the kind of refining, whether
chemical or physical. The so-called deodistillate of chemical refining of sun-
flower oil can be used as feedstock for obtaining tocopherols and sterols. The toco-
pherol composition of sunflower oil (over 90% alpha-tocopherol and only a low
proportion of the beta and gamma isomers) makes deodistillates of great value
for industrial Vitamin E production. Increased importance has been placed on vege-
tal sterols because they were found to reduce the risk of cardiovascular disease. As
a result, the demand for tocopherols and sterols was increased as food additives.
Deodistillate originated in physical refining, diluted in free fatty acids, is not an
attractive feedstock for Vitamin E and sterol producers. Deodistillate originated in
chemical refining of sunflower oil may contain 57% of total tocopherols, com-
pared with only 12% for deodistillate of physical refining (68).
5.3.4. Deodorization There are no major differences between deodorization

procedures for sunflower oil and other vegetable oils. The loss of tocopherols in
the oil is worth noting for sunflower oil, though. The average tocopherol content
of sunflower oil is medium (4401520 ppm, according to Codex-Stan 210-1999),
nearly all of which is alpha-tocopherol (403935 ppm). Table 18 shows the reduc-
tion in the tocopherol content upon deodorization of sunflower oil (69). Tocopher-
ols may be recovered from the distillate of deodorization.
TABLE 18. Total Tocopherol Content (ppm)

of Sunflower Oil in Different Refining Stages
[Based on (69)].
Refining Stage Tocopherols (ppm)
Crude 823
Neutralized 815
Bleached 843
Dewaxed 903
Deodorized 510
700 SUNFLOWER OIL
5.4. Cold-Pressed Sunflower Oil

Virgin oils currently available on the market are not restricted to virgin olive oil but
include other oils obtained by cold pressing of seed. These oils are appreciated
highly by consumers in view of their nutritional characteristics and flavor (particu-
larly those organoleptic notes that are lost in refined oils). The consumer appreci-
ates the natural characteristics of these oils, as they are not subjected to chemical
treatment. A relatively new market has developed for these oils in the U.K.,
Germany, France, Italy, and Switzerland, among other markets. Sunflower oil is
manufactured and commercialized as cold-pressed or first cold-pressed oil.
Cold-pressed sunflower oil is obtained through mechanical pressing at a low
temperature (3035 C, for example). It has a clear appearance, an agreeable golden-
yellow color, and a typical light flavor. The visible spectrum of this oil indicates
a low content of chlorophylls and carotenoids, characteristic of unbleached oils.
It may be stored in dark containers at room temperature for 1 year without the
appearance of turbiditycharacteristic of sunflower oil extracted with solvent
from undehulled seed. A high tocopherol content constitutes a natural protection
against oxidation, to a higher extent than antioxidants added to commercial oils
(70). De Panfilis et al. (71, 72) also analyzed regular cold-pressed sunflower oils
manufactured by different European countries. Relatively high acidity values
were found for these oils (0.651.59 %), characteristic of virgin oils (unrefined).
6. HYDROGENATION OF REGULAR SUNFLOWER OIL
Regular sunflower oil contains hardly any linolenic acid, a factor contributing to a
high oxidative stability. Light hydrogenation processes (low temperature) are there-
fore unnecessary to increase the stability of sunflower oil, as is the case for soybean
and rapeseed oilaiming at the elimination of linolenic acid, while avoiding the
formation of considerable amounts of trans-isomers. Sunflower oil is hydrogenated
in producer countries for use in the manufacture of shortenings and margarines. It is
a good raw material for the production of hydrogenated fat of relatively flat melting
curve, with melting point in the range of 3236 C. The behavior of sunflower oil in
the hydrogenation process is similar to that of soybean oil, and it does not require
particularly special or different reaction conditions. It does not contain compounds
that may interfere with the reaction, as is the case for rapeseed oil. Mention will be
made here only of those aspects that are specific to sunflower oil, without treatment
of the general conditions valid for hydrogenation of any vegetable oil. Process con-
trol parameters are the refractive index, melting point, iodine value, and solid fat
index.
Topallar et al. (73) studied the density and viscosity of hydrogenated regular
sunflower oil of iodine value up to 82.4. The density of hydrogenated sunflower
oil is slightly higher than that of its nonhydrogenated counterpart, varying linearly
with temperature. At a given temperature, oil viscosity was found to double upon
hydrogenation.
HYDROGENATION OF REGULAR SUNFLOWER OIL 701
45
40
0.03%
35
0.05%
30
trans fatty acids (%)
0.10%
25
20
15
10
0
120 C 150 C 180 C
Figure 28. Content of trans-isomers in a hydrogenated regular sunflower oil at different

temperatures (120 C, 150 C, and 180 C) with varying amounts of a nickel catalyst at 20% on
silica (0.03%, 0.05%, and 0.10% nickel in the oil) [based on (74)].
Around 40% of trans-isomers are formed in the traditional hydrogenation pro-

cess. Figure 28 shows the amount of trans-isomers formed as a function of hydro-
genation temperature (120, 150, and 180 C) and the amount of nickel catalyst at
20% on silica (0.03, 0.05, and 0.10% nickel in the oil), for reactions taking
place in the laboratory at a pressure of 3 atm and an agitation rate of 1500 rpm (74).
Hydrogenation of vegetable oils at a low temperature (105120 C) with a high
hydrogen concentration on the catalyst results in a minimum amount of trans-iso-
acid formation. Medium-temperature hydrogenation (150 C) may be unselective,
and high-temperature hydrogenation (180 C) generates the highest content of
trans-isomers, a value near equilibrium. Sunflower oil shows the expected behavior,
with a minimum of trans-isomers corresponding to the lowest temperature and the
least amount of catalyst, and a maximum at the highest temperature and the highest
amount of catalyst.
Variations of the hydrogen pressure and the amount of catalyst, keeping tem-
perature at a constant 180 C and the rate of agitation at 750 rpm, did not result
in significant modification in the trans-content. Figure 29 shows this effect (74).
The traditional hydrogenation process generates a high proportion of trans-
isomers. Having considerably higher melting points than their cis-isomers, trans-
isomers contribute to a large extent to an enhancement of product thermal behavior
and plasticity. On the other hand, from the nutritional viewpoint, it is recommended
to avoid intake of trans-isomers; thus, ways have been sought to reduce their con-
tent in margarines. One widely used procedure consists full hydrogenation of the oil
and later interesterification with nonhydrogenated oil. A new method consists
hydrogenation with solvent under supercritical conditions, leading to a reduction
702 SUNFLOWER OIL
50
45 3 atm
40 5 atm
35
trans fatty acids (%)
30
25
20
15
10
0
0.03% 0.05% 0.10%
Figure 29. Content of trans-isomers in a hydrogenated regular sunflower oil at different
pressures (3 and 5 atm) and with different amounts of nickel catalyst at 20% on silica (0.03%,
0.05%, and 0.10% nickel in the oil) [based on (74)].
in the trans-content below 5%. The rate of hydrogenation is some 1000 times
greater owing to an increased transfer of hydrogen toward the catalyst and the
lack of blockage of pores in the catalyst produced by stagnant oil (74).
Catalytic transfer hydrogenation (CTH) is one method currently being devel-
oped. This method differs from traditional hydrogenation with molecular hydrogen
in the use of a hydrogen donor (for example, sodium formate solution) in a catalytic
transfer reduction reaction. Naglic et al. (75) used this procedure to hydrogenate
regular sunflower oil, among other vegetable oils.
Hydrogenated sunflower oil may suffer from disadvantages for use in the man-
ufacture of margarine spreads. The appearance of sandiness in these products dur-
ing storage is because of a strong tendency of partially hydrogenated sunflower oil
to form b crystals. Under rapid cooling conditions, as used in the manufacture of
margarines, the a phase is the first to form, followed by its rapid transformation into
phase b. This is the desirable crystalline form, because, under certain conditions, it
tends to form a fine three-dimensional network capable of immobilizing a large
amount of liquid oil. From a thermodynamic viewpoint, phase b is the stable
form of hydrogenated sunflower oil. Thus, the solid state transformation of b0
into b always takes place, leading to the formation of a coarse, sandy texture
detected during storage but not during product elaboration. The transformation of
b0 to b is slow. The time required for full transformation depends on temperature
and crystallization rate of the manufacturing process, as well as storage temperature
of the margarine (7679).
Various procedures have been proposed to solve the problem of graininess of
margarines manufactured with hydrogenated sunflower oil. The hydrogenated oil
STORAGE AND DETERIORATION OF SUNFLOWER OIL 703
may be mixed with partially hydrogenated cottonseed oil with the consequent
increase of fatty acids of over 16 carbons. Cottonseed oil, crystallizing at a higher
temperature than hydrogenated sunflower oil, induces and stabilizes b0 crystalliza-
tion in the rest of the fatty material (76). Food emulsifiers provide another solution
to the problem. Addition of saturated and unsaturated fatty acid monoacylglycerols,
acting as modifiers of the crystalline structure, aids in preventing the undesired phe-
nomenon. Also, the addition of 0.3% of sorbitan tristearate inhibits the transition
from b0 to b in a margarine. The same is observed for sucrose polyesters (80).
7. STORAGE AND DETERIORATION OF SUNFLOWER OIL
Several general factors affect the oxidative stability of sunflower oilas well as
most vegetable oilsduring storage. One of these factors is the degree of unsatura-
tion, i.e., the relative content of oleic and linoleic acids. Product shelf-life is
affected by manufacturing conditions such as the type of extraction process (press-
ing, with solvent, with supercritical fluids), degree of purification (crude, refined,
deodorized, etc.), addition of antioxidants, and type of packaging (container mate-
rial, incorporation of inert atmosphere, etc.). Other major factors influencing the
oxidative stability are the particular storage conditions: time, temperature, and
light, among others.
7.1. Composition of the Refined Oil

Sunflower oil extracted from different types of hybrid may have different composi-
tions. It is expected that the degree of unsaturation will influence the oxidative sta-
bility of sunflower oil markedly. AOM time measurements were used to determine
the influence of oleic/linoleic ratio on the oxidative stability of sunflower oil (36).
Oil samples extracted (refined and deodorized) from three progenies of cultivar Per-
venets were analyzed, as well as other oil samples from different regions of the
United States. The oleic acid content thus ranged from 18% to 89%, and the linoleic
acid content decreased from 69% to 1%; the saturated fatty acid content was nearly
constant. AOM values increased from 11 hours for the oil containing the least
amount of oleic acid to 100 hours for that with the oleic acid highest content. These
results show the importance of monounsaturated fatty acid content on the oxidative
stability of sunflower oil.
Inherent stability of an oil may be calculated from the oil composition and the
relative rates of oxidation of oleic (defined as 1), linoleic (evaluated as 10), and
linolenic (evaluated as 25) acids. The higher this value, the more unstable or sus-
ceptible to oxidation an oil is. Values of 6.8 and 1.9 are obtained for regular and
high-oleic sunflower oils, respectively, taking into account their standard composi-
tions. The inherent stability of olive oil, calculated as 1.5, is slightly lower than that
of high-oleic sunflower oil, whereas even lower values are obtained for saturated
fatty materials like tallow (0.86), palm kernel (0.27), and coconut (0.24). The
704 SUNFLOWER OIL
inherent stability of high-oleic sunflower oil is 3.5 times higher than that of high-
linoleic sunflower, and higher than the values for most vegetable oils (81).
7.2. Storage Conditions

Three major factors influence the storage conditions of oils: temperature, light, and
the presence of dissolved oxygen. The combined effect of temperature and light
exposure on the stability of a refined commercial high-linoleic sunflower oil was
studied (82). Oil bottled in PET (polyethylene terphthalate)the head space being
filled with nitrogenwas stored at three different temperatures (35, 45, and 60 C),
both in a dark chamber and with 12-hour lighting per day (18-W fluorescent tube).
Lighting conditions were chosen considering bottles packed in cardboard boxes
(darkness), or displayed on supermarket shelvesexposed to artificial lighting.
The criterion for establishing oil deterioration was the rejection of an oil sample
by a panel of trained consumers. The corresponding shelf-lives were estimated for
the three temperatures (35, 45, and 60 C). These values were used to extrapolate
shelf-life at 20 C (taken as room temperature). Storage temperature was found to
have a marked influence on oil deterioration, even when bottled under inert nitrogen
atmosphere. For example, an oil stored in the absence of light at 45 C is rejected
by consumers after 102 days, compared with 1140 days estimated for oils stored
at 20 C.
Estimated shelf-lives for those samples stored under conditions of darkness were
considerably higher than for the corresponding samples exposed to light. Estimated
shelf-life at 20 C was 281 days for oil exposed to light and around 1140 days for oil
stored in dark chambers. This value is higher than the 2 years generally established
by manufacturers as useful life period of this kind of oil. Exposure to light must be
avoided to extend the life-span of sunflower oil, and it must be bottled in containers
that prevent the passage of light. PET bottles do not prove efficacious light filters in
preventing oil deterioration.
Other researchers (83) have studied the effect of light exposure on regular sun-
flower oil (iodine value 132) and on partially hydrogenated sunflower oil (iodine
value 82). Oil samples were stored in containers of glass, PET, and glass covered
by tin, and were exposed directly to solar radiation at atmospheric conditions for
30 days (no indication is given of either temperature or exposure time). As expected
for each type of container, the increase in the peroxide value was higher for the
nonhydrogenated oil than it was for its hydrogenated counterpart. A rapid increase
of the peroxide value was reported for those samples stored in glass containers,
whereas an intermediate value (yet high) was for PET containers, with practically
no deterioration of those samples in metal-covered containers. These results are in
accordance with the work of other authors (82).
Values denoting oxidative deterioration increase gradually during the shelf-life
of an oil. Determinations of induction times in rapid-aging tests also reflect the oxi-
dative stability of commercial refined oils. Grompone et al. (84) studied the stability
of commercial regular sunflower oils. The oil samples were stored at ambient con-
ditions in their original PET bottles for a one-year period. The OSI induction time
STORAGE AND DETERIORATION OF SUNFLOWER OIL 705
at 110 C was found to decrease from 4.6 hours (at the time of purchase of the oil) to
3.3 hours (one year later). A similar behavior was observed for two refined oils pro-
duced by the same manufacturer over a one-year time difference. The OSI times of
both oils were determined simultaneously: 3.9 hours for the oil procured that year
and 3.2 for the sample procured in a previous year (85).
7.3. Presence of Free Fatty Acids

The pro-oxidant effect of free fatty acids in vegetable oils is widely known. To
determine their influence (86), pure oleic acid was added (up to 3%) to a high-oleic
sunflower oil extracted by cold-pressing. The effect intensity was found to be
directly related with the concentration of free fatty acids.
7.4. Addition of Antioxidants

The effectiveness of antioxidants depends on the type of fatty material; tests are,
therefore, required for each case. Stabilization studies have been reported, for
example, of refined high-linoleic sunflower oil with the addition of three antioxi-
dants and blends of them (87). Butylated hydroxytoluene (BHT), propyl gallate
(PG), and tert-butylhydroxyquinone (TBHQ) were evaluated as antioxidants
through determinations of induction time by Rancimat method at 98 C and the
changes in the properties (acidity, peroxide value, anisidine value, polar compounds
content) of the oil samples during storage in glass jars without lid, in the dark, at
constant temperature: 30 days at 47 C and 30 C and 10 days at 67 C.
The stability of oils containing TBHQ and PG, measured by Rancimat induction
time, increased rapidly with increasing antioxidant concentration. An induction
time of 12.7 hours for the original oil was increased to 25.7 hours with the addition
of 50 ppm of TBHQ and to as much as 42.2 hours with the addition of 200 ppm. PG
had a lesser stabilizing effect, with induction times of 18.8 hours at 50 ppm and
32.3 hours at 200 ppm. BHT had a slight stabilizing effect with practically no mod-
ification of induction time: increasing only to 14.2 hours at 200 ppm. Thus, TBHQ
proved 1.7 times more effective than PG and 22 times more effective than BHT. No
cooperative interaction (synergism) was observed between antioxidants through
Rancimat method.
Different results were obtained with respect to the effectiveness of antioxidants
for oil stored at 47 C in the dark. Effectiveness was evaluated as the time required
for a sample to reach a peroxide value of 10 meq/kg, as estimated from curves of
peroxide value as a function of storage time. The estimated effectiveness of TBHQ
was approximately 9.6 times higher than that of GP and 51 times higher than for
BHT.
According to these results, the relative antioxidant effectiveness of BHT
increased with decreasing temperature, and it decreased for GP. The Rancimat
method underrated the stability of those samples containing BHT, perhaps because
of volatilization of antioxidants during Rancimat testing. The relative activity of
706 SUNFLOWER OIL
antioxidants depends on both temperature and the conditions of the method used to
measure such activity.
Independent of the method and test conditions employed, TBHQ proved to be
the most effective of all three antioxidants, followed by PG, and BHT had little anti-
oxidant effect on the sunflower oil studied. Further research should focus on eva-
luation of antioxidant effectiveness through methods of oxidative experimentation
in conditions similar to those of storage.
7.5. Effect of Heating and Microwaves

Vegetable oils exposed to high temperature undergo rapid deterioration, as can be
evidenced both through chemical and sensory analysis. For example, a refined high-
linoleic sunflower oil kept at 60 C for one hour showed changes in its ultraviolet
absorption spectrum, with the appearance of a peak at 234 nm corresponding to the
formation of conjugated dienes generated in the oxidation of linoleic acid. Another
peak was observed at 268 nm for oil kept at 180 C for 1 hour, corresponding to
oxidation byproducts, especially ethylene diketones (84).
The development of rancidity in sunflower oil is accelerated with increased tem-
perature, the process being even more complex in microwave cooking. To deter-
mine the influence of combined factors (88), both a regular sunflower oil and a
high-oleic sunflower oil were exposed to different heating conditions: (1) micro-
wave heating at approximately 170 C for 120 min; (2) conventional electric heating
at 180 C for 120 min; (3) exposure to microwave radiation for 120 min at tempera-
tures lower than 40 C. Greater degrees of alteration were found for both oil types
when microwave-heated than if heated in a conventional oven, and exposure to
microwaves without heating did not produce significant alteration. Hence, the
increase in temperature was found to be the main deterioration factor for both oil
types.
7.6. Crude Sunflower Oil

Crude sunflower oils are produced via mechanical pressing or through extraction
with hexane, followed by water degumming. Most oils are stored for relatively
long periods prior to refining operations. Cold-pressed crude oils, of superior edible
quality, are commercialized as such. Whether unrefined or cold-pressed, the oil may
be subject to temperature fluctuations.
The influence of composition, storage temperature (30, 47, and 67 C), and oxy-
gen concentration (open and stoppered bottles containing different amounts of oil,
and under nitrogen atmosphere) was studied for degummed crude sunflower oils
obtained by pressing and by hexane extraction (89, 90). The oil obtained by solvent
extraction had a lower rate of oxidation than the cold-pressed oil, although initially
at a more advanced step of deterioration. Although all oils had equal amounts of
unsaturated fatty acids and the same concentration of natural tocopherols, the sol-
vent-extracted oil had a higher concentration of phosphorous related with the phos-
pholipid content. It appears that differences in oxidative stability may be attributed
USES OF SUNFLOWER OIL 707
to the phospholipid concentration, having synergic activity and a metal scavenger

capacity.
OSI induction times were determined at 110 C for three virgin high-linoleic sun-
flower oils, corresponding to three different manufacturers (85). These OSI times,
ranging between 3.8 and 4.7 hours, were equal to or higher than the value obtained
for a refined sunflower oil of the same fabrication year (3.9 hours), yet lower than
the value for a virgin olive oil (8 hours). Virgin oils, which are not subjected to
physical or chemical processes, retain their natural antioxidants, a fact that might
explain the high oxidative stability of these oils.
7.7. Sunflower Oil Extracted by Supercritical Fluids

Although several studies have been made on the extraction of oil from seed by
supercritical fluids, few are about sunflower oil (91). Reports on supercritical car-
bon dioxide (SC-CO2) over a wide range of pressure (2070 Mpa) and temperature
(4080 C) show a maximum solubility of sunflower oil in supercritical CO2 at 80 C
and 70 Mpa, conditions similar to those obtained for other seed oils. Over 90% of
the oil content of seed can be removed under these conditions (92).
One disadvantage of oils obtained by this method is their low oxidative stability
(91). To determine the causes of such instability, one study was made of sunflower
oil extracted with SC-CO2, finding that, although the composition and organoleptic
properties of these oils were similar to those of hexane-extracted oils, their oxida-
tive deterioration rate was higher (93). The results of this study suggest that the
instability of oils extracted with SC-CO2 may be attributed to the oxygen contained
in the supercritical fluid, the tocopherols in the oil being inefficient in preventing
oxidation. Thus, the oxidative stability of sunflower oil decreased when re-extracted
with SC-CO2. Improvements may be obtained with the addition of traces of ascor-
bic acid.
8. USES OF SUNFLOWER OIL
8.1. Food Products

8.1.1. Regular Sunflower Oil In view of its light flavor, relatively high oxida-
tive stability, and light golden-yellow color, sunflower oil finds many applications
both in domestic and industrial levels. In countries where sunflower oil is a com-
mon oil, it is used mainly as a salad oil and as a cooking oil. Industrial applications
of sunflower oil include use as frying oil, as well as in the manufacture of mayon-
naise and oil-based dressings.
8.1.2. High-oleic Sunflower Oil Formulations A high oxidative stability is

required by non-dairy coffee creamer formulas for an extended shelf-life without
refrigeration. Both high-oleic sunflower oil and monoacylglycerols obtained from
it prove stable products for use in this kind of creamer. The oil may also be used in
fluid margarine, fluid spreads, dressings, and so on.
708 SUNFLOWER OIL
8.1.3. Margarines Owing to the strong tendency of hydrogenated sunflower oil

to crystallize in the b form, precautions must be taken to avoid the problem of san-
diness. The addition of crystal-modifying agents delays the transformation from
unstable a phase to b phase, or stabilizes the intermediate meta-stable b0 phase.
For optimum creaminess conditions, it is generally recommended to add 515%
of b0 -crystallizing hydrogenated oils to formulations of margarines and of some
shortenings (2). There are studies in the literature (76) about margarine formula-
tions with blends of partially hydrogenated cottonseed oil, partially hydrogenated
sunflower oil, and nonhydrogenated sunflower oil. Hydrogenated cottonseed oil,
having a strong tendency to crystallize in the b0 form, has a favorable influence
in the blend with sunflower oil.
There has been controversy over the consumption of foods rich in trans-isomers.
As partially hydrogenated oils contain high amounts of trans-components, alterna-
tive processes have been sought that do not lead to the formation of trans-isomers
as byproducts. One such alternative is total hydrogenation of an oil (which does not
form trans-isomers) and later blending with liquid nonhydrogenated oil. The oxi-
dative stability of the blend is determined by the instability of the nonhydrogenated
oil. High-oleic sunflower oil appears a most appropriate ingredient in view of its
low content in polyunsaturated fatty acids. Products of this kind are produced by
some manufacturing companies.
Soft nonhydrogenated margarines containing high-oleic sunflower oil and other
oils are available. Parmalat Canada, for example, produces a soft nonhydrogenated
margarine trademarked under the name Olivina, containing a blend of refined olive
oil, high-oleic and regular sunflower oil, and canola oil, as well as palm, and palm-
kernel oils (94).
Another alternative for increasing oxidative stability while improving margarine
texture consists in the interesterification of hydrogenated and nonhydrogenated oils.
As an example, an outstanding increase of crystals in the b0 form was observed
upon interesterification of a blend of sunflower oil and fully hydrogenated soybean
oil with a strong tendency to crystallize in the b form. This procedure leads to bases
for zero-trans-margarines of optimum texture (95).
8.1.4. Interesterified Tallow with Regular Sunflower Oil Countries rich in

livestock resources produce large amounts of tallow as a byproduct of the slaughter
industry. Tallow is used in the manufacture of food, mainly bakery products and
cookies. Among other reasons, a high hardness texture, and a melting point above
the mouth temperature, make foods manufactured with tallow of lower quality.
Interesterification of tallow with vegetable oils is an alternative for an improvement
in the behavior of tallow.
Studies have been published on chemical interesterification of tallow with reg-
ular sunflower oil (9698). Changes in the proportion of sunflower oil in blends
with tallow produce little modification of the solid content or the melting point.
However, important modifications take place in the thermal properties of these
blends when they are subjected to a process of chemical interesterification. In addi-
tion, modifications occur in the crystalline structure. Solid beef tallowcomposed
of large spherulites responsible for a sandy mouthfeelmay be interesterified with

sunflower oil into a liquid blend (at process temperature), which solidifies in the
form of tiny crystals, giving the solid product enhanced consistency and texture.
Lipase catalysis constitutes an alternative to the method of interesterification
using chemical catalysts. Products of different physical properties may be obtained,
according to lipase specificity (99,100).
8.1.5. Frying with Regular and Modified Sunflower Oil Deep-fat frying, a
common cooking procedure, may be performed in a continuous or discontinuous
manner. In repeated discontinuous fryingeither in the home or in restaurants
the oil remains hot for long time periods, in contact with the surrounding air,
with occasional cooking. The process is carried out with a relatively low rate of
fresh-oil supply (turnover).
In continuous frying, generally performed in industrial facilities for processing
of fried and prefried foods, turnover is high because of the large amount of oil
removed by products during continuous cooking. Contact of the oil with oxygen
is limited in this case by the steam protection barrier generated in the cooking
process.
The deterioration of deep-frying oil depends on a large number of factors,
including frying frequency (discontinuous or continuous), turnover, the time the
oil is hot, and type of oil. Oil decomposition products absorbed by fried foods
together with the frying oil during the cooking process affect not only the sensory
and nutritional quality of these foods but also their shelf-life. They include polar
compounds, triacylglycerol polymers and dimers, diacylglycerols, peroxides, vola-
tile compounds, and so on. Storage time and characteristics depend on the type of
product. Crisps, for instance, are generally stored at room temperature, whereas
prefried french fries are stored in a freezer.
Frozen prefried foods are prepared before ingestion. Deep-fat frying is a most
common cooking method for the manufacture of prefried foods. These foods are
thus subjected to two different frying processes and a stage in a freezer prior to final
cooking and consumption. Several studies have been performed to evaluate the
oxidative resistance of sunflower oils when used in frying processes.
8.1.5.1. Regular Sunflower Oil High-linoleic sunflower oil is of customary use

for frying in strong producer countries, as is the case of Argentina and Uruguay.
In contrast, consumption and cooking use of olive oil has decreased in traditional
consumer countries mainly because of price reasons. In Spain, the consumption of
olive oil dropped from 55 g/capita/day in 1964 to 25 g/capita/day in 1987. Olive oil
has been replaced by seed oils, especially sunflower oil (101, 102).
Although several studies have been performed to determine the characteristics of
regular sunflower oil for use in frying, few were focused on the in situ stability of
an oil contained in fried foods. In one study, french fries fried in high-linoleic
sunflower oil and crushed to different sizes, were subjected to fast-aging in OSI
equipment at 110 C and the induction period was determined (84). The values
obtained differed from those for pure sunflower oil. This method is a more realistic
710 SUNFLOWER OIL
representation of the deterioration phenomenon, as it enables the study of an oils

oxidative stability in the intact food matrix.
Recent publications are based on the simulation of frying operations without the
addition of food (103, 104). Others consider the process with the addition of food,
generally french fries. The deterioration of sunflower oil was determined in the fry-
ing oil during discontinuous frying processes (101, 105), or continuous processes
with different rates of turnover (106, 107), as well as in the oil extracted from fried
foods (101, 108, 109). Other studies are on the deterioration of a sunflower oil
absorbed in foods during storage in normal or fast-aging conditions.
Results show that regular sunflower oil rarely reaches a critical value of 25% of
polar compounds in continuous frying processes with frequent turnover, indicating
the suitability of sunflower oil for this use. However, use of this oil is recommended
only for frying of crisp-type foods with short commercialization periods, ensuring
their consumption before detectable levels of deterioration of the absorbed oil are
reached (109).
8.1.5.2. High-Oleic Sunflower Oil Studies on the use of high-oleic sunflower oil
in frying processes have similar characteristics to those performed on regular sun-
flower oil, and generally they include a comparison of both types (104, 109119).
Several reports have been published on oil deterioration in deep-fat frying processes
carried out in intermittent manner (102, 114, 119) or continuously (110, 114, 119).
Some studies simulate the frying process (in the absence of food) using Rancimat
equipment without air bubbling at 180 C (104), or simply study oils heated in con-
vection oven and on a hot plate (113). Other studies are on oil extracted from fried
foods, generally potato crisps and french fries (111), and on the deterioration of oils
contained in fried foods during storage in normal and fast-aging conditions (109,
112, 115, 116, 119).
Marquez-Ruiz et al. (119) compared regular and high-oleic sunflower oils in
continuous and discontinuous potato frying processes. Oil deterioration was mon-
itored in the fryer as well as in the absorbed oil, during the time of cooking and
during 30 days storage at 60 C. Several conclusions may be reached: (1) high-oleic
sunflower oil deteriorated to a lesser extent than regular sunflower oil, in either con-
tinuous or discontinuous frying; (2) products fried in high-oleic sunflower oil and
stored at 60 C were more stable than those fried in regular sunflower oil; (3) anti-
oxidant protection was essential for fried products requiring storage prior to their
consumption.
High-oleic sunflower oil is the most appropriate type for industrial frying, as
determined through customary deterioration indices and sensory evaluation of final
fried products. Both high-oleic and regular sunflower oils are feasible alternatives
for processing of prefried products stored in freezer; i.e., no sensory differences
were found in frozen precooked french fries stored for 19 months at 18 C,
shelf-life for this kind of product being normally 2 years. However, for crisps stored
under ambient conditions, those fried in regular sunflower oil had developed detect-
able levels of rancidity after 4 months, the process evolution being both more rapid
and intense. Despite the above difference, the legal commercialization period for
crisps is usually 3 months, making both oils suitable for manufacture of this kind of
product.
Studies have been made on the effect of additions of dimethyl polysiloxane
(DMPS; 2 mg/kg) on frying oil performance and the storage of fried potato as crisps
or french fries (109114). DMPS is an antifoaming silicone forming a monolayer
over the oil surface, protecting against oxidation by air contact. The addition of
DMPS is inefficient for both regular and high-oleic sunflower oil in continuous fry-
ing, but it is useful in discontinuous frying where the oil surface is exposed to the
atmosphere for extended time periods.
8.1.5.3. Mid-Oleic Sunflower Oil With the relatively recent appearance of mid-
oleic sunflower oil, researchers have shown an interest in its use in frying processes.
Abidi and Warner (120) used the three types of sunflower oil (regular, high, and
mid-oleic) in the preparation of french fries, potato crisps, and fresh white corn tor-
tilla chips. However, no general conclusions may be drawn regarding inherent sta-
bility as there is no indication of antioxidant type and concentration in the different
oils.
Kleingartner and Warner (48) summarized several studies performed with mid-
oleic sunflower oil in frying. Mid-oleic sunflower oil was used successfully by sev-
eral potato chip manufacturing enterprises. In July 2000, Procter & Gamble
announced it would use mid-oleic sunflower oil in the production of Pringles potato
chips. The U.S. Department of Agricultures National Center for Agricultural Uti-
lization Research in Peoria also carried out performance evaluations in processes of
frying of tortilla chips and of french-fried potatoes. Oxidative stability was evalu-
ated through the appearance of polar compounds during intermittent frying and
through flavor evolution during storage. Three years research showed that mid-
oleic sunflower oil was of higher frying quality than other nonhydrogenated oils
(soybean, canola, corn, and cottonseed).
Pan-frying is a popular frying method at home and in many restaurants. The pan-
fry stabilities of two oils with similar iodine valuesmid-oleic sunflower oil
(NuSun) and a commercial canola oilwere compared (121). Both oils have simi-
lar pan-fry stabilities, with few significant differences in the physicochemical prop-
erties during the heating process.
8.2. Industrial Products

8.2.1. Biodiesel The importance of biodiesel as a partial or total substitute for
petroleum-based fuels has increased in the last two decades. Already in the
1980s, studies had been made in the United States for the use of vegetable oils
(mustard, canola, corn, soybean, peanut) as fuel, sunflower being the preferred
oil. The North Dakota State University, Agriculture Engineering Department, con-
ducted a project to determine the effects of sunflower oil used in diesel engines. The
Iowa State University, College of Agriculture, also conducted studies on the use of
sunflower oil as fuel. Several short studies were carried out with a tractor by the
USDA-SEA-AR Subtropical Texas Area unit in Weslaco (Texas). Other studies
712 SUNFLOWER OIL
were made outside the United States; for example, sunflower oil was tested on a
range of diesel engines in South Africa.
The use of vegetable oils as a substitute for petroleum-based diesel has been
replaced by use of their methyl or ethyl esters. These are produced in several coun-
tries for use in ignition engines, especially blended with petroleum diesel. Having a
high population density and serious pollution problems, the development of biodie-
sel has been stimulated in European countries. A large number of diesel vehicles
constitute a potential market for biodiesel production. In Germany, the first com-
mercial plant was built in 1995 with a capacity of 60,000 ton/year. In 2001,
Germany produced some 500,000 ton of biodiesel. Production of methyl ester
biodiesel in Italy was 200,000 ton in 2001. In France, several plants currently oper-
ate, especially for rapeseed methyl ester production, with a total biodiesel produc-
tion over 300,000 ton in 2001. Rapeseed and sunflower biodiesels are also produced
commercially in Austria.
Sunflower oil, either regular or high oleic, may be used in biocarburants in the
form of methyl esters. These products were already in use in Italy in 1998, pure or
in blends. In Austria, high-oleic sunflower was used as carburant for tractors. The
French oilseed sector launched an experimental program for the incorporation of
methyl esters of traditional sunflower in fuels, instead of rapeseed esters, as certain
regions of southern France are not suitable for rapeseed cultivation (41). Experi-
ences with sunflower-based biodiesel were carried out in Spain, Greece, and
Portugal, among others.
The optimization of biodiesel production by transesterification of sunflower oil
was studied (122). The best combination of process parameters was found to be
three stoichiometric doses of methanol, 0.28% w/w of KOH, and 70 C temperature.
Several reports have been published on the properties of biodiesel manufactured
with different fatty materials and on their performance in compression ignition
engines, including information about sunflower oil and its esters (30, 31, 41,
123). Table 19 shows major properties of sunflower oil and its methyl esters. The
physicochemical characteristics of these esters meet the norm specifications of dif-
ferent countries, even with improvements of some properties, such as the cetane
number.
Although the production of methyl esters is the easiest alternative, the produc-
tion of ethyl esters from ethanol obtained from renewable starch sources, e.g., corn,
poses a more interesting challenge. However, the production of ethyl esters through
TABLE 19. Properties of Sunflower Oil-Based Biodiesel (30, 41, 123).
Methyl Esters of Methyl Esters of Methyl Esters of

Regular Oil (30) Regular Oil (41) Regular Oil (123) High-Oleic Oil (41)
Cetane number 37 54 49 55
Energy content 39,575 kJ/kg 39,687 MJ/kg 33.5 MJ/l 39,733 MJ/kg
Cloud point ( C) 7.2 1 1 4
Flash point ( C) 183 180 274 185
Density (kg/l) 0.9161 0.8866 0.860 0.8821
WORLD PRODUCTION AND DISTRIBUTION OF SUNFLOWER OIL 713
basic catalysis is difficult to achieve because of the formation of stable emulsions.

Studies have been conducted with a view to finding solutions to the problems
associated with the production of ethyl esters of sunflower oil (124), opening
new alternatives for its use as biodiesel.
8.2.2. Lubricants Mineral-based lubricants lead the lubricant market. However,

with the advancement of the need for biodegradable products, vegetable oils have
become popular, because of they are also better lubricants. The fatty acid composi-
tion of vegetable oils, however, is one major disadvantage. Oils rich in saturated
fatty acids have poor low-temperature flow properties, and those rich in polyunsa-
turated fatty acids are of low oxidative resistance. Vegetable oils rich in monoun-
saturated fatty acids have optimum oxidative stability and low-temperature
properties (125).
In view of the higher oxidative stability of high-oleic sunflower oil, it is used as
diesel and gasoline engine lubricant. In France, for example, a product is commer-
cialized containing 7080% of high-oleic sunflower oil and 2030% other
additives. Several studies have been aimed at the production of polyol esters (propyl
glycol, pentaerytritol, trimethylpropane) from high-oleic sunflower oil (41).
8.2.3. Vegetable Oil-Based Printing Inks Lithography and letterpress pro-

cesses require paste inks. Printing inks that are conventionally used in these appli-
cations are multicomponent systems comprising a pigment, a hydrocarbon and/or
alkyd resin, a hydrocarbon solvent, and optional additives. Vegetable oil-based non-
petroleum inks have been formulated for various specialized applications. These
ink formulations cost even less than petroleum oil-based ink formulations. Sun-
flower oil is used for the manufacture of vegetable oil-based inks (126).
8.2.4. Other Applications Containing around 70% linoleic acid, sunflower oil is
a semidrying oil. Insofar as economically feasible, sunflower oil may replace soy-
bean oil in the manufacture of resins for paint and press-ink formulations. Through
epoxidation of sunflower oil, PVC stabilizers may be obtained, and dimerization
would yield products that could be used for lubricant manufacture (41).
Epoxides have received increased attention in view of their interest both as end-
products and as chemical intermediates. Epoxidized oilsmainly high-oleic sun-
flower oiland their ester derivatives have found important applications as plasti-
cizers and additives for polyvinyl chloride (PVC). Epoxidized esters produced from
high-oleic sunflower methyl esters have hydroxyl values of 0, oxirane values of 5.2/
4.5, and iodine values of 1.7/1.5 (127)
9. WORLD PRODUCTION AND DISTRIBUTION

OF SUNFLOWER OIL
The world production of sunflower seed has grown steadily since 1950, at a lower
rate toward the last several years. Figure 30 shows the evolution since 1935 (5, 128,
714 SUNFLOWER OIL
30
25
Sunflower Seed Production (MMT)
20
15
10
0
1930 1940 1950 1960 1970 1980 1990 2000 2010
Figure 30. World production of sunflower seed in million of metric tons (MMT) (5, 128, 129).
129). World production of sunflower seed in 20002001 was of 23.3 MMT, 8.9
MMT of sunflower oil and 10.2 MMT of sunflower meal.
World production of sunflower seed (7%) was third in the world production of
oilseeds in 20012002, after soybean (57%) and canola seeds (11%). Production of
soybean being by far the largest, sunflower seed production does not amount to
much of the world total oilseed production (129).
The evolution of the sunflower oil production (MMT) in the last few years is
shown in Figure 31 by country/region [based on Gunstone (129)]. The world
9 28
EU-15
8
Total sunflower seed production (MMT)
Central Europe 27
Sunflower seed production (MMT)
Ex.USSR
7
Argentina
26
6 total
5 25
4 24
3
23
2
22
1
0 21
96-97 97-98 98-99 99-00 00-01
Figure 31. Evolution of sunflower seed production (MMT) by country/region in the last years
compared with total world production [based on (129)].
others
22%
soybean
31%
sunflower
8%
canola
14%
palm
25%
Figure 32. Distribution of world production of vegetable oils in 20012002 [based on (129)].
production of sunflower oil has changed (in both distribution and amount) in the last
decade because of economic and political reasons, such as the facts occurred in ex-
USSR and Central Europe. In Argentina, the largest world producer of sunflower oil
until 1999, the substitution of sunflower by more profitable crops since that year
resulted in a drop in sunflower seed production, making ex-USSR the largest world
producer (in this year, the production of sunflower seeds in ex-USSR also increased
markedly). In 2000/2001, EU-15 became the second largest producer. In view of the
amounts produced in Argentina, the decline in sunflower oil production in this
country was reflected in the total world production.
Sunflower oil production is determined by the production of sunflower seed.
Sunflower oil is fourth in importance among vegetable oils (including oils extracted
from fruits, as is the case of palm oil), as shown in Figure 32. World production of
sunflower oil is around 9 million metric tons, amounting to 8% of the total vege-
table oil production [based on Gunstone (129)]. Despite its fourth position in the
world production of vegetable oils, the participation of sunflower oil is fairly low.
Figure 33 shows the evolution of the world production of sunflower oil in the last
several years, compared with the total fat and oil world production (128, 130, 131);
including the projected value for 20082012, according to the literature (132). The
trend in sunflower oil production is for a steady value (declining slightly), and the
total fat and oil production tends to increase considerably.
Figure 34 shows the evolution of the sunflower oil production (MMT) by coun-
try/region in the last years, compared with the total sunflower oil production [based
on Gunstone (129131)]. The leading position in the producer market varies accord-
ing to period, with fairly similar amounts for the European Union and Argentina
(major producer countries), except for a sharp drop in Argentina in 20002001
as a result of the decrease in seed production (showed in Figure 31). As for ex-
USSR, it became the worlds largest producer country as of 19992000. The
716 SUNFLOWER OIL
14 140
sunflower oil
12 120
total
sunflower oil production (MMT)
10 100
total fats and oils (MMT)

8 80
6 60
4 40
2 20
0 0
93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 01/02 02/03 08/12
Figure 33. World production of sunflower oil (MMT) in the last years, compared with the total fat
and oil production (MMT) (Key: Solid box real production. Open box estimated production
(128, 130132).
participation of sunflower oil in the world trade has declined in the last years, as
shown in Figures 31 and 33.
Soybeans have a strong participation in the world supply of oilseeds. The supply
being oriented mainly to the production of meals rich in protein, an oversupply of
soybean oil is commercialized at a lower relative price, a fact that is reflected in the
general composition of the vegetable oil trade.
3 9.8
Total sunflower oil production (MMT)
9.6
Sunflower oil production (MMT)
2.5
9.4
9.2
2
9
1.5 8.8
8.6
1
8.4
8.2
0.5
8
0 7.8
96-97 97-98 98-99 99-00 00-01
EU-15 Ex.USSR Central Europe Argentina total

Figure 34. Sunflower oil production (MMT) by country/region, compared with total world
production of sunflower oil [based on (129131)].
Others
15% EU-15
23%
Turkey
5%
United States
4%
Central Europe
8%
exUSSR
27%
Argentina
18%
Figure 35. Participation by country/region in the world production of sunflower oil in 20012002
(based on (129131)].
Figure 35 shows the participation (%) by country/region in the world production

of sunflower oil in 20002001 [based on Gunstone (129131)]. The largest world
producer was ex-USSR for that period (27%), followed by the European Union
(23%), and Argentina (18%).
The production of sunflower oil may supply either the internal or external market
of a region or country. The world exports distribution does not follow the same pattern
as the distribution of production. The participation in sunflower oil exports is shown
by country/region in Figure 36 for 20002001 [based on Gunstone (129131)].
EU-15
Others 7%
17%
exUSSR
Central Europe
22%
4%
Argentina
50%
Figure 36. Participation by country/region in world exports of sunflower oil in 20012002 [based
on (129131)].
718 SUNFLOWER OIL
For that period, the main exporter was Argentina (50%), with a participation
twice as large as ex-USSR (22%) in exports, whereas the latter was the largest pro-
ducer in the same period. The above figures show how a large proportion of the
Argentinean production (1.18 MMT) finds its way into the external market (out
of a total 1.60 MMT produced). Most of it is exported as bulk crude oil and as
bottled refined oil in a lesser proportion.
The amount of sunflower oil produced by ex-USSR (2.40 MMT) was far larger
than the exported amount (0.53 MMT), showing a large consumption (1.87 MMT)
in this country of the produced oil. It is exported primarily as processed oil,
mainly to the Middle East countries (Algeria, Egypt, Turkey, etc.) for later proces-
sing, demand for this type of oil being scarce in the European market. Figure 37
shows the participation by country/region in world imports of sunflower oil for
20002001 [based on Gunstone (129131)]. A comparison of Figures 35 and 37
shows that major importer countries are generally not strong producers of sunflower
oil.
In the European Union, exports of sunflower oil (0.16 MMT) were balanced
by imports (0.17 MMT), whereas it was a net importer of sunflower seed for pro-
cessing (1.94 MMT). Likewise, Central Europe exported the same amount of
sunflower oil (0.10 MMT) as it imported (0.13 MMT), but it was an exporter of
sunflower seed (0.39 MMT). These relationships of the international European
market are reflected in the characteristics of the internal market. A number of vege-
table oils are available for consumption in the European Union, unlike the case of
Argentina. Production of canola oilthe main oilis closely followed by sun-
flower, soybean, palm, and olive oils, with regional variations. Olive oil is more
important in the south of Europe (Greece, Italy, Spain, and Portugal), amounting
to 36% of the total vegetable oil consumption, sunflower oil also being important.
India
19%
Algeria
10%
Others
59% EU-15
7%
Central Europe
5%
Figure 37. Participation in sunflower oil imports by country/region in 20012002 [based on
(129131)].)
SUNFLOWER OIL EXTRACTION AND PROCESSING BY-PRODUCTS 719
Consumption in the north of Europe is more varied: the market is supplied by

regional oilseeds (especially canola), with sufficient sunflower production in
France. Soybean, sunflower, canola, and palm oils together constitute nearly 80%
of total vegetable oils in northern Europe (133).
In 20002001, Argentina produced 1.60 MMT of sunflower oil and exported
1.18 MMT (of that oil), with 0.42 MMT disappearance, 75% of which was destined
for direct consumption, whereas the rest was used in the manufacture of margarine
and mayonnaise. An analysis of internal demand by oil type in Argentina shows
sunflower oil with 70% of the trade, followed by soybean (26%), and low volumes
of corn and olive oils. Use of sunflower oil is widespread in this region.
Sunflower oil, as well as canola oil, is produced annually for sale of the oil pri-
marily, unlike the case of soybean oil, mainly related to the demand for its meal.
Sunflower oil accounts for 80% of the seed price, being six to seven times higher
than the price of meal pellets. The price of sunflower seed is determined by the
price of the extracted oil.
10. SUNFLOWER OIL EXTRACTION AND PROCESSING

BY-PRODUCTS
Two coproducts are originated in the sunflower-seed oil extraction process: meal
and hulls. Sodium soapstock is obtained as a byproduct of alkali refining of the
oil, and it may contain phosphatides, depending on process type. Sunflower lecithin
is obtained by treatment of the oil. Less important byproducts are waxes, tocopher-
ols, and so on.
10.1. Soapstock
Sodium soapstock obtained from alkali refining of sunflower oil may be used as an
ingredient in animal feed meals in view of its high caloric value, in addition to
being a good phosphorous source in the case of soapstock containing phosphatides.
Use of soapstock is local because of transportation costs and ready fermentation,
resulting from a high water content. Dehydration is often carried out prior to trans-
portation, storage, and incorporation of soapstock into other meals.
Sodium soapstock may also be subjected to treatment with mineral acids, freeing
the constituent fatty acids upon decomposition of the soaps. The product thus
obtained, having a very low water content, is called acid oil. Storage and trans-
portation requirements are the same as for crude oil.
Both acid oil and the free fatty acids obtained in the physical refining may be
incorporated in the manufacture of soap. In view of its high linoleic acid content
(in particular when originating in the refining of regular sunflower oil), soapstock
does not make a fatty material of good properties for the manufacture of toilet soap.
To this end, it is blended in relatively low proportions with other more appropriate
fatty materials. It is used in cattle producer countries also producing sunflower oil
as a means to reduce the titer of beef tallow or of the beef tallow/coconut oil blend.
720 SUNFLOWER OIL
A high titer is one problem of beef tallow in Argentina and Uruguay, for example:
45.4 C mean value, although it may be as high as 48 C (134, 135). Sunflower oil,
having a titer of 1620 C, may be added to this blend to obtain appropriate values
for soap manufacture, generally considered around 42 C.
10.2. Sunflower Lecithin

Responding to the recommendations of food and nutrition scientists, studies have
been made of new phosphoacylglyerols (phosphatides or phospholipid) sources.
Oilseeds and cereals are important sources of phospholipids. The phospholipid
(lecithin) content of crude sunflower oil ranges from 0.5% to 1.2% (53, 136
138). Oils extracted by solvent generally have a higher phospholipid content than
those obtained by pressing. Major phosphoacylglyerols of sunflower oil are PC, PE,
PI, and PA. Most are hydratable and may be removed from the crude oil through a
water degumming process.
Most published research is into the content and composition of unremoved
phospholipids in sunflower oil after different degumming processes. Little
research has been done, however, of the separated lecithins. Some conclusions
may still be reached about efficiency of production method, as well as the possible
composition of these lecithins from the composition of the phospholipids remaining
in the oil.
The fact that the phospholipid composition of sunflower depends on the oil
extraction method and the degumming treatment used to remove them explains
the differences in the reported compositions in the literature. Phospholipid compo-
sitions of sunflower oil are shown by type in Table 20 (136, 139141). The overall
fatty acid composition also varies widely for the same reason. Cherry and Kramer
(140) report composition ranges of 11.131.9% of palmitic acid, 3.07.9% stearic
acid, 13.317.3% oleic acid, and 42.868.7% linoleic acid.
Sunflower lecithin is not produced in considerable amounts worldwide. This is
mainly because of the low lecithin content of crude sunflower oil as compared with
2.9% for soybean, 1.9% for canola, 2.4% for cottonseed, and 2.02.7% for corn oil
(normalized at 70% of insolubles in acetone). Lecithin removal from sunflower oil
may be justified in strong sunflower producer countries. It may be used as a food
additive in view of its high phosphatidylcholine and essential fatty acid content.
Upon refining and fractioning stages, the quality of sunflower lecithin may be
improved for the manufacture of food products and cosmetics.
TABLE 20. Phosphoacylglycerol Composition (%) of Sunflower Oil (136, 139141).
PC (%) PE (%) PI (%) PA (%)
Cherry/Kramer (140) 12.726.8; 42.264.2 9.929.4; 46.6 3.721.4; 24.036.6

Morrison (136) 52.0 19.7 26.0 2.2
Carelli et al. (141) 4448 2021 2021 1215
Chapman (139) 48.7 21.2 27.9 2.2
Hollo et al. (142) report on studies carried out in Hungary (where sunflower oil
represents 80% of the total vegetable oil production) about possible uses of sun-
flower lecithin. It may be added to sunflower meal by 2.5% or it may be dried,
representing an increase to 5570% in the phospholipid content, prior to blending
with animal feed meals. Used in swine feed meals, it leads to an increase of body
weight and a shortened fattening period. It is also useful for an adjustment of the
energetic level of poultry feed meals, and it replaces the addition of synthetic cho-
line chloride, in view of the high natural choline content of sunflower lecithin. It is
also used in the food industry as emulsifier, and as a viscosity reducer agent in the
manufacture of chocolates.
10.3. Sunflower Meal

The remaining material from the pressing stage (expeller) of sunflower seed usually
contains 1012% of oil. Sunflower meal is obtained as a byproduct of solvent
extraction from this material. In order to facilitate handling and transportation,
the meal is often compacted through pressure and temperature treatment, into the
shape of sunflower pellets. Sunflower pellets are the fourth important oleaginous
raw material used in animal feeds, after soybean, cottonseed, and canola pellets.
Production of sunflower pellets in Argentina, the main world producer and
exporter, is around 1,900,000 tons annually, 8085% of which is exported to the
Rotterdam market mainly, and the remaining 300,000 tons are destined to the local
market (143).
The evolution of the world production of sunflower meal (MMT/year) in the
last years (harvests in 1996/1997 through 2000/2001) is indicated in Figure 38,
20
18
16
14
12
MMT/year
10
0
96-97 97-98 98-99 99-00 00-01 06-10 16-20
Figure 38. World production of sunflower meal (MMT/year). (Key: Solid box actual production.
Open box estimated production) [based on (129)].
722 SUNFLOWER OIL
Others EU-15
26% 26%
Central Europe
9%
Argentina
16%
Ex-USSR
23%
Figure 39. Production of sunflower meal by major countries/regions in 20012002 [based on
(129)].
including the estimated projection for the five-year periods 2006/2010 and 2016/
2020 [based on Gunstone (129)]. Figure 39 is a representation of the production
of sunflower meal (MMT/year) by major countries/regions involved. Argentina is
the largest world producer (16% of the world production), although certain
regionsas a group of countriesrepresent larger productions. Exports are usually
from Argentina (68%) and ex-USSR (19%), and most of them reach the EU-15
(66%). Soybean meal leads the world market of oleaginous meals, sunflower
meal representing only 6% (129).
The sunflower oil extraction industry produces three kinds of meal: meal pro-
duced from undehulled seeds, containing around 28% of protein and 2528% fiber;
meal of partially dehulled seed, containing 3537% protein and 18% fiber; and
meal obtained through a two-step dehulling process of seed, containing 4042%
protein and 1214% fiber. The meal composition thus depends on the efficiency
of the dehulling process of sunflower seeds. The oil content of sunflower meals
ranges between 1.5% and 2.5%, depending on oil extraction efficiency and raw
material (3, 144146).
Undehulled sunflowerseed meals cannot match soybean meals in the meal mar-
ket, their use being limited to ruminant feeds. In addition, sunflower hulls contain a
large amount of raw fiber (6065%) of practically no nutritional value, so that they
are almost exclusively used as ruminant feed.
The amino acid composition of sunflower meal is generally balanced. The
energy content of sunflower meal compares favorably with other oilseed meals.
The energy value of a sunflower meal increases with increasing residual oil content
and a reduction of the fiber content. Sunflower meal is also valuable as a calcium
and phosphorous source and a good source of hydrosoluble B complex vitamins,
mainly nicotinic acid, thiamine, pantothenic acid, riboflavin, and biotin (146).
Sunflower meal also contains chlorogenic acid, a polyphenolic compound. For
meals extracted in the conventional manner under alkaline conditions, chlorogenic
acid reacts with a certain protein fraction, giving the product a dark green color.
Several methods have been proposed for the production of protein concentrates
obtained from sunflower meal through extraction or inactivation of chlorogenic
acid (3, 146). Compared with soybean, cottonseed, and peanut meals, dehulled
sunflowerseed meals are substantially less rich in lysine, although richer in methio-
nine and cystine. Sunflower proteins are therefore a good complement for soybean
meals when both are blended in animal feeds. Supplementation with lysine must
generally be performed for sunflower meals to be used alone as swine and poultry
feeds (145).
Sunflower meal (blended with wheat flour) can be used for human nutrition.
Despite their dark color, sunflower protein concentrates are of excellent digestibil-
ity. A method was proposed (147) for obtaining sunflower protein concentrates
from defatted wholemeal sunflower flour, through extraction of these proteins
with a basic solution. The process yields a concentrate of 71% protein (dry basis),
rich in glutamic and aspartic acids. The supernatant liquid, rich in potassium and
phosphorous, can be used as agricultural fertilizer.
10.4. Hulls
Hulls are obtained as a byproduct of sunflower seed processing for oil extraction.
The amount of hull represents around 2228% of seed weight. Hulls may be sepa-
rated either prior to or upon extraction of the oil. They may also remain in the meal,
making a wholemeal product. Hulls contain around 4% of crude protein, 5% of lipi-
dic matter (including waxes, hydrocarbons, fatty acids, sterols, and triterpenic acid),
50% carbohydrates (mainly cellulose and lignin), 26% reducing sugars (mainly
xylose), and 2% ash. The high fiber content (6065%) and the low protein and
energy content of hulls reduce their nutritional value; they are used as roughage
in certain animal feeds (146).
Hulls may be used as ruminant feed when finely milled and blended with other
ingredients, composing the nondigestible part of the meal in view of the high con-
tent in cellulose and lignin. They may be used for an additional volume of concen-
trated meals and to absorb liquids such as molasses. Owing to the high content of
nondigestible fiber, the nutritional value of hulls as feed for farm animals is extre-
mely low, so that they are used as livestock or poultry litter.
Sunflower hulls are also used as fuel, with a caloric power of 19.2 MJ/kg, and the
caloric power of hull and meal is 23.6 MJ/kg, constituting an improved fuel. Hull
burning is an alternative to the use of more expensive fuels in some countries. The
resulting ashes are rich in potassium and may be used as fertilizers (146).
Sunflower hulls may also be pressed and shaped into fireplace logs, including
wood residues. In view of their high content in reducing sugars, sunflower hulls
can be used for the production of ethyl alcohol and furfural. Other minor uses
include building or insulation board.
724 SUNFLOWER OIL
10.5. Waxes
Sunflower seed oil waxes are fatty alcohol esters of fatty acids. The fatty acids are
in the range of 14 to 30 carbons, with a predominance of linoleic (44.0%), oleic
(18.6%), behenic (9.7%), and palmitic (9.8%) acids. The fatty alcohols have
chain-lengths in the range 16 32 carbons, with a predominance of octadecanol
(23.1%), nonadecanol (18.4%), and tetracosanol (11.9%). This fatty acid and alco-
hol profile leads to esters of 36 to 48 carbons, with a predominance of 41 carbons
(14.3%), 48 carbons (13.9%), 36 carbons (13.0%), 40 carbons (11.5%), 46 carbons
(11.2%), and 37 carbons (9.2%). Other authors report sunflower waxes to contain
fatty acids with 1630 carbons (highest occurrence of 20 and 22) and fatty alcohols
in the range of 2032 carbons (highest occurrence of 24 and 26). The saponification
value of these waxes is 8590 and the iodine value of 812. The melting point is
7080 C. They behave as nonpolar lipophilic compounds at temperatures below
40 C, and at higher temperatures, they adopt a crystalline state of weak hydrophilic
character (53, 57, 64).
These waxes are located mainly in the hull of sunflower seeds (1.53%). The
concentration will depend on the hybrid or seed variety, as well as on origin and
storage factors. They are incorporated in the oil during oil extraction operations.
The quantity of extracted wax will depend on the degree of dehulling of seeds,
the extraction method (pressing or extraction by solvent), and the temperature
and technology used. The wax content of crude sunflower oil is usually in the range
0.020.35%, although it may reach values as high as 1%; the wax content of refined
sunflower oil can be as high as 60 ppm (53, 64). Sunflower waxes have been used
successfully as ingredient in livestock feeds, mixed with grain, silage, and so on.
4000
3500
total alpha
3000
gamma delta
Content (ppm)
2500
2000
1500
1000
500
0
corn soybean sunflower peanut cottonseed
Figure 40. Typical tocopherol levels in deodorizer distillate obtained from crude oils [based on
(148)].
REFERENCES 725
10.6. Tocopherols
A low percentage of tocopherols is lost during operations of oil neutralization,
bleaching, and deodorization of sunflower oil. The tocopherols lost during deodor-
ization may be recovered together other volatile compounds, from the deodorizer
distillate. The distillate obtained from sunflower oil is also a good source of phy-
tosterols. Figure 40 shows typical tocopherol levels in deodorizer distillate obtained
from crude oils [based on Walsh et al. (148)].
The distillate obtained from sunflower oil may be sold to pharmaceutical com-
panies for tocopherol and sterol isolation. Tocopherols may be used as natural
antioxidants or may be converted to vitamin E by methylating the heterocyclic
ring. The interest in phytosterols is caused by the high potential of some of them
to inhibit intestinal cholesterol absorption.
REFERENCES
1. H. Autino, F. Wnuk, and P. Vacca, Aceites Grasas, 3, 2130 (1993).

2. M. Melgarejo, Aceites Grasas, 8, 4952 (1998).
3. D. C. Zimmerman, in E. H. Pryde, L. H. Princen, and K. D. Mukherjee, eds., New Sources
of Fats and Oils, AOCS Press, Champaign, Illinois, 1981.
4. G. N. Fick, in G. Robbelen, R. K. Downey, and A. Ashri, eds., Oil crops of the world,
McGraw-Hill, New York, 1989.
5. M. Bockisch, Fats and Oils Handbook, AOCS Press, Champaign, Illinois, 1998, pp. 237
242.
6. F. Ucke, Aceites Grasas, 8, 7073 (1998).
7. S. C. Chapman and A. J. de la Vega, Field Crop Res., 73, 107120 (2002).
8. R. A. Carr, Aceites Grasas, 7, 373388 (1997).
9. A. Merrien, Aceites Grasas, 8, 7580 (1998).
10. A. B. Afzalpurkar and G. Lakshminarayana, J. Amer. Oil Chem. Soc., 57, 105106
(1980).
11. M. Rataus, Aceites Grasas, 5, 335338 (1995).
12. J. A. Robertson, G. W. Chapman, and R. L. Wilson, J. Amer. Oil Chem. Soc., 55, 266269
(1978).
13. J. A. Robertson and R. B. Russell, J. Amer. Oil Chem. Soc., 49, 239244
(1972).
14. J. A. Robertson and V. E. Green, J. Amer. Oil. Chem. Soc., 58, 698701 (1981).
15. J. R. Lajara, U. Daz, and R. Daz Quidiello, J. Amer. Oil Chem. Soc., 67, 618623
(1990).
16. E. Martnez-Force, R. Alvarez-Ortega, S. Cantisan, and R. Garces, J. Agric. Food Chem.,
46, 35773582 (1998).
17. A. Prevot, Rev. F. Corps Gras, 34, 183195 (1987).
18. E. Santalla, I. Riccobene, and S. Nolasco, Aceites Grasas, 44, 175178 (1993).
19. A. Nagao and M. Yamazaki, J. Amer. Oil Chem. Soc., 60, 16541658 (1983).
726 SUNFLOWER OIL
20. Physical and Chemical Characteristics of Oils, Fats and Waxes, in Official Methods and
Recommended Practices of the American Oil Chemists Society, AOCS Press, Cham-
paign, Illinois, 1997.
21. P. Gillatt, Inform, 5, 981986 (1994).
23. E. J. Campbell, J. Amer. Oil Chem. Soc., 60, 387392 (1983).
24. J. B. Rossell, B. King, and M. J. Downes, J. Amer. Oil Chem. Soc., 60, 333339 (1983).
25. R. Alvarez-Ortega, S. Cantisan, E. Martnez-Force, and R. Garces, Lipids, 32, 833837
(1997).
26. K. Warner, T. L. Mounts, and G. R. List, Inform, 1, 326 (1990).
27. M. S. R. Subrahmanyam, H. S. Vedanayagam, and P. Venkatacharyulu, J. Amer. Oil
Chem. Soc., 71, 901905 (1994).
28. H. Abramovic and C. Klofutar, Acta Chim. Slov., 45, 6977 (1998).
29. J. N. Coupland and D. J. McClements, J. Amer. Oil Chem. Soc., 74, 15591564 (1997).
30. Y. Ali and M. A. Hanna, Bioresource Technol., 50, 153163 (1994).
31. S. Bhattacharyya and C. S. Reddy, J. Agric. Eng. Res., 57, 157166 (1994).
32. M. A. Grompone and B. Irigaray, Personal communication.
33. B. F. Haumann, Inform, 5(11), 11981210 (1994).
34. B. F. Haumann, Aceites Grasas, 8, 8192 (1998).
35. S. Cantisan, E. Martnez-Force, and R. Garces, Plant Physiol. Biochem., 38, 377382
(2000).
36. R. H. Purdy, J. Amer. Oil Chem. Soc., 62, 523525 (1985).
37. R. H. Purdy, J. Amer. Oil Chem. Soc., 63, 10621066 (1986).
39. M. Mozzon, F. Bocci, and N. Frega, J. Agric. Food Chem., 46, 4198 4202 (1998).
40. R. Scarth, in R. Przybylski and B. E. McDonald, eds., Development and Processing of
Vegetable Oils for Human Nutrition, AOCS Press, Champaign, Illinois, 1995.
41. G. Vermeersch, Aceites Grasas, 8, 112116 (1998).
42. T. Krawczyk, Inform, 10, 555561 (1999).
43. T. Krawczyk, Aceites Grasas, 9, 525534 (1999).
44. M. I. Gurr, Lipids in Nutrition and Health, The Oily Press, Bridgwater, 1999.
45. A. Cert, Aceites Grasas, 5, 175189 (1995).
46. M. K. Gupta, Inform, 9, 11501154 (1998).
47. L. W. Kleingartner, in J. Janick and A. Whipkey, eds., Trends in New Crops and New
Uses, ASHS Press, Alexandria, Virginia, 2002.
48. L. Kleingartner and K. Warner, Cereal Foods World, 46, 399 404 (2001).
49. G. List, K. R. Steidley, and W. E. Neff, Aceites Grasas, 10, 569573 (2000).
50. S. Muller, J. Amer. Oil Chem. Soc., 60, 393394 (1983).
51. G. Mazza and D. S. Jayas, Trans. Am. Soc. Agric. Eng., 34, 534538 (1991).
52. D. C. Tandy, in P. J. Wan, ed., Introduction to Fats and Oils Technology, AOCS Press,
53. A. A. Carelli, L. M. Frizzera, P. R. Forbito, and G. H. Crapiste, J. Amer. Oil Chem. Soc.,
79, 763768 (2002).
REFERENCES 727
54. M. I. V. Brevedan, A. A. Carelli, and G. H. Crapiste, Aceites Grasas, 51, 417423 (2000).
55. A. Smiles, Y. Kakuda, and B. E. MacDonald, J. Amer. Oil Chem. Soc., 65, 11511155
(1988).
56. L. G. Pan, R. Harispe, A. Campana, M. C. Tomas, and M. C. Anon, La Alimentacion
Latinoamericana, 238, 5460 (2001).
57. S. Bloch, Aceites Grasas, 1, 2733 (1990).
58. J. Turkulov, E. Dimic, Dj. Karlovic, and V. Vuksa, J. Amer. Oil Chem. Soc., 63, 1360
1363 (1986).
59. J. Turkulov, E. Dimic, Dj. Karlovic, and V. Vuksa, Aceites Grasas, 10, 256259 (2000).
60. W. H. Morrison, J. Amer. Oil Chem. Soc., 59, 284285 (1982).
61. M. A. Baltanas, H. Molina, and C. Silva, J. Amer. Oil Chem. Soc., 75, 363370 (1998).
62. K. J. Moulton, J. Amer. Oil Chem. Soc., 65, 367368 (1988).
63. A. A. Carelli, P. R. Forbito, P. Bodnariuk, and G. H. Crapiste, Aceites Grasas, 8, 103111
(1998).
64. S. Martini and M. C. Anon, J. Amer. Oil Chem. Soc., 77, 10871092 (2000).
65. H. Liu, R. Przybylski, K. Dawson, N. A. M. Eskin, and C. G. Biliaderis, J. Amer. Oil
Chem. Soc., 73, 493497 (1996).
66. S. Bloch, Aceites Grasas, 8, 5366 (1998).
67. E. Dimic, Dj. Karlovic, and J. Turkulov, J. Amer. Oil Chem. Soc., 71, 13571361
(1994).
68. K. Kovari, J. Dense, Z. Kemeny, and K. Recseg, OCL, 7, 305308 (2000).
69. R. L. Winters, Aceites Grasas, 4, 5461 (1994).
70. M. Contini and M. Moresi, Ind. Aliment., 33, 149156 (1994).
71. F. De Panfilis, T. G. Toschi, and G. Lercker, Aceites Grasas, 7, 422429 (1997).
72. F. De Panfilis, T. G. Toschi, and G. Lercker, Inform, 9, 212221 (1998).
73. H. Topallar, Y. Bayrak, and M. Iscan, J. Amer. Oil Chem. Soc., 72, 15191522
(1995).
74. N. Ajzenberg, Aceites Grasas, 53, 229238 (2002).
75. M. Naglic, A. Smidovnik, and T. Koloini, J. Amer. Oil Chem. Soc., 75, 629633
(1998).
76. G. Rivarola, J. A. Segura, M. C. Anon, and A. Calvelo, J. Amer. Oil Chem. Soc., 64,
15371543 (1987).
77. J. A. Segura, M. L. Herrera, and M. C. Anon, J. Amer. Oil Chem. Soc., 67, 989992
(1990).
78. M. L. Herrera, J. Amer. Oil Chem. Soc., 71, 12551260 (1994).
79. M. L. Herrera, Aceites Grasas, 5, 6067 (1995).
80. M. L. Herrera and F. J. Marquez Rocha, J. Amer. Oil Chem. Soc., 73, 321325 (1996).
81. M. D. Erickson and N. Frey, Food Technol., 6367 (1994).
82. G. Ramrez, G. Hough, and A. Contarini, J. Food Qual., 24, 195204 (2001).
83. H. Topallar, Y. Bayrak, and M. Iscan, J. Amer. Oil Chem. Soc., 74, 13231327 (1997).
84. M. A. Grompone, T. Pagano, and Y. Pinchak, in Proceedings of the International
Workshop on Fats, Oils and Oilseeds Analysis (IUPAC), 2000, pp. 4661.
85. T. Pagano and M. A. Grompone, Revista Ingeniera Qumica, 15, 1120 (1999).
728 SUNFLOWER OIL
86. N. Frega, M. Mozzon, and G. Lercker, Aceites Grasas, 9, 580585 (1999).

87. A. A. Carelli, P. Bodnariuk, and G. H. Crapiste, Aceites Grasas, 10, 227231
(2000).
88. T. Albi, A. Lanzon, A. Guinda, M. Leon, and M. C. Perez-Camino, J. Agric. Food Chem.,
45, 37953798 (1997).
89. G. H. Crapiste, M. I. Brevedan, and A. A. Carelli, J. Amer. Oil Chem. Soc., 76, 1437
1443 (1999).
90. G. H. Crapiste, M. I. Brevedan, and A. A. Carelli, Aceites Grasas, 11, 6779 (2001).
91. R. Eggers, in J. W. King, and G. R. List, eds., Supercritical Fluid Technology in Oil and
Lipid Chemistry, AOCS Press, Champaign, Illinois, 1996.
92. F. Favati, J. W. King, and G. R. List, in J. W. King and G. R. List, eds., Supercritical Fluid
Technology in Oil and Lipid Chemistry, AOCS Press, Champaign, Illinois, 1996.
93. L. Calvo, M. J. Cocero, and J. M. Diez, J. Amer. Oil Chem. Soc., 71, 12511254 (1994).
95. M. A. M. Zeitoun, W. E. Neff, G. R. List, and T. L. Mounts, J. Amer. Oil Chem. Soc., 70,
467471 (1993).
96. D. Chobanov and R. Chobanova, J. Amer. Oil Chem. Soc., 54, 4750 (1977).
97. M. A. Grompone, Grasas Aceites, 42, 349355 (1991).
98. A. Rodrguez, E. Castro, M. C. Salinas, R. Lopez, and M. Miranda, J. Amer. Oil Chem.
Soc., 78, 431 436 (2001).
99. T. A. Foglia, K. Petruso, and S. H. Feairheller, J. Amer. Oil Chem. Soc., 70, 281285
(1993).
100. I. Jachmanian, M. Gil, and M. A. Grompone, Aceites Grasas, 46, 8492 (2002).
101. F. J. Sanchez-Muniz, C. Cuesta, M. C. Garrido-Polonio, and R. Arroyo, Grasas Aceites,
45, 300305 (1994).
102. A. Romero, F. J. Sanchez-Muniz, C. Tulasne, and C. Cuesta, J. Amer. Oil Chem. Soc., 72,
15131517 (1995).
103. D. B. Kupranycz, M. A. Amer, and B. E. Baker, J. Amer. Oil Chem. Soc., 63, 332337
(1986).
104. D. Barrera-Arellano, G. Marquez-Ruiz, and M. C. Dobarganes, Grasas Aceites, 48, 231
235 (1997).
105. R. Arroyo, C. Cuesta, C. Garrido-Polonio, S. Lopez-Varela, and F. J. Sanchez-Muniz,
106. C. Cuesta, F. J. Sanchez-Muniz, C. Garrido-Polonio, S. Lopez-Varela, and R. Arroyo,
107. C. Cuesta and F. J. Sanchez-Muniz, Grasas Aceites, 49, 310318 (1998).
108. M. C. Perez-Camino, G. Marquez-Ruiz, M. V. Ruiz-Mendez, and M. C. Dobarganes,
J. Food Sci., 56, 16441650 (1991).
109. M. Martn-Polvillo, G. Marquez-Ruiz, N. Jorge, M. V. Ruiz-Mendez, and M. C.
Dobarganes, Grasas Aceites, 47, 5458 (1996).
110. J. R. K. Niemela, I. Wester, and R. M. Lahtinen, Grasas Aceites, 47, 1 4 (1996).
111. J. L. Sebedio, M. C. Dobarganes, G. Marquez, I. Wester, W. W. Christie, G. Dobson,
F. Zwobada, J. M. Chardigny, Th. Mairot, and R. Lahtinen, Grasas Aceites, 47, 513
(1996).
REFERENCES 729
112. R. M. Lahtinen, I. Wester, and J. R. K. Niemela, Grasas Aceites, 47, 5962 (1996).
113. N. Jorge, G. Marquez-Ruiz, M. Martn-Polvillo, M. V. Ruiz-Mendez, and M. C.
114. N. Jorge, G. Marquez-Ruiz, M. Martn-Polvillo, M. V. Ruiz-Mendez, and M. C.
115. R. Raoux, O. Morin, and F. Mordret, Grasas Aceites, 47, 6374 (1996).
116. L. J. Van Gemert, Grasas Aceites, 47, 7580 (1996).
117. G. Dobson, W. W. Christie, and M. C. Dobarganes, Grasas Aceites, 47, 3437
(1996).
118. P. Ch. Dutta and L.-A. Appelqvist, Grasas Aceites, 47, 3847 (1996).
119. G. Marquez-Ruiz, M. Martn-Polvillo, N. Jorge, M. V. Ruiz-Mendez, and M. C.
Dobarganes, J. Amer. Oil Chem. Soc., 76, 421425 (1999).
120. S. L. Abidi and K. Warner, J. Amer. Oil Chem. Soc., 78, 763769 (2001).
121. S. Kiatsrichart, M. S. Brewer, K. R. Cadwallader, and W. E. Artz, J. Amer. Oil Chem.
Soc., 80, 479 483 (2003).
122. G. Antoln, F. V. Tinaut, Y. Briceno, V. Castano, C. Perez and A. T. Ramrez, Bioresource
Technol., 83, 111114 (2002).
123. H. Fukuda, A. Kondo and H. Noda, J. Biosci. Bioeng., 92, 405 416 (2001).
124. W. Zhou, S. K. Konar, and D. G. B. Boocock, J. Amer. Oil Chem. Soc., 80, 367371
(2003).
125. D. R. Kodali, Inform, 14, 121123 (2003).
126. S. Z. Erhan, M. O. Bagby, and H. W. Cunningham, J. Amer. Oil Chem. Soc., 69, 251256
(1992).
127. X. Pages and C. Alfos, OCL, 8, 122125 (2001).
128. A. J. M. Rodrguez, Aceites Grasas, 8, 187193 (1998).
129. F. Gunstone, Inform, 13, 159163 (2002).
130. F. Gunstone, Inform, 13, 810812 (2002).
131. F. Gunstone, Inform, 14, 42 43 (2003).
133. J. Fry and R. Stringfellow, Aceites Grasas, 9, 43 47 (1999).
134. M. A. Grompone and P. Moyna, J. Amer. Oil Chem. Soc., 60, 13311332 (1983).
135. M. A. Grompone, Revista Ingeniera Qumica, 3, 1418 (1991).
136. W. H. Morrison, J. Amer. Oil Chem. Soc., 58, 902903 (1981).
137. G. Gil, I. G. de Kuck, and R. A. Macchi, Aceites Grasas, 9, 443 446 (1999).
138. G. I. Dashiell, Aceites Grasas, 12, 197204 (2001).
139. G. W. Chapman, J. Amer. Oil Chem. Soc., 57, 299302 (1980).
140. J. P. Cherry and W. H. Kramer, in B. F. Szuhaj, ed., LecithinsSources, Manufacture and
Uses, AOCS Press, Champaign, Illinois, 1989.
141. A. A. Carelli, M. I. V. Brevedan, and G. H. Crapiste, J. Amer. Oil Chem. Soc., 74, 511
514 (1997).
142. J. Hollo, J. Peredi, A. Ruzics, M. Jeranek, and A. Erdelyi J. Amer. Oil Chem. Soc., 70,
9971001 (1993).
143. P. A. Piras, Aceites Grasas, 8, 205206 (1998).
730 SUNFLOWER OIL
144. S. P. Clark, P. J. Wan, and S. W. Matlock, J. Amer. Oil Chem. Soc., 57, 275A279A
(1980).
145. Z. Liebowitz and C. Ruckenstein, Aceites Grasas, 3, 4044 (1993).
146. M. A. Pickard, T. J. Jones, and R. T. Tyler, Aceites Grasas, 8, 233250 (1998).
147. C. Ordonez, M. G. Asenjo, C. Benitez and J. L. Gonzalez, Biores. Technol., 78, 187190
(2001).
148. L. Walsh, R. I. Winters, and R. G. Gonzalez, Inform, 9, 7883 (1998).
BAILEYS INDUSTRIAL
OIL AND FAT
PRODUCTS
Sixth Edition
Volume 3
Specialty Oils and Oil Products
Edited by
Fereidoon Shahidi


1
Conjugated Linoleic
Acid Oils
Rakesh Kapoor, Martin Reaney, and Neil D. Westcott
Bioriginal Food and Science Corp.
1. INTRODUCTION
The discovery of conjugated linoleic acid (CLA) dates back to 1933, when it was
found that treatment of polyunsaturated fatty acids with alkali increased the UV
absorbency (13). It was later found that the treatment produced a one to one
mixture of cis-9-, trans-11- (9,11-ct) and trans-10-, cis-12 (10,12-tc) CLA. In
1935, it was noted that UV absorbency at 230 nm of milkfats was higher in milk
from cows fed polyunsaturated fatty acids than cows fed saturated fats (4). This
phenomenon was shown to be a result of conjugation of the double bonds of the
polyunsaturated fatty acids (5). The predominant isomer in dairy products (milk,
cheese, butter) or meat is 9,11-ct CLA (6).
Since the time of their discovery, conjugated fatty acids have been the subject of
intense investigation. Spectroscopic analysis using ultraviolet light was the major
analytical instrument available to researchers in the 1930s. As a result of high absor-
bance of UV light at 230 nm or higher, conjugated fat became a useful research tool
for the study of fat metabolism. The first animal study used naturally conjugated
fats of tung oil, but it was poorly tolerated (7). During this period, essential fatty
1
2 CONJUGATED LINOLEIC ACID OILS
acids were discovered. Aaes-Jorgensen (8) studied the possibility of using CLA to
treat/prevent the symptoms of essential fatty acid-deficiency. They observed that in
essential fatty acid-deficient animals, CLA could not prevent deficiency but showed
toxicity. Synthetic conjugated fatty acids produced from linoleic acid (LA), usually
mixtures of 9,11-ct and 10,12-tc CLA, replaced the natural substances as a pre-
ferred biological marker.
Over the last two decades, the conjugated fats and, particularly, conjugated linoleic
acid (CLA) have been intensively studied for their biological activity. As a result
of the ease of synthesis, blends of two CLA isomers, namely 9,11-ct and 10,12-tc-
CLA, have been the focus of most research into biological activity. Recent research,
however, has been expanding to include pure or enriched isomer preparations.
2. METABOLISM
Early studies using a CLA mixture revealed that animals absorb and incorporate
some of the CLA in their tissues in phospholipid, glycolipid, and acylglycerol
fractions. In 1950, Reiser (9) observed faster and better absorption and incorpora-
tion of CLA when administered as triacylglycerol (TAG) versus free fatty acid.
Following administration of CLA as TAG, the maximal levels appear in blood,
liver, and organs at 16 hours compared with 24 hours for the free fatty acid
form. Incorporation was higher in mesentery fat followed by perirenal and subcu-
taneous fat (9). Barnes et al. (10), in 1941, reported the kinetics of CLA absorption
from the mixture. The absorption rate was highest in the first hour following admin-
istration and then gradually declined. Following a single oral dose, over 50% of
conjugated CLA was absorbed in neutral mucosal lipids in the first hour followed
by a decline, whereas in mucosal phospholipids, the incorporation was much
slower, reaching maximum at about 8 hours followed by decline. In 1951, it was
found that poultry incorporate CLA into egg lipids (5). Recent studies confirm
the earlier finding of preferential incorporation in neutral lipids followed by phos-
pholipids in liver (11) and mammary tissues (11, 12). Incorporation of CLA in tis-
sues is associated with a reduction in the amounts of arachidonic acid and linoleic
acid in neutral lipids (11) and in the liver TAG levels with an increase in the levels
of 18:0 (13). In pig heart lipids, 11,13-ct isomer was the major CLA isomer
followed by 9,11-ct isomer following an oral administration of CLA mixture
(14). CLA isomers compete with desaturases and elongases to produce desaturated
and elongated products maintaining the geometry of double bonds (1517). CLA
has been shown to inhibit delta-9-desaturase enzyme in vitro (18) and in vivo
(19). CLA feeding also resulted in a decrease in arachidonic acid, and also
reduced the desaturation of linoleic acid without affecting the desaturation of
alpha-linolenic acid to any significant level. In rat, 9,11-ct isomer was preferen-
tially metabolized to conjugated C20:3, whereas 10,12-tc isomer was metabolized
to conjugated C16:2 and 18:3 compounds (19). The CLA mixture and its indivi-
dual isomers (9,11-ct and 10,12-tc) inhibited basal and calcium ionophore stimu-
lated production of prostaglandin from human sapheneous vein endothelial cells in
PHYSIOLOGICAL ACTIONS OF CLA 3
a dose-dependent manner. The mixture of CLA isomers was reported to inhibit the pro-
duction of eicosanoids at all doses, whereas 10,12-tc isomer was shown to inhibit pro-
duction at lower dose but stimulate at higher dose (20).
3. PHYSIOLOGICAL ACTIONS OF CLA
3.1. Effect on Body Composition
While working with essential fatty acid deficient rats in 1951, Holman observed
that CLA-fed rats had significantly less total fat than control rats, and that they
lost weight (21). Subsequent studies demonstrated that CLA inhibited fat accumu-
lation and promoted lean muscle mass in growing animals, including pigs and mice
(2224). The results on the effect of CLA on body fat composition in animals are
unequivocal, whereas studies in humans are providing mixed results. In a double
blind, randomized clinical trial on obese and overweight humans, Blankson et al.
(25) observed a clinically significant reduction in body fat mass in groups adminis-
tered various doses of CLA ranging from 1.7 g per day to 6.8 g per day. Reduction
in body fat mass was significant for groups administered 3.4 g CLA per day and 6.8 g
CLA per day. Interestingly, this study found that 3.4 g of CLA per day provided
maximum reduction in body fat mass; increasing the dose above this level provided
no additional effect (25). Lean body mass and body mass index were similar in all
the groups, although there was a slight increase in lean body mass in the CLA
group. The increase in lean body mass did not achieve significance compared
with a placebo group administered olive oil. Additionally, the CLA group presented
a significant reduction in total-, LDL-, and HDL-cholesterol (25). In another study,
Riserus et al. (26) observed a reduction in saggital abdominal diameter in abdom-
inally obese humans without affecting total body weight. A reduction in total fat
mass in healthy, nonobese, exercising males was observed when CLA was given
at a total daily dose of 1.8 g (divided in 3 doses) for 12 weeks. Body weight was
not affected in this double blind clinical trial (27). A recent study in type II diabetic
patients who were not on any medication found an inverse relationship between
plasma CLA levels and weight loss and serum leptin levels (28). The inverse rela-
tionship was significant for 10,12-tc isomer of CLA and not for 9,11-ct isomer. A
study in nonobese individuals using a mixture of CLA isomers containing about
20% each of 9,11-ct and 10,12-tc-CLA isomers, along with 20% to 25% other iso-
mers, did not observe a reduction in body fat mass (29). Another study investigated
the effect of CLA on weight regain after weight loss in overweight subjects (30).
This study observed no effect of CLA on weight gain after weight loss; however,
the weight gain in the CLA group was a result of an increase in fat free mass and
was independent of dose (30). Comparison between studies is difficult as these stu-
dies differed in the degree of obesity of the subjects, duration of treatment, dose,
and the isomer composition of the CLA preparation. Earlier commercial products
of CLA contained equal amounts of 9,11-ct and 10,12 tc isomers with other iso-
mers in small amounts. The other isomers include all trans-isomers as well as other
cis-, trans-isomers, including 11,13-ct, 11,13-tc, 8,10 ct,8,10-tc, etc. This illustrates
the need for research on specific isomers and standardized protocols.
These reported actions of CLA could be mediated through a number of physio-
logical mechanisms including increased fat oxidation (31) or inhibition of lipid
accumulation in fat cells. Recent studies have started to investigate the physiological
actions of individual isomers. It appears that the 10,12-tc isomer of CLA is mainly
responsible for the effect of CLA on adiposity (3234).
It was demonstrated that 10,12-tc-CLA reduces leptin, a hormone involved in
regulation of fat deposition, in cultured fat cells (35), and in mice (36). In the lat-
ter study, feeding 10,12-tc-CLA to mice caused a comparatively small gain in
weight with no gain in adipose fat. The 10,12-tc isomer of CLA was also shown
to inhibit differentiation of preadipocytes in murine (3T3-L1) (37) and human
preadipocytes (38). This was associated with decreased accumulation of TAGs
in differentiating preadipocytes; inhibition of peroxisome proliferator-activated
receptor g (PPAR-g) gene (38) and its downstream gene products including lipo-
protein lipase (LPL), GLUT-4 (glucose transporter gene 4), and inhibited expres-
sion of fatty acid synthase (FAS) gene. CLA isomer 9,11-ct, on the other hand,
increased accumulation of TAGs in adiposities and also stimulated GLUT-4 and
LPL. This suggests that isomer 10,12-tc may be responsible for inhibition of glu-
cose uptake and oxidation in the adipocytes, leading to decreased TAG accumula-
tion. These actions also underlie the effect of 10,12-tc isomer in inducing insulin
resistance leading to lipoatrophic diabetes observed in animal (39, 40) and human
studies (41).
3.2. Anticancer Properties

In 1985, Pariza and Hargrave discovered an antimutagenic fraction in cooked and
raw beef during their studies on identification of carcinogenic compounds present
in cooked beef (42). This fraction was identified to be a mixture of 4 isomers of
CLA (9,11-ct, 9,11-tt, 10,12-tc, and 10,12-tt) (43). Studies in animal models and
cell lines demonstrated the antimutagenic activity of a CLA mixture against known
chemical carcinogens (7,12-dimethylbenz[a]anthracene, DMBA, and benzo (a) pyr-
ene) (4347). CLA has been shown to have anticancer effects against breast, colon,
and prostate cancer cell lines (4850). In rat models of breast cancer, CLA was
reported to affect the breast structure when given during development stages. In
this study, dietary CLA reduced the proliferation of terminal end bud and lobuloal-
veolar bud structures, whereby breast tissue became resistant to neoplastic transfor-
mations associated with cancer at later stages in life (4850).
The exact mechanism of anticancer effects of CLA is not clear, and several pos-
sible mechanisms could underlie the anticancer properties of CLA. These actions
may include its ability to interfere with the proliferation of cancer cells, increased
apoptotic cell death, inhibition of angiogenesis, or increased oxidative stress. In a
study comparing the effects of CLA on estrogen receptor positive human breast
cancer cells (MCF-7) and estrogen receptor negative (MDA-MB 231) cells, CLA
was shown to selectively inhibit proliferation of estrogen receptor positive cells.
CLA-treated MCF-7 cells selectively remained in G0/G1 phase and the expression
of c-myc was inhibited. CLA had no effect on the growth of MDA-MB 231 cells.
This study suggests that CLA acts by interfering with the estrogen-mediated second
messenger system (51). CLA is also known to interfere with eicosanoids pathway
and inhibits production of prostaglandin E2 (PGE2) (50, 52). Reduced production of
PGE2 may play a role in anticancer actions of CLA. CLA is also shown to stimulate
apoptotic death of cancer cells (50, 53, 54). CLA isomers increased apoptosis
by stimulating the expression of caspase 3 and 9 activities and by reducing the
expression of Bcl-2, an apoptosis repressor gene. The 10,12-tc isomer of CLA
was found to be more potent in mediating these actions than either the 9,11-ct iso-
mer or a mixture of the two (54). The other possible mechanism for anticancer
properties of CLA is its ability to inhibit angiogenesis. In a mouse model of breast
cancer, both isomers inhibited angiogenesis in mammary fat pad and reduced
the concentration of vascular endothelium-derived growth factor (55). The CLA
isomer 10, 12-tc also inhibited secretion of leptin and induced apoptosis in white
and brown adipocytes, whereas 9,11-ct isomer was without effect on these para-
meters. CLA was also shown to reduce cell proliferation by reducing the expression
of proteins involved in cell cycle regulation (p16 and p27) and DNA synthesis (56).
The above discussion focused on the role of CLA as a chemoprotective agent.
Information regarding its effect on cancer treatment is limited. Feeding CLA for
four or eight weeks after carcinogen administration was reported to be ineffective in
preventing tumor formation, whereas continuous administration protected against
tumor development (12, 57). A recent case control study in Finnish women suggested
that dietary CLA may be protective against breast cancer (58). The role of CLA in
prevention or treatment of cancer in humans is not clear and requires more research.
3.3. Insulin Resistance and Diabetes

CLA has been shown to normalize impaired glucose tolerance and improve hyper-
insulinemia in prediabetic ZDF rats (59). These actions appeared to be mediated
through PPAR-g pathway as CLA treatment induced expression of mRNA for
aP2. Recently, it has been observed in Zucker diabetic rats that either a 50:50 mix-
ture of 9,11-ct-CLA and 10,12-tc-CLA isomers or a 90% 9,11-ct-CLA isomer sti-
mulated insulin action in fat and muscle cells (60). These observations suggest that
CLA can prevent or delay the onset of diabetes and that the 9,11-ct isomer may
contribute much of this activity. Recently, 10,12-tc isomer of CLA was shown to
induce insulin resistance and hyperinsulinemia in mice, whereas 9,11-ct isomer
had no effect (61). Both these isomers were shown to be equally efficient in stimu-
lating PPAR-a and g receptors, indicating that the hyperinsulinemia may not be
mediated through nuclear receptor pathway. In another study using either high
metabolic rate mice or low metabolic rate mice, Hargrave et al. (62) demonstrated
that CLA increased the insulin resistance in high metabolic rate mice only, whereas
in Zucker diabetic rat, CLA was shown to prevent a rise in insulin and glucose
levels that might have been mediated through an increased production of adiponec-
tin, a hormone released by adipose tissues (63). In pigs, dietary CLA had no effect
on plasma glucose or insulin levels or on the ability of insulin to mobilize plasma

glucose (64). These studies indicate that the effect of CLA in diabetes is not clear
and the reported differences may be species specific. In human subjects, the effects
of CLA on insulin resistance and glucose homeostasis are not well studied. In one
clinical trial, the 10,12-tc isomer, but not a mixture of 9,11-ct and 10,12-tc isomers,
was shown to increase the insulin resistance and blood glucose levels in abdomin-
ally obese people (41), whereas another study on normolipidemic subjects failed to
observe any effect of either a 50:50 or 80:20 mixture of 9,11-ct and 10,12-tc iso-
mers on blood glucose or insulin levels (65). Belury et al. (28) observed a reduction
in fasting plasma glucose levels in type II diabetics when they were treated with
6.0 g of a mixture of CLA for eight weeks. This necessitates the need for controlled
studies in human to delineate the effect of CLA and its isomers on blood glucose
homeostasis and insulin resistance.
3.4. Cardiovascular Actions

An isomeric mixture of CLA was reported to inhibit the development of athero-
sclerosis in rabbits (66, 67) or hamsters (68) fed a high cholesterol diet. The CLA mix-
ture caused a regression of atherosclerosis in rabbits (67) and Apo E-/- mouse (69).
In hamsters, CLA was shown to lower the levels of total- and LDL-cholesterol and
TAG levels (33, 70). The reduction in plasma levels of cholesterol could be
mediated via an increase in LDL receptor expression in the liver leading to
increased clearance from the circulation (71). CLA has also been reported to reduce
secretion of apolipoprotein B in animals (72). The effect of CLA could be mediated
through induction of the PPAR-g pathway. CLA has been shown to inhibit cyclo-
oxygenase enzyme in vitro (69). Its anti-inflammatory actions might be playing a
role in prevention of atherosclerosis but does not appear to play a role in regression
of atherosclerosis in animals. The 9,11-ct, and 10,12-tc isomers of CLA and the
metabolite of 9,11-ct isomer (13-hydroxy-9c,11t- octadecadienoic acid) have
been reported to inhibit arachidonic acid and collagen-induced platelet aggregation
(73), which could have been mediated through inhibition of thromboxane A2 for-
mation from arachidonic acid. CLA was also shown to prevent the development of
hypertension in Zucker diabetic rats (63) that was associated with increased expres-
sion of mRNA for adiponectin, a hormone released by adipose tissues. Similar
results on prevention of hypertension were shown in Otsuka Long-Evans Tokush-
ima fatty (OLETF) rats (74). In this rat model, the 10,12-tc isomer of CLA was
shown to inhibit angiotensinogen production from adipose tissues. The effect of
CLA on plasma cholesterol levels and atherogenic potential in human is not clear.
In healthy, normocholesterolemic human subjects, Benito et al. (75) reported no
effect of CLA on cholesterol levels, platelet aggregation, or bleeding time, whereas
Noone et al. (65) reported a TAG lowering effect of 50:50 mixture of CLA isomers.
These differences could be due to differences in the composition of CLA, as Noone
et al. (65) used 50 : 50 or 80 : 20 mixtures of 9,11-ct and 10,12-tc isomers and
Benito et al.s (75) preparation contained 11.4% 9,11-ct, 10.8% 8,10-tc, 15.3% 11,13-ct,
and 14.7% 10,12-tc, with 6.7% c,c and 5.9% tt isomers of CLA. To establish the effects
of CLA on cardiovascular risk factors in humans, more research is needed using

pure isomers or standardized mixture of CLA isomers.
3.5. CLA and Immune Function

In animal studies, CLA was reported to enhance immune response and attenuate
allergic reactions (76, 77). CLA has also been shown to prevent age associated
reductions in immune function (78). In Guinea pigs, feeding CLA was shown to
reduce the release of histamine and PGE2 from isolated trachea challenged with
antigens (79), suggesting a strong antiallergic action. In mice, feeding CLA dose
dependently increased splenic lymphocyte proliferation in response to phytohemag-
glutinin but not to lipopolyssaccharide or conclavin A, suggesting that CLA has
selective actions on immune function. CLA stimulated IL-2 production (80) and
also increased the basal and mitogen stimulated natural killer cell activity of splenic
lymphocytes in mice that was associated with increased number of NK-cells but no
change in the ratio of NK-cells to total splenic lymphocytes (81). CLA has also
been shown to reduce the release of proinflammatory mediators, including PGE2,
IL-1, and IL-6, from macrophages stimulated by interferon gamma (IFN-g) (82).
These actions appeared to be mediated through stimulation of PPAR-g pathway.
When fed to pigs, CLA enhanced cellular immunity by modulating white blood
cell types that control adaptive and innate immunity (83, 84). Feeding CLA to preg-
nant and lactating pigs caused an increase in IgG levels of colostrums without
affecting the serum IgG levels (85). In the same study, feeding CLA to suckling
piglets resulted in increased serum levels of IgG and lysosomes. These results indi-
cate enhancement of immune function in piglets. Following the feeding of CLA to
young healthy women, no effects were observed on several measures of immune
function including delayed type hypersensitivity response and numbers of circulat-
ing white cells and antibodies to vaccine, although an increase in CLA content of
mononuclear cells was observed (86, 87). Differences in immune responses may be
due to the selection of a young healthy population with optimal immune function,
species differences, or dietary CLA isomer composition.
CLA has recently been studied for its actions on peroxisome proliferator-
activated receptors (PPARs), most notably of the PPARg. As PPARg plays a role
in macrophage activity, Yu et al. (82) observed a stimulation of PPARg in
RAW264.7 mouse macrophage (RAW) cells by various CLA isomers. CLA also
decreased the production of PGE2, TNF-a, nitric oxide (NO), IL-1b, and IL-6 in
RAW cells treated with interferon-g (IFNg) (82). The inhibition of production of
these inflammatory mediators was associated with a reduced expression of
mRNA for cyclo-oxygenase 2 (COX2), inducible NOS (iNOS), and tumor necrosis
factor alpha (TNF-a). Cheng et al. (88) observed similar inhibition of COX-2 and
NOS mRNA in lipopolysaccharide (LPS) stimulated macrophage cell, which was
associated with an inhibition of LPS-induced protein expression of the cytoplasmic
phosphorylated inhibitor kappaBalpha (IkBa) and nuclear p65 as well as NF-
kappaB nuclear protein-DNA binding affinity. This observation indicate a role of
NF-kappaB in regulation of anti-inflamaotry actions of CLA.
4. STRATEGIES TO INCREASE DIETARY INTAKE OF CLA
As the interest in beneficial effects of CLA is increasing, so are the efforts to

increase its dietary intake. Various strategies are being employed, which include
increasing the content of CLA in eggs, milk, and meat. In 1951, hens were shown
to incorporate CLA into egg lipids following dietary administration (5), but the egg
production was significantly reduced. Recently, strategies to reduce the population
of problem birds, based on feeding CLA to the females to reduce the egg hatchabil-
ity, have been patented (89). Feeding 0.5% CLA in the diet of hens was shown to
increase the CLA content of egg lipids, which was associated with a significant
increase in saturated fatty acids and a reduction in monounsaturated fatty acids
in the egg lipids. These changes in the egg composition also changed the properties
of the egg yolk in that it became hard when stored at cool temperatures (refrigera-
tor). This observation suggests that feeding CLA to poultry may not be an attractive
strategy to increase the CLA content of eggs as poultry breeders cannot bear the
economic burden of significantly reduced fertility. Strategies are needed to
increase the content of CLA without affecting hatchability of eggs. Recently,
methods to increase the content of CLA in eggs have been patented (90, 91). These
strategies include incorporating CLA along with monounsaturated (91) or mono-
and polyunsaturated fatty acids (90) in the diet. Similarly, feeding CLA in a
diet to cattle increased CLA content in milk and meat and simultaneously reduced
total milkfat content (92) and increased milkfat saturated fatty acids. Other
strategies include feeding 11-trans-octadecaenoic (C18:1) acid (93) and other
sources of polyunsaturated fatty acids to cattle. Feeding either fresh forage or
supplementing the cattle diet with high linoleic acid meals/oils is another effective
way of increasing the content of CLA in milkfat (9497) and muscle. Most
common sources of polyunsaturated fatty acids include sunflower oil, linseed
(flaxseed oil), safflower oil, fish oil, or marine algae (98101). When CLA is fed
to cattle, it is better to protect it from rumen hydrogenation by converting to
calcium salts.
5. COMMERCIAL PRODUCTION OF CLA
Industrial conjugated linoleic acid (CLA) is a poorly defined blend of compounds

(102). Early commercial syntheses focused on maximizing total CLA content.
Many early products were rich in CLA but contained a number of positional iso-
mers. Market demand has now shifted for a product that contains two predominant
isomers, specifically 9,11-c,t-octadecadienoic acid and 10,12-tc-octadecadienoic
acid. It is not surprising that alkali isomerization produced some undesirable posi-
tional isomers of CLA. In 1970, Mounts and Dutton (103) had shown unequivocally
that when potassium t-butoxide was used, at least four positional isomers of CLA
were produced. It was not until 1997, after the use of CLA as a dietary supplement
COMMERCIAL PRODUCTION OF CLA 9
TABLE 1. Linoleic and Linolenic Acid Contents of Some Vegetable Oils (104).
% Linoleic % Linolenic Commercial Oil Commercial

Oil Source Acid Acid Available Fatty Acid Available
Corn 57 0 Yes
Cottonseed 53 0 Yes
Cucumber 72 0 No
Grapeseed 70 0 Yes
LinolaTMFlaxseed 72 3 Yes
Poppy 77 0 Yes
Safflower 75 0 Yes
Sunflower 64 0 Yes Yes
Soybean 51 8 Yes Yes
Squash (pumpkin) 60 0 Yes
Walnut 62 12 Yes
began, that Christie et al. (102) elegantly demonstrated that commercial CLA was a
blend of positional isomers. In response to this discovery, new commercial CLA
products have been introduced that have comparatively high levels of the preferred
isomers. In spite of the improvements, all current available commercial CLA
products contain some level of the less desirable isomers and other components,
which may or may not be desirable.
Commercial processes for the synthesis of any compound of economic value is
normally proprietary information and the commercial methods of CLA production
are no exception. The process by which each brand of commercial CLA is synthe-
sized is not known by the authors of this review. Therefore, this review is directed
at the patent literature on CLA synthesis, major problems encountered in CLA
synthesis, and analysis of CLA from commercial suppliers.
5.1. CLA Production Raw Materials

The raw material for CLA production must be a material that is rich in linoleic
acid. This product could be in the triacylglycerol form, fatty acids or fatty acid
esters. The concentration of CLA in the final product is directly dependent on
the level of linoleic acid in the starting material. The highest level of linoleic
acid available from botanical sources is not available in commercial products.
Extraction and refining equipment would be required to obtain oils with the highest
linoleic acid levels. Table 1 lists the commercial and noncommercial sources of oil
and fatty acids that are known to be rich in linoleic acid and their availability as
TAGs and fatty acids.
If commercial CLA is to be synthesized from a fatty acid, it must be recognized
that commercial fatty acids are generally not intended for use in the production of
CLA. Commercial fatty acids are usually produced by the reaction of water (steam)
and TAG oil at high temperatures in a continuous reaction (Reaction 1).
Reaction 1: Alkali or acid hydrolysis of acylglycerols.
O
O R
O
R O
O
R
O
H+ or
OH
H2O
OH O OH O
HO + 3 HO + 3
OH R OH OH R O
This reaction is accelerated through the use of a solid phase acid catalyst, which
is readily separated from the fatty acid and glycerol products after hydrolysis (105).
The disadvantage of this hydrolysis process is that the reaction is reversible and
products generated by this process contain appreciable amounts of mono- and dia-
cylglycerols, which may have undesirable side reactions in CLA synthesis. Fatty
acids may also be produced by the hydrolysis of TAGs in a pressurized reactor
at 200 C without the addition of a catalyst (105). This reaction may be catalyzed
at lower temperatures using zinc oxide in a batch reactor (105). The product of these
batch reactions also contains substantial amounts of mono- and diacylglycerols.
Hydrolysis of TAGs is possible using water and strong base to produce soaps
(Reaction 1). This reaction proceeds to completion and can be conducted at the
modest temperatures required to maintain the reaction mixture as a fluid. More
than three moles of potassium hydroxide or sodium hydroxide are required to
hydrolyze one mole of TAG oil. As the caustic alkali cannot catalyze the reverse
reaction, this process can produce soaps that are virtually free of acylglycerols in
a single step (105). The soaps from alkali hydrolysis of TAGs are readily converted
to fatty acids by acidification with the addition of citric acid or strong mineral acids,
which include HCl, H2SO4, or H3PO4. Regardless of the method chosen for produc-
tion of fatty acids, the acids should be dried under vacuum after washing with brine
or by a combination of other acceptable methods (105).
There are commercially available fatty acids suitable for use in CLA production.
For example, Henkel Corporation (106) sells a series of fat products including those
shown in Table 2. However, none of the products listed in Table 2 would be pre-
ferred as starting materials for CLA production for reasons that will be discussed.
5.2. Enrichment of Linoleic Acid

A commercial interest may wish to produce CLA at concentrations greater than can
be obtained by modifying high linoleic acid plant oils. Several methods exist that
will improve the starting material by increasing the concentration of linoleic acid,
TABLE 2. Henkel Products (106).
Product 14:0 (%) 16:0 (%) 18:0 (%) 16:1 (%) 18:1 (%) 18:2 (%) 18:3 (%)
Soya fatty acids 0.5 16 4 1 25.5 48 5

(Emersol 610)
Linoleic acid 0.5 3.5 0.5 Trace 19.5 65.5 10.5
(Emersol 315)
Methyl Linoleate 0.5 3.5 0.5 Trace 19.5 65.5 10.5
(Emery 2221)
but only a few methods are used in industrial settings. Industrial separation of fatty
acids has been reviewed by others (104, 107). A limited discussion of these
methods will be presented.
Crystallization is used to separate saturated fats and oleic acid from linoleic acid.
If a highly concentrated product is required, the linoleic acid may be crystallized
once or repeatedly as the last step in purification. Crystallization is a mild procedure
but usually requires the use of a solvent (108) such as acetone or methanol. The use
of low boiling point and flammable solvents raises concerns over plant safety, gov-
ernment regulations on manufacturing, and market acceptance of the product.
Furthermore, the removal of oleic acid by crystallization in solvent is only possible
by lowering the temperature of the liquor to below 40 C (108). To crystallize
linoleic acid, the temperature must be reduced to 75 C.
Dry or solvent free crystallization is also possible; but these methods often
require the addition of crystal modifiers that become incorporated into the product
(108). Losses during crystallization can be very high as the crystals entrain large
amounts of fatty acid. However, these losses may be reduced by physically pressing
the crystals to remove the entrained solution (109). Linoleic acid-rich products of
dry crystallization would be preferred starting materials for CLA production over
those of solvent crystallized products; but the losses incurred in dry crystallization
may prohibit this method of manufacture. Crystal modifiers may be selected so that
they do not adversely affect the quality or acceptance of the final product.
Specific fatty acids may be concentrated by sequentially removing contaminat-
ing fatty acids as urea adducts and forming the urea adduct of the desired fatty acid.
This process requires dissolving the fatty acids or esters in urea and hot methanol
(or other alcohol) and cooling to effect adduct formation. The adduct is filtered
from the liquor and, if conditions are carefully controlled, the adducts can be
used to sequentially crystallize saturates, monounsaturates, diunsaturates, and triun-
saturates. A urea adduct rich in linoleic acid could be produced by first removing
adducts of saturates and monounsaturates from a suitable oil and then forming the
desired adduct. Once formed, the adduct may then be decomposed by the addition
of water to the solid phase. Enriched linoleic acid could be recovered by solvent
extraction of the urea : water solution with a nonpolar solvent such as hexane.
All problems associated with crystallization in solvent mentioned previously also
occur in formation of urea adducts, with the exception of the requirement for
very low temperatures. Typically, urea adducts form between room temperature and
0oC (108).
Fatty acids may also be enriched by the use of various absorption media. Mole-
cular sieves can separate saturated fatty acids from unsaturated fatty acids dissolved
in acetone (110). Oleic acid and linoleic acid dissolved in blends of solvents,
including acetonitrile, tetrahydrofuran, water, and formamide, may be separated
using cross-linked polystyrene polymers such as AmberliteTM XAD-2 or XAD-4.
Selective extraction methods using two-phase solvent systems may also be used
to enrich fatty acids. Solvent systems such as dimethyl formamide, hexane, and
ethylene glycol can form a two-phase system that effectively partitions sunflower
oil TAGs rich in linoleic acid from those depleted in linoleic acid (111). TAGs
partitioned in this way may contain up to 84.7% linoleic acid. This method would
not likely be used in industry because the magnitude of the losses are usually
unacceptable.
5.3. Approaches to CLA Production

CLA has been produced by the reaction of soaps with strong alkali bases in alcohol,
ethylene glycol, and glycerol (112114) (Reaction 2).
Reaction 2: Isomerization of cis,cis-9,12-octadecadienoyl soaps.
COO
base
COO
COO
The CLA product is generated by acidification of the soap solution with a strong
acid (sulfuric or hydrochloric acid) and repeatedly washing the product with brine
or an aqueous CaCl2 solution.
CLA has been synthesized from fatty acid and soap blends using SO2 in the
presence of a substoichiometric amount of soap forming base (115). This reaction
produced predominantly the all trans-configuration of CLA.
Of these methods, alkali isomerization of soaps is the least expensive process for
bulk preparation of CLA isomers; however, the use of either monohydric or poly-
hydric alcohols in alkali isomerization of CLA can be problematic. Lower alcohols
are readily removed from the CLA product, but they require that the production
facility be constructed to support the use of flammable solvents. Higher molecular
weight alcohols and polyhydric alcohols are considerably more difficult to remove
from the product and residual levels of these alcohols (e.g., ethylene glycol) may
not be acceptable in the CLA product.
Water may be substituted for the alcohols in the production of CLA by alkali
isomerization of soaps (116, 117). When water is used in this reaction, it is neces-
sary to perform the reaction in a pressure vessel, whether in a batch (116) or con-
tinuous mode of operation (117). The process for synthesis of CLA from soaps
dissolved in water still requires a complex series of reaction steps. Bradley and
Richardson (118) were able to produce CLA directly from TAGs by mixing sodium
hydroxide, water, and oil in a pressure vessel. Their method eliminated the need to
synthesize fatty acids followed by soap formation prior to the isomerization reac-
tion. However, the authors reported that they were able to produce an oil with only
40% CLA. Quantitative conversion of the linoleic acid in soybean oil to CLA would
have produced a fatty acid mixture with approximately 51% CLA.
5.4. Reaction Kinetics and Production of Positional Isomers

The kinetics of conversion of linoleyl soaps to conjugated linoleyl soaps has been
well described by first-order reaction kinetics (119). Total conjugation is readily
measured by simple methods such as increases of UV absorbance at 231.5 nm. In
industry, it is necessary to allow the reaction to proceed until most of the linolyl
soaps are conjugated but desirable to stop the reaction as soon as possible after
that point. The reaction is between 97% and 98.5% complete after 5 half-lives
and 6 half-lives respectively. There is little advantage in continuing the reaction
longer than this time and, as will be discussed below, undesired reactions may occur
with longer reaction times. The reaction constant is readily determined early in the
reaction when changes in the level of conjugation are large. The task of determining
the rate constant in a large reactor is complicated by the mass of the reactor and its
contents. A large batch reactor often requires several hours to reach the optimum
heat of reaction, at which time the reaction may be almost complete. Similarly,
cooling the contents of a large reactor as a means of stopping a reaction is usually
impractical. A reaction that nears 99% completion in 2 hours in a laboratory has a
half life of less than 20 minutes. Control of a batch reaction in a commercial opera-
tion may use analytical data from periodic sampling, but the analytical method must
be very rapid to be an effective tool for decision making.
The authors have found that the half-life of the reaction may vary by approxi-
mately 10% per 1 C. It follows that precise reactor temperature control is essential
to standardize quality control. A reaction planned to continue for 6 half-lives could

vary from 5 half-lives to 7 half-lives if the reactor temperature control is 2 C.
The isomerization reaction that leads to the production of positional isomers
(Reaction 3) has similar first-order kinetics. Using this assumption, we have mod-
eled the sequential conversion of linoleyl soaps to a mixture of 9,11-cis-, trans-
octadecadienoyl and 10,12-trans-, cis-octadecadienoyl soaps (Reaction 2) and final-
ly, the conversion of these two soaps to 8,10-octadecadienoyl and 11,13-octadeca-
dienoyl soaps. Reaction 3 shows the mechanism of one of these two isomerization
processes.
Reaction 3: Isomerization of cis-,trans-9,11-octadecadienoyl soap to trans-,cis-
8,10-octadecadienoyl soap via sigmatropic rearrangement.
10
9 11
8
R H R1 R H R1 R H R1
R = (CH2)6COOH
R1 = (CH2)5CH3
In an earlier review of commercial CLA production, Reaney et al. (120) devel-

oped a model of the base catalyzed sequential conversion of unconjugated linoleic
acid to a mixture of two isomers, 9,11-c,t- and 10,12-t,c-CLA, by first-order
kinetics. After the formation of the two primary isomers, two additional isomers,
8,10-t,c-CLA and 11,13-c,t-CLA, were produced by a sequential reaction. It was
later reported by Saebo (121) that the second reaction producing the additional
isomers was actually a thermal sigmatropic rearrangement as shown in reaction 3.
The intramolecular rearrangement is independent of catalyst concentration and,
thus, the accumulation of additional isomers is a function of the reaction tempera-
ture and the duration of reaction. Saebo (121) reports that prolonged heating results
in the accumulation of isomers but that reaction solvent that allows for low-
temperature reactions may be used to prevent sigmatropic rearrangement and the
consequent formation of isomers.
5.5. Solvents Used in Production of CLA by Alkali Catalysts

Research reports describe the use of at least eight solvent systems for the produc-
tion of CLA using alkali catalysts (112116) (Table 3). The choice of solvent
greatly affects the reaction conditions of CLA production. The choice of solvent
by the manufacturer is determined by a number of considerations. Many markets
will not accept low levels of ethylene glycol, ethylene glycol monomethylether,
t-butanol, dimethyl sulfoxide (DMSO), or dimethyl formamaide (DMF) in the final
product. This limitation could restrict the choice of solvents to only molten alkali,
glycerol, propylene glycol, water, and ethanol. The reaction in water and ethanol
only proceeds above the boiling temperature of these solvents and, therefore, a
pressure reactor would be required to operate using these solvents. Glycerol is
TABLE 3. Summary of Solvents Used in CLA Production.
Solvent Mol B.P. B.P. Temp Time % Phase

(reference) wt. (760) (Vac) ( C) (h) Reaction Sepn{ Color z Toxicity Other
Ethylene
glycol (148) 62.07 198.00 9313 180 2.0 100.0 yes Poor yes
Glycerol (148) 92.11 290.00 18220 180 0.75 100.0 yes Good none viscous
Propylene
glycol (149) 76.11 189.00 9721 170 2.5 99.1 yes Good minimal
t-butyl
alcohol (103) 74.18 82.41 Low 90 4.0 98.5 no unknown yes High
Pressure
Water (118) 18 100.00 Low 225 2.5 40.0 yes Good none High
Pressure
DMSO (149) 78.14 189.00 8317 30 1.5 78.0 no unknown yes fp 95
DMF (149) 73.09 153.00 7639 no unknown yes fp 67
{Phase Separation Yes if a two-phase system is formed after acidification of the soap.
zColor As described in references.

fp flash point.

Superscript numbers refer to Vacuum in mm of Hg.
DMF, Dimethylformamide; DMSO, Dimethylsulfoxide.
expensive, but it could be recovered from a commercial operation that produces its
own fatty acids. The quality of glycerol necessary to produce high-quality CLA has
not been investigated, but refining this glycerol stream to remove the salt might
prove difficult to a small operation. Recovery would have to be very efficient as
fatty acid production only generates 10% of the weight of the oil as glycerol.
5.6. Catalyst Selection

Numerous catalysts have been used in the production of CLA. We have found that
hydroxides of lithium, sodium, and potassium are all capable of generating CLA in
various solvents. As fatty acids neutralize the catalyst, it is necessary to add at least
one mole of catalyst for every mole of fatty acid in the reaction to ensure soap is
generated. We have found that, on a molar basis, potassium hydroxide has proven to
be a more effective catalyst than sodium hydroxide, with lithium hydroxide the
least effective and not suited for industrial CLA production. On a weight basis,
sodium and potassium hydroxide have similar efficiency of conversion. Although
sodium hydroxide is much less expensive than potassium hydroxide, the disposal
costs for the waste neutralized alkali should also be considered. Potassium salts
are easily used as fertilizer and can be applied to fields, whereas sodium salts
cannot be disposed of in a similar fashion.
The effective form of the catalyst is not necessarily determined by the added
catalyst itself but rather by the solvent used. When water is used as the solvent
and sodium ethoxide is the catalyst, the effective form of the catalyst is likely
the hydroxide ion. If t-butanol is used as the solvent and sodium methoxide is added
to catalyze the reaction at 90 C, the reaction mixture will quickly release methanol
vapours and t-butoxide ion will become the effective catalyst.
Water consumes alkoxide catalysts and, in industrial production, maintaining
alkoxide catalysts in a water-free environment is difficult. Water is produced by
the neutralization of fatty acids with alkali hydroxide. As many alkoxide catalysts
contain some alkali hydroxide, it is not uncommon for the catalyst to be consumed
by this reaction.
5.7. Reaction Vessels

Alkali isomerization of linoleyl soaps requires a containment vessel that is both tol-
erant of heat and caustic. When a low boiling point solvent such as ethanol or water
is used, the vessel must also be capable of maintaining the reaction under pressure.
There are a limited number of materials that will meet these criteria. Polytetrafluor-
oethylene and other fluoropolymers are capable of withstanding both the heat of the
reaction and the caustic environment, but they cannot withstand pressure and are
poor heat conductors (122). Fluoropolymer coated parts and nickel and nickel
alloys such as Monel may be used in the construction of reaction vessels for pro-
duction of CLA. The high cost of these materials rules out their use in construction
of large batch reactors. Furthermore, none of these materials has sufficient strength
for use in pressure reactors if a reaction in water or alcohol is planned. The pre-
ferred choice for reactor construction is nickel-plated steel, which has the desired
strength, heat transfer, and chemical properties for conducting reactions in strong
caustic solutions. A coated vessel of this design requires regular inspection, as a
flaw in the coating could lead to vessel failure.
5.8. Microbial Production of CLA

Pariza and Yang (123) have recently described the microbial production of 9,11-c,t-
CLA from linoleic acid using cultures of Lactobacillus sp. In their patented method,
early stationary phase Lactobacillus cultures were incubated with linoleic acid dis-
solved in propylene glycol. A total CLA level of 7 mg/g cells was produced, which
was over 96% 9,11-c,t-CLA. This type of conversion may lead to improved CLA
products in the future.
5.9. Synthesis of CLA by Dehydration of Ricinoleic Acid

(12-Hydroxy-cis-9-Octadecadienoic Acid)
The most attractive method for production of pure 9,11-c,t-CLA is through the
dehydration of ricinoleic acid. Synthesis from this relatively inexpensive starting
material has proven elusive as it is difficult to control the formation of dehydration
products (124). Synthesis of 9,11-c,t-CLA from ricinoleic acid has been reported
(125), which, although an efficient reaction, uses expensive elimination reagents
such as 1,8-diazobicyclo-(5,4,0)-undecene. For most applications, the high cost
of the elimination reagent increases the production cost beyond the level at which
commercial production of CLA is economically viable.
5.10. The Quality of Commercial CLA Products

The fate of other fatty acids and minor components during processing has not been
investigated. The conditions used to conjugate linoleic acid have little or no effect
on either monounsaturated or saturated fatty acids, however, any polyunsaturated
fatty acids may be conjugated. The products of the reaction of alkali catalysts on
these fatty acids are more complex than that discussed for linoleic acid (Reaction 4)
and will not be discussed except to note that these reactions may produce undesir-
able products.
Reaction 4: Isomerization of cis-,cis-,cis-9,12,15-octadecatrienoyl soap to isomers.
COO cis-,cis-,cis-9,12,15
base
COO cis-,trans-,cis-9,11,15
COO trans-,cis-,cis-10,12,15
COO cis-,trans-,cis-9,13,15
COO cis-,cis-,trans-9,12,14
From our observations, glycerol does not form undesirable compounds under the
conditions of alkali isomerization of linoleic acid. However, we have found that a
number of commercial fatty acids and CLA preparations contain appreciable levels
of monoacylglycerols (MAGs). MAGs themselves are not toxic, but it is possible

that toxic compounds (such as ethylene glycol) may be incorporated into the final
product through alcoholysis of the MAG by the low volatility alcohols used as a
reaction medium (Reaction 5).
Reaction 5: Alcoholysis of MAG with ethylene glycol.
OH
O
OH
O
OH
OH
HO
NMR and liquid chromatographic analysis of CLA samples from all sources
indicated that, although CLA was the predominant compound, CLA esters and other
unknown compounds may also have been formed.
The minor components of vegetable oil, such as tocopherol, sterol, or squalene,
are stable to heat a strong alkali. However, tocopherol and other components pre-
sent in vegetable oil readily react with oxygen in the presence of metals. Tocopherol
stabilizes vegetable oils against oxidation and tocopherol loss through processing or
high metal content may lead to a decreased shelf-life of the CLA product. Three
commercial CLA products were analyzed for metal content using ICP. The metal
contents are given in Table 4.
The CLA samples tested were free of potentially toxic metals at the levels tested,
with the exception of a trace level of barium in product C. Products A and B had
very low total metal contents, whereas product C had appreciable levels of calcium.
The soaps probably derived from a CaCl2 wash in a late stage of processing.
TABLE 4. Metal Contents of Three Commercial CLA Products (ppm).
Sample Na K Fe Cr B P Ca Ba Mg Li Total
A 3 3 1 0 5 2 5 0 1 0 20
B 2 28 0 0 3 8 1 0 1 0 43
C 4 6 11 8 3 3 917 1 2 1 956

Not detected Si, Pb, Cu, Sn, Al, Ni, Ag, Ti, Zn, Mo, V, Sb, Be.
Product C also contained iron and chromium suggesting the use of a stainless-steel
reactor for processing. Even these small levels of metals may contribute to rapid
oxidation of product C, and it probably has a shortened shelf-life when compared
with the other products.
Three commercial samples of CLA were subjected to a series of tests, which are
reported in Table 5. Viscosity was measured using ASTM test D445. It was pre-
sumed that increased viscosity might be related to increased oxidation if it had
occurred. Large differences in viscosity were not found. Color of the three samples
is reported, but it is only a subjective statement regarding the apparent quality of the
three materials. Included in the color analysis is the absorbance maximum observed
for CLA dissolved in hexane (1:1000 CLA:hexane) using a Cary UV/Vis spectro-
meter. The UV spectra indicated that conjugated dienes were the major source of
absorption of the oil samples.
Proton and carbon 13 NMR spectra were obtained from the three commercial
samples. A series of unknown spectral components were observed in product A
at 3.68 ppm, 3.98 ppm, 4.08 ppm, 4.15 ppm, 5.05 ppm, and 5.17 ppm. Comparison
of the unknown peaks with the spectra of MAGs indicated that most of the observed
peaks correlated with those present in the spectra of MAGs. An exact assignment of
the spectra was not attempted as many forms of MAGs are possible. Observation of
the spectral region of the olefinic protons revealed that the three samples were
TABLE 5. Summary of Analyses of Three Commercial CLA Products.

1
Viscosity H-NMR Size Isomer
Sample cSt (40 C) Color (400 MHz) Exclusion RP-LC GC-FID Mix
A 28.9 Clear Acylglycerols Acylglycerols Polar Contamination Good

l 231.5 contamination 95%
A 0.720
B 27.9 Light No No Polar OK Poor

yellow acylglycerols acylglycerols contamination 70%
l 231.5
A 0.667
C 29.7 Brown No No Non-Polar OK Good
l 231.5 acylglycerols acylglycerols contamination 90%
A 0.676
predominantly cis-, trans- or trans-, cis-fatty acids (126). We have found that a
convenient measure of the CLA content of an oil can be obtained by comparing
the integrated values of the protons at 6.0 ppm and 6.3 ppm with the integrated
value of alpha methylene group (adjacent to the carbonyl) at 2.3 ppm. The olefinic
protons were chosen as they do not overlap with oleic acid olefinic protons or
olefinic protons from conjugated linoleic acid with trans-, trans-configurations.
The ratio of cis-, trans-olefin to alpha methylene protons is a useful measure of
CLA purity.
Carbon-13 olefinic carbons were observed at 400 MHz. The carbon-13 spectrum
clearly demonstrates the formation of positional CLA isomers. As pure standards
were not available, it was not possible to unequivocally assign the spectra of
8,10- or 11,13- CLA, but it is clear that these isomers are predominantly cis-, trans-
or trans-, cis-fatty compounds.
Size exclusion chromatography was performed with a Waters GPC-StyragelTM
HR 0.5 column (7.8 300 mm) at 35 C using tetrahydrofuran as a solvent flowing
at 1.0 mL/min and detecting compounds with both UV absorbance and evaporative
light scattering detection (ELSD, 40 C, 4.2 L/min gas flow). The goal of this ana-
lysis was to observe polymerization of the CLA products or the occurrence of acyl-
glycerols. Under these conditions, 18-carbon fatty acids had a retention time of 6.6
minutes and MAGs had a retention time of 5.9 minutes. All three samples had two
peaks observed by ELSD, one at the position expected for 18-carbon fatty acids and
the other as expected for their respective MAGs. Product A had the largest peak at
the position expected for MAGs. Observations made using UV absorbance at
233 nm reflected the observations made with the ELSD.
Reversed phase liquid chromatography was performed on a 150 mm 3.0 mm
Waters SymmetryTM C-8 column at 30 C and a flow rate of 1.0 mL/min. The
solvent phase was acetonitrile:tetrahydrofuran : 0.1% aqueous phosphoric acid
(50.4:21.6:28v/v/v). Under these operating conditions, most of the UV absorbance
occurred as a peak at 3.83 minutes for all samples. Chromatograms of all samples
had some small peaks (presumably the more polar compounds) eluting prior to the
major peak. Product C presented a small but significant UV absorbing peak that
eluted after the peak at 3.83 minutes.
5.11. Rapid Analytical Methods

Industrial CLA syntheses must be controlled to both maximize the content of
preferred isomers such as 9,11-cis-, trans-octadecadienoic acid and minimize the
formation of undesirable isomers. For the analysis to be useful, the results of the
analysis should be available online or as quickly as possible. The authors are not
aware of any existing online tests for the quality of CLA preparations, but several
methods may have promise for use offline.
The potential offline analytical methods include UV, FTIR, proton NMR, car-
bon-13 NMR, gas chromatography, and capillary electrophoresis. With the excep-
tion of capillary electrophoresis and UV absorbance, none of these methods can
ANALYSIS OF CONJUGATED LINOLEIC ACIDS 21
effectively analyze soap solutions and, therefore, for most analytical methods, neu-
tralization of soaps would be necessary.
A reaction medium that contains 40% soaps by weight can usually be dissolved
in ethanol. We have found that 100 mg of reaction mixture will totally dissolve in
10 mL of 95% ethanol when glycerol, water, or ethylene glycol are the reaction sol-
vent and alkali hydroxides are used as the catalyst. It is then possible to dilute the
reaction mixture solution 1000-fold to determine the UV absorbance at 231.5 nm.
When there is no other interfering UV absorbance, this method is an excellent indi-
cation of the total conjugated double bonds. This method is also sensitive to the
presence of conjugated linolenic acids derived from linolenic acid, which is indi-
cated by a UV absorbance at 268 nm. None of the samples observed show three
conjugated double bonds.
To obtain more detailed information regarding the composition of fatty acids
requires additional analytical methods, some of which require extensive and
time-consuming sample preparation. For these methods, we have found that it is
possible to rapidly prepare a free fatty acid fraction. Alkali soaps from most reac-
tion mixtures are readily dissolved in a mixture of hexane and ethanol (1 : 1v/v) or
ethanol alone. When the soaps are neutralized by the addition of hydrochloric acid
and water, a two-phase system is evolved. The fatty acids remain in the nonpolar
phase while the polar solvent used in the reaction medium dissolves in the water.
The solution of dissolved fatty acids can be directly injected onto a GC column
specifically designed for separation of free fatty acids, such as the DB-FFAP col-
umn, or a suitable nonpolar GC column, such as the HP-5 column. Analysis by
chromatography without derivitization affords the potential for rapid analytical
feedback. We have found that a 30 m, DB-FFAP column (0.32 mm id, 0.25 m
film thickness) gave baseline resolution of most fatty acids without derivitization
(program 50 C for 1 min, 50200 C @ 25 C/min, 200220 C @ 2 C/min,
hold 20 min, He carrier 3 ml/min). The nonpolar HP-5 column (0.32 ID, 0.25
film) gave poorer resolution of underivatized fatty acids with some tailing (program
50 C for 1 min, 50150 C @ 25 C/min, 150290 C @ 10 C/min, hold 6 min, He
carrier 2 ml/min). Both columns also partially separated isomers of CLA. It was
possible to observe the formation of the all trans-isomers, but detailed analysis
of positional isomers was not possible without additional effort.
6. ANALYSIS OF CONJUGATED LINOLEIC ACIDS
An analyst needs to recognize three major variables before the selection of a sui-
table method of analysis of CLA. CLA preparations can differ in degree of conju-
gation, position of double bonds, and configuration of double bonds. The most basic
variable is the degree of conjugation of the double bonds. Not all of the isolated
double bonds may become conjugated either in biological or chemical conversions.
The second issue to be addressed is the position of the double bonds in conjugated
linoleic acids. Initially, in linoleic acid, the double bonds begin at the ninth and
twelfth carbon atoms. After alkali isomerization, the predominant positional isomers
are 9,11 and 10,12 dienes. However, minor amounts of other positional isomers are
also reported (127). In biological samples, the number of positional isomers is more
varied with 16 isomers being separated by silver-ion chromatography (128). The
final consideration is the geometric isomers present in CLA. Whereas the double
bonds in linoleic acid are cis-, cis-, the isomerized product contains a predominance
of cis-, trans- and trans-, cis-isomers plus lesser amounts of cis-, cis- and trans-,
trans-isomers. Thus, the task for the analyst is to decide what information is
required and then select a method or methods that will provide that information.
6.1. Ultraviolet Spectroscopy

The early observation of an increase in the ultraviolet absorption (UV) near 233 nm
was recognition that the composition of milkfat was not consistent throughout the
year (129). This increase in absorption was not initially attributed to the increase in
conjugated dienes present in milk fatty acids; however, it is now accepted that this
is the case. There are subtle differences in the absorption maxima for the geometric
isomers. Czauderna (130) reported a maximum of 231.9 nm for the trans-, trans-
isomer, 234.3 nm for the cis-, trans- (or trans-, cis-) isomer and 235.4 nm for the
cis-, cis-isomer. These small differences would not likely be discernable in isomer
mixtures typically found in either biologically or chemically produced CLA. The
UV measurement provides no information on the position of the double bonds.
The advantage of using UV is the comparatively low cost of the spectrophotometer
and the fact that it can be performed on either intact acylglycerols, fatty acids, or
esters. In Figure 1, the UV spectra of linoleic acid and a commercial CLA product
are shown.
UV-Vis Spectra of Linoleic Acid and CLA

1.2
9,11, c,t,CLA + 10,12,t,CLA

1.0
Linoleic Acid
0.8
Absorbance
0.6
0.4
0.2
0
200 250 300 350 400 450
Wavelength (nm)
Figure 1. The ultraviolet spectra of linoleic acid and conjugated linoleic acid (CLA).
6.2. Infrared Spectroscopy

Infrared spectroscopy (IR) initially used double beam optics, photocells, and alkali
halide sample cells. These features at times tended to reduce the sensitivity of the
instrument and to decrease the signal to noise ratio. Newer instruments have
improved the signal to noise ratio by using Fourier Transform (FT) methodology
and modern electronics. The application of infrared spectroscopy to the problem
of identification of geometric isomers was first reported about 50 years ago. It
was recognized that cis-, trans- (or trans-, cis-) dienes had a characteristic doublet
at 948 cm1 and 982 cm1 (131). The corresponding trans-, trans-diene had a
strong absorption band at 988 cm1. Using FTIR and a direct deposit interface,
Mossaba (132) reported the cis-, trans-doublet to occur at 949 cm1 and
988 cm1 and the trans-, trans-singlet at 993 cm1 for 4,4-dimethyloxazoline
(DMOX) derivatives of CLA. An example of the spectra of linoleic acid and two
CLA positional isomers obtained with a FTIR using a synthetic sample disc are
shown in Figure 2A. In addition to these absorptions, there are other characteristic
frequencies attributed to carbon-hydrogen stretching. In one report of 4,4-
dimethyoxazoline derivatives (DMOX) of CLA isomers, it was possible to attribute
unique differences among geometric isomers (133). The cis-, trans-isomers had
characteristic bands at 3020 cm1 and 3002 cm1 (see Figure 2B); the cis-, cis-iso-
mers had bands at 3007 cm1 and 3005 cm1; and the trans-, trans-isomer had a
single band at 3017 cm1. The use of these bands to determine geometric composi-
tion may be complicated by the much more intense absorption bands from other
carbon-hydrogen bands present in the same region of the spectrum. Although it
is theoretically possible to obtain a quantitative measurement of the amount of a
substance present by IR, it is not often done. In the case of CLA, the bands that
needed to be measured are minor bands compared with other bands, making the
determination imprecise and difficult.
100 100
Linoleic Acid
% Transmittance
% Transmittance
80 80 9,11,c,t,CLA
10,12,t,c,CLA
60 60
40 40
Linoleic Acid
9,11,c,t,CLA
20 20
a 10,12,t,c,CLA b
0 0
1000 980 960 940 920 900 880 3040 3020 3000 2980 2960 2940
Wavelength (nm) Wavelength (nm)
Figure 2. Infrared spectra of linoleic acid and 2 cis-, trans-, isomers of conjugated linoleic acid
(A) 1000 to 900 nm of the
CH vibration band and, (B) 3050 to 2950 nm of the C H stretching
bands.
6.3. Thin Layer Chromatography

Thin layer chromatography (TLC) has proven very useful in the separation of a vast
array of chemicals both synthetic and naturally occurring. When it comes to separa-
tion of closely related diene isomers, it is a difficult task with unmodified adsor-
bents such as silica gel. Silver nitrate-modified thin layer chromatography (Ag-
TLC) has been used to separate CLA isomers. Some cis-, trans-isomers of CLA
have been separated as their methyl esters using hexane/diethyl ether, benzene,
or toluene (102). The initial solvent mixture caused the cis-, trans-isomers to
migrate faster than the cis-monoenes. Either of the later solvents resulted in the
cis-, trans-isomer migrating between the cis- and trans-monoenes. Ackman (134)
reported that the cis-, trans-isomer had a relative retention (Rf) of 0. 61, whereas the
trans-, trans-isomer had a Rf of 0.65. Although the difference in the Rf values is not
great, it is possible to separate the isomers to either collect larger quantities or
investigate possible metabolites.
6.4. Gas Chromatography

Gas chromatography (GC), initially using packed columns and later using capillary
columns, provides another method for the analysis of fatty acids and esters or other
derivatives. As binding of the fatty acids to the column was problematic, fatty acid
methyl esters (FAME) were traditionally used. The method of forming the methyl
esters can be divided into three procedures: acid-catalyzed, base-catalyzed, and dia-
zomethane alkylation (135). No one method is suitable for all situations and all suf-
fer some deficiency.
Acid-catalyzed procedures typically are either boron trifluoride/methanol (BF3/
CH3OH) or hydrochloric acid/methanol (HCl/CH3OH). Many analysts use sulfuric
acid instead of hydrochloric acid. Acid-catalyzed procedures are used for free fatty
acids, phospholipids, or triacylglycerols, often at elevated temperatures. Although
the procedures are relatively efficient at production of methyl esters, there will be
some isomerization of some cis-, trans-isomers to the trans-/trans-isomers (136).
This isomerization can be reduced by using lower temperatures, for instance
60 C, for HCl/CH3OH or room temperature for BF3/CH3OH (137). However,
under these milder conditions, some phospholipids may not be esterified (137). In addi-
tion, methoxy adducts may be formed and hydroxy fatty acids may produce artifacts.
A base, such as sodium methoxide (NaOCH3), is useful in esterifying lipids as
found in acylglycerols, sterol esters, and phospholipids (138). Free fatty acid and N-
acyl lipids in sphingolipids will not be methylated. The NaOCH3 method does not
apparently change the cis-, trans-isomer composition or form methoxy artifacts.
However, some artifacts that could interfere with shorter chain fatty acids were
observed.
Diazomethane is effective in esterifying free acids to their corresponding methyl
esters under mild conditions and is fast. It will not produce methyl esters from acyl-
glycerols, cholesterol esters, or phospholipids. Many researchers are concerned
about the potential hazard of the diazomethane, its precursors, and the actual
25.218
25.722
90 cis-9,trans-11 CLA
20.522
80 trans-10,cis-12 CLA
15.236
70
60 Oleic
pA
50
trans-10,trans-12 CLA
19.577
40
27.880
21.944
26.634
30 Palmitic
20 Stearic trans-9,trans-11 CLA
Linoleic
10
0
0 5 10 15 20 25 30 35
Retention Time (min)
Figure 3. Gas chromatogram of free fatty acids on a 30-m DBFFAP column (0.32 mm id,
0.25-um film thickness) using a temperature gradient of 50 C for 1 min; 50 C150 C @ 25 C/
min, 150 C220 C @ 3 C/min; hold at 220 C for 3 min; He carrier gas @ 2 mL/min.
preparation of the reagent. Trimethylsilyl diazomethane is commercially available

and can be used as a source of diazomethane. However, some artifacts (trimethyl-
silyl CLA esters) and other trimethylsilyl impurities may interfere with analysis
(139). As there are limitations to each method of making methyl esters, the
analyst will need to select reagents and conditions that are most appropriate for
the substrate. For intact acylglycerols, sterol esters, and phospholipids, the
NaOCH3 method would be appropriate. If only free fatty acids are present, the dia-
zomethane method would be appropriate, particularly if double bond isomerization
is of concern.
The earliest of GC analyses were performed on columns packed with a solid sup-
port coated with a nonvolatile liquid phase. Packed columns are not frequently used
today as they have been replaced by capillary columns where the liquid phase is
immobilized on the internal surface of the capillary. As there are numerous liquid
phases available, it is now possible to obtain commercial columns that will separate
not only the methyl esters but also the underivatized fatty acids. This advancement
obviates the need for derivatization and the associated problems. A typical chroma-
togram of free fatty acids is displayed in Figure 3. Individual isomers of CLA
are now available to aid in the identification of isomers in the chromatogram.
Gas chromatography can provide quantitative information on the degree of
conjugation, positional, and geometric isomer distribution when suitable standards
are available.
6.5. High-Performance Liquid Chromatography

High-performance liquid chromatography (or less common, high-pressure liquid
chromatography, HPLC) is a preferred method of analysis for many compounds
because it does not require the high temperatures used in gas chromatography.
Separations in HPLC can be based on either a size exclusion or on an adsorption
principle. The size exclusion mode is useful for separating fatty acids from
acylglycerols and has been applied to CLA analysis. Reaney et al. (120) have used a
Waters GPC-StyragelHRTM column for the purpose of observing possible polymer-
ization of CLA or the presence of acylglycerols. In three commercial samples
of CLA, the 18-carbon fatty acids could be separated from the MAGs. Both UV
detection at 233 nm and evaporative light scattering detection (ELSD) were used.
There was no reported separation of individual fatty acids or CLA isomers. Use
of a reverse phase column (Waters C-8 SymmetryTM column) for separation of
commercial CLA resulted in a major peak being observed that would correspond
to the free fatty acids (120). As with the size exclusion column, there was no
separation of either isomeric conjugated fatty acids or acids of different chain
length. A preferred technique is silver ion-modified high-performance liquid
chromatography (Ag-HPLC). Using this modification, several groups have
reported successful separation of both positional and geometric isomers of CLA
(14, 140).
6.6. Mass Spectrometry

Mass spectrometry (MS) is a method to determine the mass of either an intact
ionized molecule or an ionized fragment. When MS is combined with the separa-
tion power of gas or liquid chromatography, much valuable information can be
obtained for structural determination. Although GC-MS or HPLC-MS may seem
to be an ideal tool for CLA analysis, it suffers a major problem as originally prac-
ticed. In order to obtain a mass spectrum, it is necessary to produce an ionized spe-
cies. With CLA, as well as other molecules, some ionization conditions cause the
double bonds to migrate to new positions. If the double bonds migrate, it is only
possible to obtain information on chain length and number of double bonds. To
allow for less harsh ionization conditions, it is possible to make derivatives at either
the carboxylic acid (remote site) or at the site of the conjugated double bonds (on
site). The derivatives have been selected to allow for easier ionization. In the case of
the on-site derivatives, the position of the diene is fixed at one location. At the car-
boxylic acid site pyrrolidide, picolinyl ester and 4,4-dimethyloxazoline (DMOX)
derivatives are used. The fragmentation of the remote site derivative basically pro-
duces ions that are 14 mass units lighter than the previous ion for each methylene
(CH2) group. When a double bond is encountered, the decrease observed is 12 mass
units. A review on this subject has been compiled (141).
Two different approaches have been reported for on-site or double bond site deri-
vatization of fatty acids or other conjugated dienes. One method involves the com-
plete hydroxylation of the double bonds with osmium tetroxide (health caution)
followed by trimethylsilylation (142). The second method is based on the well
known Diels-Alder cyclo addition. MTAD (4-methyl-1,2,4-triazoline-3-5-dione)
has been shown to be a useful reagent for adduct formulation with FAME conju-
gated dienes. Fragmentation patterns of adducts are usually dominated by cleavage
fragments that include the ring formed during the cyclo addition plus either of the
residual carbon chains. In Figure 4 of the MTAD derivative of methyl cis-, trans-
9,11-octadecadienoate, these fragments occur at 322 m/z and 250 m/z, indicating
70 250
322 m/z
60 250 m/z
Abundance (thousands)
50
CH3(CH2)5 (CH2)7COOCH3
40 N N
30 O N O 290 322
CH3
20
M+
10
376 407
0
50 100 150 200 250 300 350 400 450

m/z
Figure 4. The mass spectrum of the adduct of methyl cis-, trans-9,11-octadecadienoate and 4-
methyl-1,2,4-triazoline-3,5-dione (MTAD). Reproduced by permission of AOCS Press (143).
that the starting FAME was a 9,11 diene (143). Methyl cis-, trans-9,11-octadeca-
dienoate and methyl trans-, trans-9-11-octacadienoate readily formed adducts with
similar mass spectra but with different retention times when analyzed by GC-MS
(143). Methyl cis-, cis-9,11-octadecadienoate reacted more slowly to produce two
products with similar fragmentation patterns but with different retention times.
Unfortunately, it was demonstrated that the cis-, cis- and cis-, trans-isomers pro-
duced the same adduct, limiting usefulness when studying geometric isomers of
CLA products.
6.7. Nuclear Magnetic Resonance

Nuclear magnetic resonance (NMR) is a powerful technique for CLA analysis
(144, 145). NMR spectroscopy can provide information on the environments of both
the proton (1H) and the carbon-13 (13C) atoms in CLA and also provide correlation
between the two atoms. In one of the most basic analyses, it is possible to observe
the disappearance of the signals as a result of the isolated olefinic protons
(5.35 ppm, Figure 5A) and the isolated methylene group (2.75 ppm) linoleic acid
as well as the appearance of new signals associated with the conjugated diene
(5.26.5 ppm, Figure 5B). Reaney et al. (120) reported that a convenient measure-
ment of CLA content can be obtaining by comparing the integrated areas of the
olefinic protons at 6.0 ppm and 6.3 ppm with the area of the methylene protons
adjacent to the carbonyl (2.3 ppm). The NMR 13C spectra for both linoleic and
cis-, trans-9,11-octadecadienoic acids are presented in Figure 5C and 5D, respec-
tively. The complete assignment for this and other isomers of CLA has been
reported (146, 147). A notable feature for CLA is the appearance of four well sepa-
rated signals between 129135 ppm (Figure 5D) compared with the two pairs of
7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm
7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm
190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 ppm
190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 ppm
Figure 5. (A) The 1H spectrum of linoleic acid; (B) the 1H spectrum of cis-, trans-9,11-
octadecadienoic acid; (C) the 13C spectrum of linoleic acid; and, (D) the 13C spectrum of cis,
trans-9,11-octadecadienoic acid.
narrowly separated doublets observed in linoleic acid spectrum (Figure 5C).

Detailed 1H and 13C and NMR spectroscopy have the potential to provide
information on both positional and geometric isomers of CLA and can provide
semiquantitative information on CLA concentration.
REFERENCES 29
7. CONCLUSIONS
The research on CLA in growing animals is consistently showing effect on modu-

lation of body mass and fat, however, the effect in humans is not consistent. More
research is needed to delineate the effect of CLA and isomers on body composition
in humans. Major research emphasis, at present, is focused on the effects of CLA
and its isomers on body composition and carcinogenesis. Other areas that are
attracting attention include the effects of CLA and isomers on cardiovascular, meta-
bolic, and immune functions and the strategies to increase the content of CLA iso-
mers in meat and dairy products.
At the same time, research is still needed to improve the commercial production
of CLA. Although the content of desirable isomers in commercial CLA products
has improved, there is still a demand for highly enriched or pure 9,11-cis-, trans-
octadecadienoic acid products. The kinetic control of CLA synthesis will allow the
development of CLA products that are virtually free of isomers other than 9,11-ct
and 10,12-tc. Kinetic control of reactions requires exceedingly rapid analytical
techniques that can be applied inexpensively and online or virtually online.
REFERENCES
1. W. J. Dann and T. Moore, Biochem. J., 27, 11661169 (1933).

2. R. Kapoor, N. D. Westcott, M. J. T. Reaney, and S. Jones, Inform, 14, 482484
(2003).
3. M. J. T. Reaney, N. D. Westcott, R. Kapoor, and S. Jones, Inform, 13, 802807 (2002).
4. W. J. Dann, T. Moore, R. L. Booth, J. Golding, and S. K. Kon, Biochem. J., 29, 138146
(1935).
5. R. Reiser, Arch Biochem. Biophys., 32, 113120 (1951).
6. P. W. Parodi, J. Dairy Sci., 60, 15501553 (1977).
7. E. S. Miller and G. O. Burr, Proc. Soc. Exp. Biol. Med., 36, 726729 (1937).
8. E. Aaes-Jorgensen, J. Nutr., 66, 465483 (1958).
9. R. Reiser, Proc. Soc. Exp. Biol. Med., 74, 666669 (1950).
10. R. H. Barnes, E. S. Miller, and G. O. Burr, Proc. Soc. Exp. Biol. Med., 45, 233240
(1941).
11. M. A. Belury and A. Kempa-Steczko, Lipids, 32, 199204 (1997).
12. C. Ip, C. Jiang, H. J. Thompson, and J. A. Scimeca, Carcinogenesis, 18, 755759 (1997).
13. J. L. Sebedio, E. Angioni, J. M. Chardigny, S. Gregoire, P. Juaneda, and O. Berdeaux,
Lipids, 36, 575582 (2001).
14. J. K. Kramer, N. Sehat, M. E. Dugan, M. M. Mossoba, M. P. Yurawecz et al., Lipids, 33,
549558 (1998).
15. J. L. Sebedio, P. Juaneda, G. Dobson, I. Ramilison, J. C. Martin et al., Biochim. Biophys.
Acta, 1345, 510 (1997).
16. J. L. Sebedio, S. Gnaedig, and J. M. Chardigny, Curr. Opin. Clin. Nutr. Metab. Care., 2,
499506 (1999).
17. J. L. Sebedio, P. Juaneda, S. Gregoire, J. M. Chardigny, J. C. Martin, and C. Ginies,

Lipids, 34, 13191325 (1999).
18. L. Bretillon, J. M. Chardigny, S. Gregoire, O. Berdeaux, and J. L. Sebedio, Lipids, 34,
965969 (1999).
19. J. L. Sebedio, E. Angioni, J. M. Chardigny, S. Gregoire, P. Juaneda, and O. Berdeaux,
Lipids, 36, 575582 (2001).
20. P. Urquhart, S. M. Parkin, J. S. Rogers, J. A. Bosley, and A. Nicolaou, Biochim. Biophys.
Acta, 1580, 150160 (2002).
21. R. T. Holman, Proc. Soc. Exp. Biol. Med., 76, 100102 (1951).
22. A. Akahoshi, Y. Goto, K. Murao, T. Miyazaki, M. Yamasaki et al., Biosci. Biotechnol.
Biochem., 66, 916920 (2002).
23. E. Ostrowska, D. Suster, M. Muralitharan, R. F. Cross, B. J. Leury et al., Br. J. Nutr., 89,
219229 (2003).
24. M. Yamasaki, A. Ikeda, M. Oji, Y. Tanaka, A. Hirao et al., Nutrition, 19, 3035 (2003).
25. H. Blankson, J. A. Stakkestad, H. Fagertun, E. Thom, J. Wadstein, and O. Gudmundsen,
J. Nutr., 130, 29432948 (2000).
26. U. Riserus, L. Berglund, and B. Vessby, Int. J. Obes. Relat. Metab. Disord., 25, 1129
1135 (2001).
27. E. Thom, J. Wadstein, and O. Gudmundsen, J. Int. Med. Res., 29, 392396 (2001).
28. M. A. Belury, A. Mahon, and S. Banni, J. Nutr., 133, 257S260S (2003).
29. K. L. Zambell, N. L. Keim, M. D. Van Loan, B. Gale, P. Benito et al., Lipids, 35, 777782
(2000).
30. M. M. Kamphuis, M. P. Lejeune, W. H. Saris, and M. S. Westerterp-Plantenga, Int.
J. Obes. Relat. Metab. Disord., 27, 840 847 (2003).
31. Y. Park, K. J. Albright, W. Liu, J. M. Storkson, M. E. Cook, and M. W. Pariza, Lipids, 32,
853858 (1997).
32. J. M. Brown and M. K. McIntosh, J. Nutr., 133, 30413046 (2003).
33. E. A. De Deckere, J. M. van Amelsvoort, G. P. McNeill, and P. Jones, Br. J. Nutr., 82,
309317 (1999).
34. Y. Park, J. M. Storkson, K. J. Albright, W. Liu, and M. W. Pariza, Lipids, 34, 235241
(1999).
35. K. Kang and M. W. Pariza, Biochem. Biophys. Res. Commun., 287, 377382 (2001).
36. K. Nagao, Y. M. Wang, N. Inoue, S. Y. Han, Y. Buang et al., Nutrition, 19, 652656
(2003).
37. K. Kang, W. Liu, K. J. Albright, Y. Park, and M. W. Pariza, Biochem. Biophys. Res.
Commun., 303, 795799 (2003).
38. J. M. Brown, M. S. Boysen, S. S. Jensen, R. F. Morrison, J. Storkson et al., J. Lipid. Res.,
44, 12871300 (2003).
39. N. Tsuboyama-Kasaoka, M. Takahashi, K. Tanemura, H. J. Kim, T. Tange et al.,
Diabetes, 49, 15341542 (2000).
40. N. Tsuboyama-Kasaoka, H. Miyazaki, S. Kasaoka, and O. Ezaki, J. Nutr., 133, 1793
1799 (2003).
41. U. Riserus, P. Arner, K. Brismar, and B. Vessby, Diabetes Care, 25, 15161521 (2002).
42. M. W. Pariza and W. A. Hargraves, Carcinogenesis, 6, 591593 (1985).
REFERENCES 31
43. Y. L. Ha, N. K. Grimm, and M. W. Pariza, Carcinogenesis, 8, 18811887 (1987).

44. Y. L. Ha, J. Storkson, and M. W. Pariza, Cancer Res., 50, 10971101 (1990).
45. T. D. Shultz, B. P. Chew, and W. R. Seaman, Anticancer Res., 12, 21432145 (1992).
46. C. Ip, M. Singh, H. J. Thompson, and J. A. Scimeca, Cancer Res., 54, 12121215 (1994).
47. F. Lavillonniere, V. Chajes, J. C. Martin, J. L. Sebedio, C. Lhuillery, and P. Bougnoux,
Nutr. Cancer, 45, 190194 (2003).
48. A. Cesano, S. Visonneau, J. A. Scimeca, D. Kritchevsky, and D. Santoli, Anticancer
Res., 18, 14291434 (1998).
49. J. L. Cheng, M. Futakuchi, K. Ogawa, T. Iwata, M. Kasai et al., Cancer Lett., 196, 161
168 (2003).
50. H. S. Park, J. H. Ryu, Y. L. Ha, and J. H. Park, Br. J. Nutr., 86, 549555 (2001).
51. V. R. Durgam and G. Fernandes, Cancer Lett., 116, 121130 (1997).
52. C. J. Kavanaugh, K. L. Liu, and M. A. Belury, Nutr. Cancer, 33, 132138 (1999).
53. H. J. Cho, W. K. Kim, E. J. Kim, K. C. Jung, S. Park et al., Amer. J. Physiol. Gastrointest.
Liver Physiol., 284, G9961005 (2003).
54. A. Miller, C. Stanton, and R. Devery, Anticancer Res., 22, 38793887 (2002).
55. P. A. Masso-Welch, D. Zangani, C. Ip, M. M. Vaughan, S. F. Shoemaker et al., J. Nutr.,
134, 299307 (2004).
56. C. Ip, Y. Dong, H. J. Thompson, D. E. Bauman, and M. M. Ip, Nutr. Cancer, 39, 233238
(2001).
57. C. Ip, J. A. Scimeca, and H. Thompson, Nutr. Cancer, 24, 241247 (1995).
58. A. Aro, S. Mannisto, I. Salminen, M. L. Ovaskainen, V. Kataja, and M. Uusitupa, Nutr.
Cancer, 38, 151157 (2000).
59. K. L. Houseknecht, J. P. Vanden Heuvel, S. Y. Moya-Camarena, C. P. Portocarrero,
L. W. Peck et al., Biochem. Biophys. Res. Commun., 244, 678682 (1998).
60. J. W. Ryder, C. P. Portocarrero, X. M. Song, L. Cui, M. Yu et al., Diabetes, 50, 1149
1157 (2001).
61. L. Clement, H. Poirier, I. Niot, V. Bocher, M. Guerre-Millo et al., J. Lipid Res., 43, 1400
1409 (2002).
62. K. M. Hargrave, M. J. Azain, S. D. Kachman, and J. L. Miner, Obes. Res., 11, 11041115
(2003).
63. K. Nagao, N. Inoue, Y. M. Wang, and T. Yanagita, Biochem. Biophys. Res. Commun.,
310, 562566 (2003).
64. E. Ostrowska, R. F. Cross, M. Muralitharan, D. E. Bauman, and F. R. Dunshea, Br. J. Nutr.,
88, 625634 (2002).
65. E. J. Noone, H. M. Roche, A. P. Nugent, and M. J. Gibney, Br. J. Nutr., 88, 243251
(2002).
66. K. N. Lee, D. Kritchevsky, and M. W. Pariza, Atherosclerosis, 108, 1925 (1994).
67. D. Kritchevsky, S. A. Tepper, S. Wright, P. Tso, and S. K. Czarnecki, J. Amer. Coll. Nutr.,
19, 472S477S (2000).
68. R. J. Nicolosi, E. J. Rogers, D. Kritchevsky, J. A. Scimeca, and P. J. Huth, Artery, 22,
266277 (1997).
69. S. Toomey, H. Roche, D. Fitzgerald, and O. Belton, Biochem. Soc. Trans., 31, 10751079
(2003).
70. V. C. Gavino, G. Gavino, M. J. Leblanc, and B. Tuchweber, J. Nutr., 130, 2729

(2000).
71. S. Yu-Poth, D. Yin, G. Zhao, P. M. Kris-Etherton, and T. D. Etherton, J. Nutr., 134, 6871
(2004).
72. M. W. Pariza and K. N. Lee (assigned to Wisconsin Alumni Research Foundation,
(Madison, Wisconsin), U.S. Patent 5,837,733, 1998.
73. A. Truitt, G. McNeill, and J. Y. Vanderhoek, Biochim. Biophys. Acta, 1438, 239246
(1999).
74. K. Nagao, N. Inoue, Y. M. Wang, J. Hirata, Y. Shimada et al., Biochem. Biophys. Res.
Commun., 306, 134138 (2003).
75. P. Benito, G. J. Nelson, D. S. Kelley, G. Bartolini, P. C. Schmidt, and V. Simon, Lipids,
36, 229236 (2001).
76. M. E. Cook, B. Ellen, J. L. Stahl, F. M. Graziano, M. Frank, and M. W. Pariza (assigned to
Wisconsin Alumni Research Foundation, Madison, Wisconsin), U.S. Patent 5,585,400,
1996.
77. M. E. Cook, M. W. Pariza, X. Yang, and D. De Voney (assigned to Wisconsin Alumni
Research Foundation, (Madison, Wisconsin), U.S. Patent 5,674,901, 1997.
78. M. G. Hayek, S. N. Han, D. Wu, B. A. Watkins, M. Meydani et al., J. Nutr., 129, 3238
(1999).
79. L. D. Whigham, E. B. Cook, J. L. Stahl, R. Saban, D. E. Bjorling et al., Amer. J. Physiol.
Regul. Integr. Comp. Physiol., 280, R908R912 (2001).
80. M. W. Wong, B. P. Chew, T. S. Wong, H. L. Hosick, T. D. Boylston, and T. D. Shultz,
Anticancer Res., 17, 987993 (1997).
81. M. E. Cook, S. Kim, M. W. Pariza, and D. De Voney (assigned to Wisconsin Alumni
Research Foundation, Madison, Wisconsin), U.S. Patent 5,914,346, 1999.
82. Y. Yu, P. H. Correll, and J. P. Vanden Heuvel, Biochim. Biophys. Acta, 1581, 8999
(2002).
83. J. Bassaganya-Riera, R. Hontecillas, D. R. Zimmerman, and M. J. Wannemuehler,
Vaccine, 20, 14351444 (2002).
84. J. Bassaganya-Riera, R. M. Pogranichniy, S. C. Jobgen, P. G. Halbur, K. J. Yoon et al.,
J. Nutr., 133, 32043214 (2003).
85. V. Bontempo, D. Sciannimanico, G. Pastorelli, R. Rossi, F. Rosi, and C. Corino, J. Nutr.,
134, 817824 (2004).
86. D. S. Kelley, P. C. Taylor, I. L. Rudolph, P. Benito, G. J. Nelson et al., Lipids, 35, 1065
1071 (2000).
87. D. S. Kelley, V. A. Simon, P. C. Taylor, I. L. Rudolph, P. Benito et al., Lipids, 36, 669674
(2001).
88. W. L. Cheng, C. K. Lii, H. W. Chen, T. H. Lin, and K. L. Liu, J. Agric. Food Chem., 52,
7178 (2004).
89. M. E. Cook, M. W. Pariza, K. N. Lee, and B. C. Wentworth (assigned to Wisconsin
Alumni Research Foundation, Madison, Wisconsin), U.S. Patent 5,504,114, 1996.
90. R. H. Stock and J. D. Compton (assigned to LifeRight Foods LLCPN, Princeton,
New Jersey), U.S. Patent 6,316,041, 2001.
91. M. E. Cook, A. Rahim, and M. W. Pariza (assigned to Wisconsin Alumni Research
Foundation, Madison, Wisconsin), U.S. Patent 6,113,973, 2000.
REFERENCES 33
92. M. L. Kelly, J. R. Berry, D. A. Dwyer, J. M. Griinari, P. Y. Chouinard et al., J. Nutr., 128,

881885 (1998).
93. K. Nurmela and M. Griinari (assigned to Valio Oy, Helsinki, Finland), U.S. Patent
6,635,271, 2003.
94. A. T. Ward, K. M. Wittenberg, H. M. Froebe, R. Przybylski, and L. Malcolmson, J. Dairy
Sci., 86, 17421750 (2003).
95. T. R. Dhiman, G. R. Anand, L. D. Satter, and M. W. Pariza, J. Dairy Sci., 82, 21462156
(1999).
96. T. R. Dhiman, E. D. Helmink, D. J. McMahon, R. L. Fife, and M. W. Pariza, J. Dairy Sci.,
82, 412419 (1999).
97. T. R. Dhiman, L. D. Satter, M. W. Pariza, M. P. Galli, K. Albright, and M. X. Tolosa,
J. Dairy Sci., 83, 10161027 (2000).
98. A. T. Ward, K. M. Wittenberg, and R. Przybylski, J. Dairy Sci., 85, 11911196 (2002).
99. L. A. Whitlock, D. J. Schingoethe, A. R. Hippen, K. F. Kalscheur, R. J. Baer et al.,
J. Dairy Sci., 85, 234243 (2002).
100. J. J. Loor and J. H. Herbein, J. Nutr., 128, 24112419 (1998).
101. S. T. Franklin, K. R. Martin, R. J. Baer, D. J. Schingoethe, and A. R. Hippen, J. Nutr., 129,
20482054 (1999).
102. W. W. Christie, G. Dobson, and F. D. Gunstone, Lipids, 32, 1231 (1997).
103. T. L. Mounts and H. L. Dutton, Lipids, 5, 9971005 (1970).
104. F. B. Padley, F. D. Gunstone, and F. B. Harwood, in F. D. Gunstone, H. Hashimoto, and
F. B. Padley, eds., The Lipid Handbook, 2nd ed., Chapman and Hall, London, 1994,
pp. 47224.
105. R. W. Johnson and R. W. Daniels, in K. Othmer, ed., Encyclopedia of Chemical
Technology, 4th ed, John Wiley and Sons, New York, 1996, pp. 168177.
106. Specifications and characteristics of Emery oleochemicals, Technical Bulletin A-473,
Henkel Corporation, Emery Group, Cincinnati, Ohio, 1991.
107. E. Fritz and R. W. Johnson, Fatty Acids in Industry, Marcel Dekker, New York, 1988.
108. F. D. Gunstone, M. Kates, and J. L. Harwood, in F. D. Gunstone, J. L. Harwood, and F. B.
Padley, The Lipid Handbook, 2nd ed., Chapman and Hall, London, 1994, pp. 47224.
109. C. Keulemans, Eur. Patent, 1990.
110. M. T. Cleary, S. Kulprathipanja, and R. W. Neuzil (assigned to UOP Inc., Des Plaines,
Illinois), U.S. Patent 4,524,029, 1985.
111. M. T. Cleary and S. Kulprathipanja (assigned to UOP Inc., Des Plaines, Illinois), U.S.
Patent 4,353,838, 1982.
112. H. G. Kirshner (assigned to Colgate-Palmolive Peet Company, Jersey City, New Jersey),
U.S. Patent 2,389,260, 1945.
113. G. O. Burr, U.S. Patent 2,242,230, 1941.
114. J. D. Cawley, U.S. Patent 2,343,644, 1944.
115. A. Struve (assigned to Henkel Kommanditgesellschaft Auf Aktien, Dusseldorf-
Holthausen, Denmark), U.S. Patent 4,381,264 1983.
116. T. F. Bradley, U.S. Patent 2,350,583, 1944.
117. K. E. Krajca (assigned to Sylvachem Corporation, Jacksonville, Florida), U.S. Patent
4,164,505, 1979.
118. T. F. Bradley and D. Richardson, Indust. Engin. Chem., 34, 237242 (1942).
119. P. L. Nichols, R. W. Riemenschneider, and S. F. Herb, J. Amer. Chem. Soc., 27, 329336
(1950).
120. M. J. T. Reaney, M. M. Liu, and N. D. Westcott, in M. P. Yurawecz, M. M. Mossoba,
J. K. Kramer, M. W. Pariza, and G. J. Nelson, eds., Advances in Conjugated Linoleic Acid
Research, AOCS Press, Champaign, Illinois, 1999, pp. 3954.
121. A. Saebo, in J. L. Sebedio, W. W. Christie, and R. Adolf, eds., Advances in Conjugated
Linoleic Acid Research. [5], AOCS Press, Champaign, Illinois, 2003, pp. 7181.
122. T. A. Richardson, Industrial Plastics: Theory and Application, South West Publishing
Co., Cincinnati, Ohio, 1993.
123. M. W. Pariza and Y. M. Yang (assigned to Wisconsin Alumni Research Foundation,
Madison Madison, Wisconsin), U.S. Patent 5,856,149, 1999.
124. D. H. Solomon, The Chemistry of Natural Film Formers, 2nd ed., Robert E. Krieger
Publishing Co., Huntington, New York, 1977, pp. 3374.
125. D. A. Zabolotskij, P. M. Demin, and G. I. Mylagkoya, Russian Patent 2021,252, 1994.
126. M. S. F. Lie Ken Jie, M. K. Pasha, and M. S. Alam, Lipids, 32, 10411044 (1977).
127. N. Sehat, M. P. Yurawecz, J. A. Roach, M. M. Mossoba, J. K. Kramer, and Y. Ku, Lipids,
33, 217221 (1998).
128. N. Sehat, R. Rickert, M. M. Mossoba, J. K. Kramer, M. P. Yurawecz et al., Lipids, 34,
407413 (1999).
129. R. L. Booth, S. K. Kon, W. J. Dann, and T. Moore, Biochem. J., 29, 133137 (1935).
130. M. Czauderna and J. Kowalczyk, J. Chromatogr. B. Biomed. Sci. Appl., 760, 165178
(2001).
131. J. E. Jackson, R. F. Paschke, W. Tolberg, H. M. Boyd, and D. H. Wheeler, J. Amer. Oil
Chem. Soc., 29, 229234 (1952).
132. M. M. Mossoba, M. P. Yurawecz, J. K. Kramer, K. Eulitz, J. Fritsche et al., in
M. P. Yurawecz, M. M. Mossoba, J. K. Kramer, M. Pariza, and G. J. Nelson, eds.,
Advances in Conjugated Linoleic Acid Research, AOCS Press, Champaign, Illinois,
1999, pp. 141151.
133. J. Fritsche, M. M. Mossoba, M. P. Yurawecz, J. A. G. Roach, N. Sehat, and H. Steinhart,
Z. Lebensm. Unters Forsch. A 205, 415418 (1997).
134. R. G. Ackman, C. A. Eaton, J. C. Sipos, and N. F. Crew, J. Can. Inst. Food Sci. Technol.,
14, 103107 (1981).
135. M. P. Yurawecz, J. K. Kramer, and Y. Ku, in M. P. Yurawecz, M. M. Mossoba,
J. K. Kramer, M. Pariza, and G. J. Nelson, eds., Advances in Conjugated Linoleic Acid
136. E. Ostrowska, F. R. Dunshea, M. Muralitharan, and R. F. Cross, Lipids, 35, 11471153
(2000).
137. S. A. Werner, L. O. Luedecke, and T. D. Shultz, J. Agric. Food Chem., 40, 18171821
(1992).
138. J. K. Kramer, V. Fellner, M. E. Dugan, F. D. Sauer, M. M. Mossoba, and M. P. Yurawecz,
Lipids, 32, 12191228 (1997).
139. Y. Park, K. J. Albright, Z. Y. Cai, and M. W. Pariza, J. Agric. Food Chem., 49, 11581164
(2001).
REFERENCES 35
140. K. Eulitz, M. P. Yurawecz, N. Sehat, J. Fritsche, J. A. Roach et al., Lipids, 34, 873877
(1999).
141. V. Spitzer, in M. P. Yurawecz, M. M. Mossoba, M. W. Pariza, and G. J. Nelson, eds.,
Advances in Conjugated Linoleic Acid Research, AOCS Press, Champaign, Illinois,
1999, pp. 110125.
142. G. Janssen and G. Paramentier, Biomed. Mass Spectrom., 5, 439443 (1978).
143. M. J. T. Reaney, L. Y. Dong, and W. G. Taylor, J. Amer. Oil Chem. Soc., 78, 10831086
(2001).
144. M. S. F. Lie Ken Jie, M. K. Pasha, and M. S. Alam, in M. P. Yurawecz, M. M. Mossoba,
145. A. L. Davis, G. P. McNeill, and D. C. Caswell, in M. P. Yurawecz, M. M. Mossoba,
Research, vol. 1, AOCS Press, Champaign, Illinois, 1999, pp. 164179.
146. A. L. Davis, G. P. McNeill, and D. C. Caswell, Chem. Phys. Lipids, 97, 155165 (1999).
147. M. S. Jie, M. K. Pasha, and M. S. Alam, Lipids, 32, 10411044 (1997).
148. B. A. Brice, S. F. Swain, S. F. Herb, P. L. Nicholas, Jr., and R. W. Reimenschneider,
149. J. Ritz and J. Reese (assigned to Hoechst Aktiengesellschaft, Frankfurt, Germany), U.S.
Patent 3,984,444, 1976.
2
Diacylglycerols
Brent D. Flickinger1 and Noburo Matsuo2
1
Archer Daniels Midland Company
Decatur, Illinois
2
Kao Corporation
Tochigi, Japan
1. INTRODUCTION
Diacylglycerols (DAG) are natural components of various edible oils (1, 2). Typi-
cally, the level of DAG in edible oils is below 5% of total oil; however, several
edible oils have levels above 5% (Table 1). Also, DAG have been used as emulsi-
fiers for use in food systems, particularly baked goods, and are approved as such
(3). Human consumption of DAG has been estimated at 3 g per day.
The traditional method of producing DAG is interesterification of triacyl-
glycerols (TAG) with glycerol in the presence of a chemical catalyst at elevated
temperatures (4). Preferred catalysts are alkali such as sodium/potassium hydro-
xide, sodium methoxide, or potassium acetate. Formation of monoacylglycerols
(MAG) and DAG can be controlled to a certain degree by the molar ratio of glycerol-
to-TAG in the initial reaction mixture. Nonetheless, the resulting DAG is part of a
mixture of glyceridic components that present difficulties in obtaining a high-purity
DAG fraction using standard industrial separation processes (i.e., chromatography,
distillation, or winterization) caused by similar chemical and physical properties of
these components.
DAG oil is prepared through the process of enzymatic esterification. Starting
with a blend of soybean and canola oils, fatty acids are prepared then mixed
37
38 DIACYLGLYCEROLS
TABLE 1. Glyceride Content of Various Edible Oils.1
% of Total Oil

Oil TAG DAG MAG Others
Cottonseed 87.0 9.5 0.2 3.3

Palm 93.1 5.8 0.0 1.1
Olive 93.3 5.5 0.2 2.3
Corn 95.8 2.8 0.0 1.4
Safflower 96.0 2.1 0.0 1.9
Lard 97.9 1.3 0.0 0.8
Soybean 97.9 1.0 0.0 1.1
Rapeseed 96.8 0.8 0.1 2.3
1
Data sourced from (1) and (2).
with glycerol. This mixture undergoes esterification using a 1,3-specific lipase from
Rhizomucor miehei, which has been immobilized on a resin bed. After several
finishing processes, including deodorization, the resulting edible oil is obtained.
This oil is bland in flavor and light in color making it suitable for use as a bottled
oil or oil ingredient for various food applications. For more information on DAG oil
manufacturing, physical properties, and food application uses, a book containing
chapters solely devoted to these issues of DAG oil has been published (5).
The nutritional characteristics of DAG oil (80%) have been compared with
dietary TAG of similar fatty acid composition. In particular, the 1,3-DAG isoform
appears to have distinct metabolic characteristics that can impact postprandial lipid
metabolism and use of macronutrients for energy compared with TAG.
The rationale for anticipating metabolic differences is a result of the difference
in metabolism of the digestion product of 1,3-DAG. The body digests DAG oil
yielding monoacylglycerols (MAG) and free fatty acids (FFA) just as observed
with TAG oil. As a result of the significant content of 1,3-DAG, 1-monoacylglycer-
ol (1-MAG) is a major digestion product of DAG oil that is not observed in any
significant amount upon TAG oil digestion. The small intestine typically reassem-
bles digested monoacylglycerols and free fatty acids into TAG beginning with
2-MAG. Previous reports in the literature indicate that providing 1-MAG results
in lower amounts of fat-rich particles appearing in serum following consumption.
This difference in fat metabolism with DAG oil is apparent in fewer fat-rich parti-
cles appearing in the blood after a meal containing DAG oil. As a result, fatty acids
not appearing as chylomicron triacylglycerols must be excreted in the feces or used
by the gut or liver for energy or triacylglycerol storage.
The following review focuses on experimental data supporting different meta-
bolic characteristics of 1,3-DAG or DAG oil containing 1,3-DAG (Table 2). Rele-
vant areas of observed differences between 1,3-DAG/DAG oil and TAG/TAG oil
metabolism include postprandial lipid metabolism and use of macronutrient fuels.
Observations from animal and human experimental data are included.
COMPARISON OF DAG OIL VS. TAG OIL 39
TABLE 2. General Differences in 1,3-DAG or DAG Oil Metabolism (vs. TAG).
Topic Animal Human

s
1. Digestion products 1(3)-MAG vs. 2-MAG.
2. Portal appearance of free Observed change.t
fatty acids
3. Lymph/serum appearance of Decreased amount and rates. Decreased amount and rate.s
chylomicron triglycerides
4. Oxidation and synthesis of fat ChronicIncreased activities ChronicDecreased respira-
of enzymes associated with tory quotient over 36
fat oxidation in liver,s hours.s
increased expression of
mRNA for genes asso-
ciated with fat oxidation in
small intestine,s decreased
respiratory quotient.s
Decreased activities of
enzymes associated with
fat synthesis in liver.s
5. Energy expenditure Enhanced energy expendi- AcuteEnhanceds energy
ture.s expenditure.
ChronicNo change in
energy expenditure
observed.
6. Appetite suppression No observations of appetite Initial appetite suppression.s
suppression.
7. Body weight and body fat Decreased body weight and Greater losses of body weight
body fat.s and body fat.s
s
Statistically significant experimental observations.
t
Trend in experimental observations.
Differences caused by 1,3-DAG/DAG oil in the following discussion are made

under the assumption of being compared with TAG/TAG oil control of similar fatty
acid composition. DAG oil refers to an oil containing 8090% DAG, 1020% TAG,
with the DAG portion containing 70% 1,3-DAG, 30% 1,2-DAG (Figure 1). TAG oil
refers to a conventional edible oil.
2. COMPARISON OF DAG OIL VS. TAG OIL
2.1. Digestion Products

Digestion of 1,3- and 1,2-DAG (70%:30%, respectively, as diolein) results in the
preferential formation of 1(3)-MAG and FFA in rats (6). 1(3)-MAG is 65% of
the monoacylglycerol after 60 minutes of interaction with the small intestine.
This observation has been demonstrated also in mice using labeled fatty acids
incorporated into 1,3-DAG (7). After exposure of labeled 1,3-DAG to the small
intestine in mice, the percentage of 1(3)-MAG from total lipid content increased
to 14.2% compared with 1.5% (p < 0:001), 2-MG decreased to 7.2% compared
40 DIACYLGLYCEROLS
Figure 1. Glyceride composition, structure, and fatty acid profile of DAG oil. (This figure is
available in full color at http://www.mrw.interscience.wiley.com/biofp.)
with 22.3% (p < 0:05), and FFA increased to 59.1% compared with 22.3%
(p < 0:001). This change was reflected to some degree in 1(3) and 1,3 in mucosa
of the small intestine (p < 0:001). Although no corresponding data exists in
humans with regard to discrete digestion products of DAG or DAG oil, fat digestion
is not believed to be significantly different between rodents and humans.
2.2. Portal Appearance of FFA

A mix of 1,3- and 1,2-DAG (70%:30%, respectively, as diolein) was administered
orally to rats followed by collection of blood from the portal vein at predetermined
time periods. Oleic acid tended to be elevated from 0.5 hour to 1 hour after admin-
istration compared with triolein (8).
2.3. Lymph/Serum Appearance of Chylomicrons

Rats were infused continuously with emulsions containing diacylglycerol (66% 1,3-
DAG, 33% 1,2-DAG) by stomach cannulation. DAG significantly retarded the rate
of chylomicron TG formation over a 5-hour period with the difference between
23 hours being significant (6). Using labeled fatty acids, the rate of appearance
of label from DAG in chylomicron triglycerides from lymph of rats was significantly
decreased during the initial 1 hour following administration by approximately 50%

(17% in DAG oil vs. 31%) (9).
Two human studies have focused on serum levels of triglycerides after consump-
tion of test emulsions. Using single doses of 10 g, 20 g, or 44 g of DAG oil, signifi-
cant differences (p < 0:05) were observed in the serum triglycerides for all doses at
6 hours after consumption (10). Additionally, the 44-g dose demonstrated a signif-
icant reduction (23%, p < 0:05) in the rate of postprandial triglycerides, whereas
the 20-g dose showed a significant reduction in serum triglycerides at 4 hours
(50%, p < 0:05). A decrease in VLDL level was also observed at 4 hours during
the adminstration of 20-g dose (p < 0:05). Using a single dose of 30 g/m2 of body
surface area (55 g), significant differences (p < 0:05) were observed in the level
of change in postprandial serum TG at 2 hours, 3 hours, and 8 hours after consump-
tion of test emulsions (11). The suppression of the increase in serum TG after DAG
oil consumption was approximately 50% at hours two and three. A significant
decrease in VLDL was reported subsequently for this study at a recent scientific
meeting (12).
2.4. Oxidation and Synthesis of Fat

Significantly higher activities (p < 0:05) of enzymes involved in fat (or beta) oxi-
dation were observed in liver homogenates of rats fed diets containing varying
levels of DAG oil for 14 days (13). Correspondingly, a significant increase
( p < 0:05) in the oxidation rate of added palmitate was observed. At the same
time, significantly lower activities ( p < 0:05) involved in fatty acid synthesis
(including fatty acid synthase) were observed. In mice, mRNA for genes associated
with beta-oxidation and lipid metabolism as well as small intestinal beta-oxidation
were increased significantly after 10 days on DAG oil-containing diets (14). Signif-
icantly greater beta-oxidation (139%, p < 0:001) was observed by incubating
labeled fatty acid in homogenates of small intestine. Increased oxygen consumption
indicating increased energy expenditure, increased portal vein FFA, increased beta-
oxidation in the liver and small intestine, and increased mRNA expression of acyl
Co-A oxidase (ACO) and uncoupling protein-2 (UCP-2) in the small intestine after
1,3-DAG or DAG oil consumption has been shown in animal models (7, 8, 13, 14).
Enzymes involved in fatty acid oxidation in the liver [carnitine palmitoyltransfer-
ase, acyl-CoA dehydrogenase, acyl-CoA oxidase (ACO), enoyl-CoA hydratase,
3-hydroxyacyl-CoA dehydrogenase, 2,4-dienoyl-CoA reductase, and d3/d2-enoyl-
CoA isomerase] showed increased activity whereas enzymes related to fatty acid
synthesis correspondingly showed decreased activity after 14 days of DAG oil con-
sumption (13). Lower hepatic triacylglycerol content and lower serum cholesterol
were observed along with these differences in liver enzyme activities. Different fat-
digestion products in the gut corresponded with a tendency to use a greater amount
of oxygen after DAG oil consumption in animals (8). Using an animal model for
diet-induced obesity, significantly lower body weight during the lifespan of mice
(C57BL/6J) was observed during ad libitum DAG oil consumption in place of con-
ventional oil (7, 14). In mice fed DAG oil, ACO mRNA levels in the liver were
42 DIACYLGLYCEROLS
increased also significantly, consistent with increased ACO activity, compared with
a conventional triacylglycerol oil (14). Enhanced beta-oxidation in the small intes-
tine has been reported in mice fed DAG oil compared with triacylglycerol oil (7).
Increased beta-oxidation in the small intestine caused by DAG oil consumption was
associated with increased expression of genes involved in beta-oxidation and lipid
metabolism, including ACO, medium-chain acyl-CoA dehydrogenase (MCAD),
liver-fatty acid binding protein (L-FABP), fatty acid transporter (FAT), and UCP-
2. These changes in beta-oxidation and mRNA expression are not apparently con-
sistent with regard to tissue specificity. In mice, these changes occurred solely in the
small intestine, whereas in rats, these changes occurred solely in the liver.
Indirect calorimetry provides the ability to measure the relative contribution of
macronutrients toward energy use. The measurements of expired carbon dioxide
and consumed oxygen are used to calculate respiratory quotient (RQ). An RQ of
1.0 indicates use of carbohydrate solely, an RQ of 0.7 indicates use of fat solely,
whereas an RQ of 0.85 indicates mixed use of macronutrients. Data in rats demon-
strates a significant increase in using fat as an energy substrate following DAG oil
infusion (gastric) as observed by a decreased respiratory quotient value between 3
5 hours post-infusion (15).
In humans, a recently published study reported the influence of DAG oil versus
conventional oil on energy expenditure, energy substrate use, and subjective appe-
tite ratings during a 36-hour stay in a metabolic chamber (16). Using healthy
women (n 12), each subject consumed a defined, eucaloric diet for 3 days prior
to each chamber. During the chamber stay, DAG oil or conventional oil with a simi-
lar fatty acid composition provided 40% of the fat consumed, which was as part of a
defined (55% en from carbohydrate, 15% en from protein, 40% en from fat) euca-
loric diet. For data analysis, differences over the entire 36-hour experimental period
were evaluated. A significant decrease in respiratory quotient (0.006, p < 0:05)
and a significant increase in fat oxidation (4.9 g on day 1, p < 0:004; 4 g on
day 2, p < 0:05) were observed following consumption of a eucaloric diet contain-
ing 12% of total energy from DAG (16% energy from DAG oil) over 36 hours com-
pared with TAG oil (Table 3).
TABLE 3. Summary of Substrate Utilization for Energy

and Fat Oxidation Data after DAG Oil Consumption.1
Parameter DAG Oil TAG Oil
RQ
Total, 36 h (8 A.M. to 8 P.M.) 0.849 0.855
Day 1 (8 A.M. to 8 A.M.) 0.846 0.853
Day 2 (8 P.M. to 8 P.M.) 0.851 0.857
Fat oxidation2 (g/d)
Day 1 60.7 55.8
Day 2 64.6 60.6
1
Adapted from (14).
2
Estimated from Figure 2 of (14).

p < 0:05 vs. TAG oil.

p 0:004 vs. TAG oil.
Serum levels of ketones, as measured by beta-hydroxybutyrate, and free fatty

acids tended to be greater following DAG oil consumption with serum ketone levels
being significantly greater on day 2 (at first draw) after DAG oil consumption
( p < 0:05 vs. TAG oil using post-hoc analysis). Serum ketone levels after DAG
oil consumption remained substantially below the level of serum ketones observed
in ketogenic diets.
2.5. Energy Expenditure

Following ten days of DAG oil consumption, increased mRNA expression of
UCP-2, a protein reported to be involved in thermogenesis, was observed in mouse
small intestine (7). In an earlier rat study, oxygen consumption was observed to
increase, but no RQ or EE was determined (8). Additionally, energy expenditure
over 22 hours was significantly increased in rats on DAG oil diet after an acclima-
tion to DAG oil for 6 days (17). A pronounced difference was observed during the 4
hours after DAG ingestion during the nocturnal cycle.
In humans, the area of energy expenditure continues to be investigated. A sig-
nificant increase in energy expenditure ( p < 0:05) has been measured over the
course of 6 hours after the consumption of 20-g DAG oil (18). In the human study
involving a metabolic chamber, no change was observed in either overall energy
expenditure or individual components of energy expenditure (sleeping metabolic
rate, diet-induced thermogenesis, and activity-based energy expenditure) over the
36-hour experimental period (16).
2.6. Appetite Suppression

Subjective data suggested significant responses toward appetite suppression as esti-
mated by decreased feelings of hunger ( p < 0:01) and appetite ( p < 0:05), estimate
of prospective food intake ( p < 0:01), and desire to eat ( p < 0:05) over the final 12
hours during the 36-hour study (16). Significant differences were not observed after
day 1; however, these significant differences caused by DAG oil consumption
in cumulative day 2 ratings for hunger, appetite, estimate of prospective food
intake, and desire to eat were observed. Decreased cumulative scores inferred
that subjects felt less hunger, had less appetite, indicated lower estimate of pro-
spective food intake, and had less desire to eat after DAG oil consumption
compared with TAG oil. Similar tendencies have been observed after consumption
of medium-chain triacylglycerols with regard to fuel substrate shifts and potential
effects on appetite suppression (1921).
2.7. Body Weight and Body Fat

In a long-term study using mice prone to diet-induced obesity, mice consuming a
DAG-oil diet for five months gained significantly less weight ( p < 0:01) and had
significantly reduced-fat pad weight ( p < 0:01) (14). Levels of mRNA from liver
were greater for genes controlling enzymes used for beta-oxidation. This significant
44 DIACYLGLYCEROLS
TABLE 4. Change in Body Weight Caused by DAG Oil versus TAG Oil Consumption.1
% Change in body weight from baseline2

1 month 2 months 4 months 6 months Overall Trend
DAG oil 2.0 2.4 3.3 3.4 p < 0:025 vs. TAG
TAG oil 0.8 1.5 2.6 2.4
1
Adapted from (22).
2
Estimated from Figure 1.
change in body weight ( p < 0:05) and fat pad weight (epididymal, p < 0:05 and
perirenal, p < 0:01) was observed also in a subsequent study conducted for 8
months (7). Using the same mouse model, 18:3(n-3)-rich DAG oil diets reduced
body weight significantly as well as certain regional areas of visceral fat (22).
Neither of these studies showed decreases in food intake as a result of consuming
DAG oil-containing diets.
Two well-controlled studies have been conducted in humans examining the
impact of DAG oil on body weight and body fat. In subjects consuming approxi-
mately 5% total calories from DAG oil for 16 weeks, significantly greater reduc-
tions in body weight ( p < 0:01) and body fat area ( p < 0:05) were observed
(23). In a larger study, overweight subjects consuming 15% of total energy from
DAG oil for 6 months as part of a mildly hypocaloric diet (500800 kcal/d)
observed a greater extent of body weight ( p < 0:025) and body fat loss
( p < 0:037) (24). In a larger study using overweight individuals consuming 15%
of total energy from test oils for 6 months as part of a diet with mild caloric restric-
tion (500800 kcal/d), subjects consuming DAG oil demonstrated a greater extent
of body weight ( p < 0:025) and body fat loss ( p < 0:037) over the period of diet-
ary intervention when compared with subjects consuming a conventional triacylgly-
cerol oil (24) (Table 4). As importantly, DAG oil functioned as the oil ingredient for
food items that included mayonnaise, crackers, muffins, and instant soups. No
apparent decrease in subject compliance caused by product quality or flavor has
been reported.
3. SUMMARY
From digestion to body weight, differences are observed caused by pure 1,3-DAG,
mixed DAG (70:30 ratio for 1,3: 1,2) or DAG oil compared with TAG or TAG oil.
These differences demonstrate that 1,3-DAG follows a metabolic route resulting in
changes in fatty acid metabolism. Fatty acids appear to a greater extent in portal
circulation and chylomicron triacylglycerols are formed to a smaller extent after
DAG oil consumption. In supporting experiments, beta-oxidation is increased by
the small intestine and liver from animal experiments whereas fatty acid synthesis
is inhibited in the liver after extended consumption of dietary DAG oil diets. Ani-
mal experiments also demonstrate changes in respiratory quotient indicating a shift
SUMMARY 45
toward greater fat use. Animal experiments demonstrate a smaller accumulation of

body weight and body fat over extended periods of consumption.
In humans, DAG oil consumption decreases serum triglycerides (postprandial
and fasting) and enables greater degrees of body fat and body weight loss compared
with conventional oil when used as a dietary aid as part of a healthy diet or caloric
management plan. Studies in Japanese individuals and overweight adult Americans
show similar apparent changes in body weight and body fat with greater differences
in both measures occurring in groups consuming DAG oil.
Relatively mild losses in body fat and body weight loss have occurred following
consumption of DAG oil over a period of several months with an approximate dif-
ference in daily caloric expenditure or intake between 50100 calories. The differ-
ence in greater weight loss over six months observed between DAG oil and
conventional oil could be attributed to a daily caloric deficit of approximately 48
calories (24). Although the difference in weight loss over 4 months observed
between DAG oil and conventional oil could be attributed to a daily caloric deficit
of approximately 90 calories (23). Correspondingly, the data from the metabolic
chamber study did not indicate a significant change in daily energy expenditure
after DAG oil consumption compared with TAG oil in humans (16). Detecting
an energy expenditure difference of 50100 calories over a 2436-hour period
would be difficult in a metabolic chamber setting under the best of conditions.
These investigators also indicate a significant increase in fat oxidation of approxi-
mately 45 g on a daily basis, which corresponds to a shift in using approximately
3545 calories from fat rather than other fuel sources. Apparently, this shift in sub-
strate use may affect overall energy intake (i.e., appetite) to a greater degree than
overall energy expenditure. Interestingly, individuals subjected to acute overfeeding
of 1-MAG did not show any differences with regard to energy intake and appetite
regulation compared with TAG (25, 26), whereas jejunal infusions of free fatty
acids have shown the ability to reduce food intake and body weight in rats (27).
These studies may indicate that the combination of 1-MAG and FFA, potentially
more so the FFA, caused by the digestion of 1,3-DAG results in observed differ-
ences in energy expenditure and appetite regulation.
From a mechanistic viewpoint, fatty acid oxidation as a metabolic control for
energy intake appears to be an important relationship (2830). Products of fatty
acid oxidation have been implicated in playing a role in control of food intake.
Historically, ketones have received considerable attention, whereas glycerol and
free fatty acids have received less attention. Using ketogenic diets (defined as
severe CHO restriction) for weight loss has been long espoused with an underlying
assumption that elevated serum ketone levels provide a certain degree of appetite
suppression. This effect of appetite suppression by elevated serum ketones remains
to be conclusively documented.
Effects of compounds that inhibit enzymes of fat metabolism have been reported
in recent literature. Changes in fatty acid oxidation have been inversely correlated
to changes in food intake using rodent models as well as human subjects. After
administration of compounds that directly inhibit fatty acid oxidation or inhibit
fatty acid synthesis, significant increases or decreases in food intake, respectively,
46 DIACYLGLYCEROLS
as well as changes in body weight over extended durations have been observed (31
36). Etomoxir, a carnitine palmitoyl transferase I inhibitor, causes decreased mito-
chondrial oxidation of long-chain fatty acids and increases insulin resistance in
mice as a result of intramyocellular lipid accumulation over prolonged administra-
tion (37). Limited studies with acute administration of etomoxir in human subjects
show mixed evidence regarding an increase in food intake (38, 39). C75, a fatty
acid synthase inhibitor, results in decreased fatty acid synthesis and decreased
food intake and body weight in mice through its proposed modulation of hypotha-
lamic neuronal activity and neuropeptide Y levels (40, 41). Studies in human sub-
jects administered C75 have yet to appear in the scientific literature.
Use of fatty acids for energy during the postprandial period after DAG oil con-
sumption has been reported to occur in the liver or small intestine via beta-oxida-
tion based on rodent studies (13, 14, 42). Regulation of hepatic fat oxidation is
believed to be under the control of hepatic glucose metabolism, rather than dietary
fat intake, which implies that the status of glycogen stores and carbohydrate meta-
bolism may be more important parameters in controlling total energy balance and
fat mass than fat intake (43, 44). This mechanism suggests that increased postpran-
dial fatty acid oxidation may potentially lead to smaller glycogen stores, which
have been observed to correlate with lower body fat stores and decreased effort
to replenish the loss of body fat.
REFERENCES
1. A. A. Abdel-Nabey, A. A. Y. Shehata, M. H. Ragab, and J. B. Rossell, Riv. Ital. Sostanze

Grasse, 69, 443447 (1992).
2. R. P. DAlonzo, W. J. Kozarek, and R. L. Wade, J. Amer. Oil Chem. Soc., 59, 292295
(1982).
3. Code of Federal Regulations, Title 21 184.1505, U.S. Government Printing Office,
4. N. O. V. Sonntag, J. Amer. Oil Chem. Soc., 59, 795A802A (1982).
5. Y. Katsuragi, T. Yasukawa, N. Matsuo, B. D. Flickinger, I. Tokimitsu, and M. G. Matlock,
Diacylglycerol Oil, AOCS Press, Champaign, Illinois, 2004.
6. M. Murata, K. Hara, and T. Ide, Biosci. Biotech. Biochem., 58, 14161419 (1994).
7. T. Murase, M. Aoki, T. Wakisaka, T. Hase, and I. Tokimitsu, J. Lipid Res., 43, 13121319
(2002).
8. H. Watanabe, K. Onizawa, and H. Taguchi et al., J. Japan Oil Chem. Soc., 46, 301308
(1997).
9. T. Yanagita, Digestion and Absorption of Diacylglycerol, 5th International Congress on
Essential Fatty Acids and Eicosanoids, Diacylglycerol and Human Health Session, Taipei,
Taiwan, 2002.
10. H. Taguchi, H. Watanabe, and K. Onizawa et al., J. Amer. Coll. Nutr., 19, 789796 (2000).
11. N. Tada, H. Watanabe, N. Matsuo, I. Tokimitsu, and M. Okazaki, Clin. Chem. Acta, 311,
109117 (2001).
REFERENCES 47
12. N. Tada, Effect of Oral Diacylglycerol Loading on Postprandial Hyperlipidemia in Men,

5th International Congress on Essential Fatty Acids and Eicosanoids, Diacylglycerol and
Human Health Session, Taipei, Taiwan, 2002.
13. M. Murata, T. Ide, and K. Hara, Br. J. Nutr., 77, 107121 (1997).
14. T. Murase, T. Mizuno, T. Omachi et al., J. Lipid Res., 42, 372378 (2001).
15. S. Kimura, Studies on Energy Metabolism of Diacylglycerol Using Expiration Analysis,
The Japanese Society of Nutrition and Food Science, 2002.
16. M. M. Kamphuis, D. J. Mela, and M. S. Westerterp-Plantenga, Amer. J. Clin. Nutr., 77,
11331139 (2003).
17. H. Watanabe, Effect of dietary DAG on energy expenditure in rats, in press.
18. K. Mori, Effects of diacylglycerol diet on energy expenditure, in press.
19. V. Van Wymelbeke, J. Louis-Sylvestre, and M. Fantino, Amer. J. Clin. Nutr., 74, 620630
(2001).
20. M. Krotkiewski, Int. J. Obes. Relat. Metab. Disord., 25, 13931400 (2001).
21. R. J. Stubbs and C. G. Harbron, Int. J. Obes. Relat. Metab. Disord., 20, 435444
(1996).
22. T. Hase, T. Mizuno, and K. Onizawa et al., J. Oleo. Sci., 50, 701710 (2001).
23. T. Nagao, H. Watanabe, and N. Goto et al., J. Nutr., 130, 792797 (2000).
24. K. C. Maki, M. H. Davidson, and R. Tsushima et al., Amer. J. Clin. Nutr., 76, 12301236
(2002).
25. A. M. Johnstone, L. M. Ryan, C. A. Reid, and R. J. Stubbs, Eur. J. Clin. Nutr., 52, 603609
(1998).
26. A. M. Johnstone, L. M. Ryan, C. A. Reid, and R. J. Stubbs, Eur. J. Clin. Nutr., 52, 610618
(1998).
27. J. E. Cox, W. J. Tyler, and A. Randich et al., Amer. J. Physiol. Regul. Integr. Comp.
Physiol., 278, R604R610 (2000).
28. E. Scharrer, Nutrition, 15, 704714 (1999).
29. M. I. Friedman, Amer J. Clin. Nutr., 67, 513S518S (1998).
30. A. Tremblay, Int. J. Obes. Relat. Metab. Disord., 16, 953957 (1992).
31. E. Del Prete, T. A. Lutz, J. Althaus, and E. Scharrer, Physiol. Behav., 63, 751754
(1998).
32. M. I. Friedman, R. B. Harris, H. Ji, I. Ramirez, and M. G. Tordoff, Amer. J. Physiol., 276,
R1046R1053 (1999).
33. N. E. Rawson, P. M. Ulrich, and M. I. Friedman, Amer. J. Physiol., 271, R144R148
(1996).
34. J. N. Thupari, L. E. Landree, G. V. Ronnett, and F. P. Kuhajda, Proc. Natl. Acad. Sci. USA,
99, 94989502 (2002).
35. T. M. Loftus, D. E. Jaworsky, and G. L. Frehywot et al., Science, 288, 23792381 (2000).
36. J. Triscari and A. C. Sullivan, Int. J. Obes., 8 Suppl 1, 227239 (1984).
37. R. L. Dobbins, L. S. Szczepaniak, B. Bentley, V. Esser, J. Myhill, and J. D. McGarry,
Diabetes, 50, 123130 (2001).
38. V. B. Hinderling, P. Schrauwen, W. Langhans, M. S. Westerterp-Plantenga, Amer. J. Clin.
Nutr., 76, 141147 (2002).
39. A. Kahler, M. Zimmermann, and W. Langhans, Nutrition, 15, 819828 (1999).
48 DIACYLGLYCEROLS
40. T. Shimokawa, M. V. Kumar, and M. D. Lane, Proc. Natl. Acad. Sci. USA, 99, 6671
(2002).
41. S. Gao and M. D. Lane, Proc. Natl. Acad. Sci. USA, 100, 56285633 (2003).
42. T. Murase, M. Aoki, T. Wakisaka, T. Hase, and I. Tokimitsu, J. Lipid Res., 43, 13121319
(2002).
43. J. P. Flatt, Amer. J. Clin. Nutr., 61, 952S959S (1995).
44. J. P. Flatt, Int. J. Obes. Relat. Metab. Disord., 20 Suppl 2, S1S11 (1996).
3
Citrus Oils and Essences
1. INTRODUCTION
Citrus fruits are among the most popular fruits nowadays and have a very long his-
tory of production and use. However, within the past century, industrial technolo-
gies began to develop in order to convert citrus fruits into commercial products (1).
Each year, millions of tons of citrus fruits are delivered to factories for processing
and juice production. Historically, the oldest citrus product is the oil. In ancient
Sicily, where early Italian citrus industry had just been introduced, lemons were
primarily grown for production of lemon oil, and juice was treated as a waste pro-
duct until its later use for citric acid recovery. The early use of lemon and orange
oils was mainly in perfumery and pharmaceuticals (1). With rapid development of
science and technology, more areas of use of citrus oils were found, for which more
detailed information on chemical composition and properties were required. The
modern perfume and flavor industries have benefited from further research on citrus
peel oil and essence. Besides, the yield of citrus seed oil has increased since citrus
seeds were discovered as a new source of edible oil.
Citrus has proven to be a very good option for the oil and essence production.
The genus citrus, according to Tanakas system, has been divided into eight groups,
Papeda, Limonellus, Citruphorum, Cephalocitrus, Aurantium, Osmocitrus, Acrumen,
and Pseudofortunella, with a number of species within each group and a larger
49
50 CITRUS OILS AND ESSENCES
number of fine-quality hybrids as well (2). In this chapter, the chemical composi-
tion, properties, and uses of citrus oils and essences are provided.
2. OIL EXTRACTION
2.1. Extraction of Citrus Peel Oil

Citrus peel oils of very complex composition are contained in oval, balloon-shaped
oil sacs, or vesicles, located in the outer rind, or flavedo, of the fruit (3). The oil is
usually extracted by mechanical separation or hydrodistillation. The five main types
of citrus from which peel oils are recovered are orange, grapefruit, tangerine,
lemon, and lime (4). Mechanical separation, known as cold-pressing of peel oils,
does not use heat in order to avoid loss of volatile components. Swisher and
Swisher (1) described three general commercial methods that are widely used in
citrus industry to extract crude oils from fruit peels:
(1) Oil recovery from peel after juice extraction

(2) Simultaneous extraction of juice and oil emulsion from whole fruit
(3) Recovery of oil from the peel flavedo after removal from the whole fruit by
abrasion or shaving
Citrus peel oils for small-scale use may be obtained by hand-pressing. Fruits are
sliced, and mesocarp and albedo layers are peeled from the flavedo before hand-
pressing. Peel oils are collected in brine solution on ice, and oil extract is centri-
fuged at 4 C. Afterwards, the supernatant is dehydrated with anhydrous sodium
sulfate and filtered (5). The total final oil extract is about 1% of the flavedo by
weight (6).
Citrus peel oils other than cold-pressed oils have a lower price in the market-
place and are known as distilled oil, which is recovered from peels by steam dis-
tillation. This oil possesses an odor and flavor that is generally inferior to that of the
cold-pressed oil (1). The final oil extract is a liquid with its color varying depending
on the species of the fruit. For instance, bergamot oil is green to greenish-yellow,
grapefruit oil is greenish-yellow, lemon oil is pale yellow to pale greenish-yellow,
lime oil is colorless to pale yellow, mandarin oil is greenish-yellow to reddish-
orange, bitter orange oil is pale yellow to yellowish-brown, and sweet orange oil
is yellow to reddish-yellow (7). In addition, different citrus peel oils have different
physicochemical properties. Table 1 shows the comparison of physicochemical
properties of two grapefruit species.
2.2. Extraction of Citrus Seed Oils

Citrus seeds are regarded as a new source of edible oil, especially in some devel-
oping countries where nutritionists and food chemists have been searching for
OIL EXTRACTION 51
TABLE 1. Average Values for Physical and Chemical Properties of Cold-pressed Red
and White Grapefruit Peel Oils.
White Grapefruit Oil Red Grapefruit Oil
Specific gravity, 25 C/25 C 0.8534 0.8522

Refractive index, n20D 1.4759 1.4766
Refractive index,10% distillate,n20D 1.4718 1.4718
Optical rotation, a[25D] 92.67 91.07
Optical rotation, 10% distillate,a[25D] 97.04 96.84
Evaporation residue,% 6.32 7.12
Aldehyde content,% 1.56 1.38
From Wolford et al., 1971 (8).
economical sources of material with oil of high industrial potential to alleviate the
shortage of oil (9). Citrus seeds contain about 36% oil, which can be recovered
from seeds by crushing and solvent extraction.
After juice extraction, seeds are separated from the waste product by a paddle-
type finisher. They are crushed after washing and drying, and the oil is extracted. In
some cases, a commercial solvent extraction of press-cake, using n-hexane or pet-
roleum ether, is used to improve oil recovery (1). Unrefined citrus seed oil is pale
to light yellow in color and may possess a bland or almond-like aroma (1). The
physicochemical characteristics of citrus seed oils, namely the refractive index,
specific gravity, melting point, color, and viscosity, vary slightly from each other
(Table 2) (9). Citrus seed oils can be used for cooking after refining. However, crude
oils are used for preparation of detergents and soaps.
TABLE 2. Physicochemical Characteristics of Different Citrus Seed Oils.
Property Citron Orange Mandarin Mixed

Refractive index (25 C) 1.4681 0.001 1.4684 0.002 1.4672 0.001 1.4682 0.001
Density (25 C) [g/cm3] 0.884 0.01 0.914 0.06 0.962 0.09 0.910 0.08
Melting point [ C] 7 7 7 7
Viscosity [101Pa s] 0.05 0.07 0.08 0.07
Color 2.6R 3.1R 7.4R/0.2B 4.0R
Acid value [mg KOH/g oil] 0.953 0.08 0.673 0.09 1.120 0.09 0.762 0.06
Saponification value 189.5 1.41 190.2 1.87 187.2 1.73 188.3 1.76
[mg KOH/g oil]
Ester value [mg KOH/g oil] 188.5 1.40 189.5 1.72 186.8 1.70 187.5 1.41
Peroxide value 5.95 0.48 6.37 0.51 5.90 0.62 5.98 0.57
[meq O2 /kg oil]
Iodine value(Hanus) 96.23 1.34 102.57 1.57 91.54 1.09 99.25 1.07
[mg I2 /g oil]
From El-Adawy et al., 1999 (9).

2.3. Extraction of Citrus Essences

Citrus essences are distilled aqueous solutions of the more volatile components
from the corresponding citrus juices, as defined by Shaw (10). Commercially,
they are added to concentrated citrus juices to impart fresh fruit flavor that may
be lost during the concentration process. Essence may be collected from fresh juice
either by partial distillation prior to juice evaporation or by condensation of vola-
tiles from the early stages of evaporation (11). Two phases, namely, aqueous
essence and essence oil, are obtained during recovery.
3.1. Chemical Composition of Citrus Peel Oils

Different methods have been adopted to analyze the chemical composition of citrus
oils as well as their odor activities, including gas chromatography-mass spectrome-
try (GC-MS) (12, 13), gas chromatography-olfactometry (GCO) (6), nuclear mag-
netic resonance (NMR) (14), near infrared (NIR) (15), high-performance liquid
chromatography (HPLC) (16), thin layer chromatography (TLC) (9), and other
chromatographic procedures. Improvements in the available techniques has given
rise to more precise and detailed data for qualitative and quantitative determination
of citrus oils. Accordingly, 55 volatile components in lemon oil, 62 in lime oil (14),
79 in mandarin oil (17), 88 in tangerine oil (18), and some more components in
hybrid citrus oils have been identified (5).
Citrus peel oils are characterized by complex mixtures containing mostly ter-
penes as well as oxygen-containing compounds (19). Volatile components account
for more than 90% of the oil mass, whereas nonvolatile residues are present in very
small amounts. For most citrus fruits, peel oils consist almost exclusively in the
form of monoterpenes, sesquiterpenes, and other aliphatic hydrocarbons. The
hydrocarbon contents in Vietnamese pummelo, orange, tangerine, and lime, respec-
tively, exceeded 98.7%, 97.6%, 98.6%, and 95.4% (20). Among these, monoter-
penes were dominant. Limonene, a monoterpene often used as a functional index
of ripeness, was the major component of all citrus peel oils, followed by b-pinene in
lemon oil (21), g-terpenene in tangerine and lime oils (14), myrcene in sweet
orange oil, and p-cymene in mandarin oil (19). The compounds a-pinene, sabinene,
b-phellandrene, and terpenolene were present at lower levels. A few sesquiterpenes
were also found in very small amounts but made appreciable contributions to flavor
and odor; these included trans-b-farnesene, trans-a-bergamotene, trans-caryophyl-
lene, germacrene A to D, and b-bisobolene (14, 22).
Although terpene hydrocarbons, especially monoterpenes, are the most abundant
constituents of citrus peels oil, they serve only as a flavor carrier and contribute
little to flavor on their own (23). These terpenoid hydrocarbons are usually removed
by deterpenation in order to increase the concentration of flavor and fragrance com-
pounds. Furthermore, unsaturated hydrocarbons (terpenes) are unstable to heat and
light, and may oxidize rapidly to produce undesirable off-flavor compounds that
adversely affect the desirable aroma of products (24). Therefore, concentrated
and deterpenated oils have become popular in the citrus oil market.
Oxygenated compounds, mainly oxygenated terpenes, rather than terpene hydro-
carbons, have been found to be responsible for the characteristic odor and flavor of
citrus fruits, although they occur in relatively small amounts. When the hydrocar-
bon fraction is removed from the oil, the oxygenated fraction becomes more odor-
ous due to a higher concentration (25). Characterized by quantitative abundance in
aldehydes and a relatively wide variety of alcohols, oxygenated fraction includes
aldehydes, alcohols, ketones, esters, oxides, acids, and trace amounts of fugenol
methyl ether (5). Geranial and neral are the major aldehydes, both of which account
for the fresh floral and citrus-like character of lemon and lime oils (21, 26). Citro-
nellal has a green-citrusy odor rather than a sweet and fruity odor (5). In addition,
many simple aliphatic aldehydes, such as octanal, decanal, and dodecanal, impart a
characteristic aroma to citrus peel oils (17, 26). Among alcohols, monoterpene
alcohols such as linalool, followed by octanol and aterpeneol, are most predomi-
nant (5). Nerol and geraniol are also found in high levels. Among these, linalool and
octanol are regarded as the most odor-active compounds in such citrus as Hyuga-
natsu (Citrus tamurana) (27). Ketones, esters, oxides, and acids are less repre-
sented, but make appreciable contributions to flavor. Nootkatone is an important
flavor compound of grapefruit oils (28); neryl acetate, geranyl acetate, and bornyl
acetate have been used as sweeteners, and linalool oxide provides a powerful sweet
odor (5). Flavor dilution (FD) factor is employed to express the odor potency or
intensity of volatile components in citrus peel oils (Table 3). FD factor is defined
as the ratio of concentration of a compound in the initial concentration to that in the
most diluted concentration in which the odor could be detected by GCO (29).
Compositions of volatiles in different orange oils are shown in Tables 4 and 5. It
is evident that most of the constituents belong to the terpene family and may be
arranged into two groups, terpene hydrocarbons (terpenes and sesquiterpenes)
and oxygenated terpene products (21). Aside from the volatile components, there
are small amounts (215%) of nonvolatile residues in citrus peel oils that possess
antioxidative property; these include coumarins, psoralens, and polymethoxylated
flavones (3034).
3.2. Chemical Composition of Citrus Seed Oils

Generally, lipids in citrus seed oil may be classified as nonpolar, nonionic polar, and
ionic polar. Nonplolar lipids are mainly composed of triacylglycerols; nonionic
polar lipids refer to sugar-containing lipids, including glycosylacylglycerols, sterol
glucosides, and sphingosine-containing lipids. Meanwhile, ionic polar lipids con-
tain phospho, sulfo, amino, or carboxyl reactive groups (35).
Analyzed by thin layer chromatography, crude citrus seed oils are reported to
have eight classes of chemical constituents: triacylglycerols, free fatty acids, diacyl-
glycerols, monoacylglycerols, sterols, phospholipids, alcohols, and hydrocarbons
TABLE 3. Potent Odorants (FD 2) Identified in the Peel Oil of Clementines

(Citrus Reticulata Blanco cv. Clementine)
No. Odorant Odor Quality FD Factor
1 a-pinene pinetree-like 1024

2 (Z)-hex-3-enal grassy 128
3 myrcene geranium leaf-like 1024
4 limonene mint-, citrus-like 512
5 octanal citrus-like, fresh 1024
6 oct-1-en-3-one mushroom-like 2
7 (Z)-octa-1,5-dien-3-one geranium-leaf-like 2
8 nonanal green, citrus-like 16
9 (Z)-non-6-enal green, fresh, cucumber 64
10 citronellal citrus-like, fresh 64
11 decanal green, fresh 256
12 (E)-non-2-enal fatty, green 128
13 (Z)-dec-4-enal green, musty 8
14 linalol flowery 16384
15 octan-1-ol citrus-like, soapy 8
16 unknown waxy 16
17 undecanal citrus-like, soapy 4
18 unknown metallic 16
19 (E,Z)-nona-2,4-dienal deep-fried 8
20 unknown flowery, citrus-like 16
21 (E,E)-nona-2,4-dienal deep-fried 512
22 dodecanal green coriander 16
23 carvone mint-like 256
24 (E)-undec-2-enal green coriander 16
25 unknown soapy 8
26 unknown flowery,soapy 64
27 unknown flowery 64
28 (E,E)-deca-2,4-dienal deep-fried 4096
29 (E)-dodec-2-enal green coriander 128
30 unknown fatty, cucumber-like 64
31 (E)-tridec-2-enal green, metallic 8
32 unknown fruity, pinetree-like 8
33 tr-4,5-epoxy-(E)-dec-2-enal metallic 128
35 unknown fruity 8
36 unknown fruity 8
37 sotolone spicy 16
3a, 4, 5, 7a-tetrahydro-3,6-
38 dimethylbenzofuran-2(3H)-one sweet, cocos 4096
39 b-sinensal metallic, green, sweaty 64
41 a-sinensal metallic, green, sweaty 64
42 vanillin vanilla-like 16
*
FD factor (Flavor dilution factor): The ratio of concentration of a compound in the initial concentration to that in
the most diluted concentration in which the odor could be detected by GCO.
From Buettner et al., 2003 (29).
TABLE 4. Major Components of Cold-pressed Navel Orange Oil.
Compound Cold-Pressed Oil (%) Compound Cold-Pressed Oil (%)
monoterpenes 97.52 alcohols 0.58

limonene 94.74 linalool 0.4
b-myrcene 1.66 n-octanol 0.12
a-pinene 0.46 a-terpineol 0.06
sabinene 0.46 oxidized limonene 0.33
carene-3 0.09 limonene oxide (cis) 0.08
b-pinene 0.03 carvone 0.08
a-phellandrene 0.03 limonene oxide (trans) 0.05
ocimene 0.03 carveol (cis) 0.05
terpinolene 0.02 limonene oxide (trans) 0.05
a-thujene <0.01 dehydro carveol (iso) 0.01
a-terpene <0.01 esters 0.02
p-cymene <0.01 neryl acetate 0.01
g-terpinene <0.01 geranyl acetate 0.01
aliphatic aldehydes 0.75 sesquiterpenes 0.13
decanal 0.38 valencene 0.03
n-octanal 0.22 b-farnesene 0.02
dodecanal 0.11 a-farnesene 0.02
n-nonanal 0.04 d-cadinene 0.02
terpen aldehydes 0.23 a-copaene 0.01
perillaldehyde 0.06 b-cubebene 0.01
citronellal 0.05 b-caryophyllene 0.01
a-sinensal 0.04 germacrene-D 0.01
geranial 0.03
b-sinensal 0.03
neral 0.02
From Shen et al., 2002 (23).
(Table 6). Triacylglycerols are the major oil class in all citrus seed oils, followed by
free fatty acids and then diacylglycerols. The presence of partial acylglycerols and
free fatty acids is due to partial enzymatic hydrolysis of reserve triacylglycerols
during seed storage (9).
The composition of citrus seed oil varies with species and storage conditions.
For instance, Habib et al. (36) reported that mandarin seed oil is very high in its
triacylglycerol content whereas citron oil has a large amount of free fatty acids.
With respect to compositional patterns, lime seed oil is similar, in its degree of
unsaturation, to soybean oil, and orange oil is similar to cottonseed oil. In general,
citrus seed oil has a larger amount of volatile fatty acids than other edible oils (36).
Crude citrus seed oils need to be refined before use as edible oils. Only triacyl-
glycerols, diacylglycerols, and polar lipids remain after degumming, refining,
bleaching, and deodorization. However, trace amounts of phosphatides (lecithin)
and plant sterols may also remain in the oil (37).
Citrus seed oil, as a potential edible oil, serves as a good source for essential
fatty acids. More than 60 fatty acids have been found in various citrus seed
oils, among which unsaturated fatty acids are present in a high amount (>65%)
TABLE 5. Composition of Oils from Five Different Orange Species.
% Total Area
Component Bergamot Bitter Orange Mandarin Sweet Orange Tangerine
limonene 36.54 93.42 68.8 96.1 88.15

b-pinene 8.63 0.91 2.15 0.52 0.86
p-cymene 7.24 1.66 16.06 0.06 2.75
g-terpinene 2.2 5.85 3.11
a-pinene 1.47 0.94 2.87 0.69 1.27
myrcene 0.95 2.05 1.66 1.77 1.78
a-bergamotene 0.39
a-thujene 0.36 0.9 0.18
b-caryophyllene 0.29 0.05 0.03 0.03 0.01
terpinolene 0.27 0.51 0.01 0.08
linalool 8.6 0.1 0.06 0.18 0.57
neral 0.07 0.04 0.05
geranial 0.05 0.02 0.02 0.03 0.02
octanal 0.03 0.06 0.04
linalyl acetate 31.73 0.16
geranyl acetate 0.52 0.11
sum of 16 compounds 99.3 99.5 98.9 99.5 98.8
% hydrocarbons 58.73 99.13 99.22 99.29 98.68
% oxygenated compounds 41.27 0.87 0.78 0.71 1.32
From Veriotti et al., 2002 (19).
(35, 38). Linoleic (>30%), oleic (>18%), and linolenic (212%) acids are the most
predominant unsaturated fatty acids present (9). Lemon, lime, and citron oils con-
tain the highest amount of linolenic acid. In addition, very small amounts of myr-
istoleic acid (C14:1) in polar lipids fraction, myristoleic (C14:1) and palmitoleic
(C16:1) acids in diacylglycerols fraction, and myristoleic (C14:1) and eicosaenoic
(C20:1) acids in triacylglycerols fraction of citrus seed oil were also identified (39).
Saturated fatty acids were less abundant than their unsaturated counterparts in
citrus seed oils and consisted mainly of palmitic (C16:0) and stearic (C18:0) acids
TABLE 6. Composition of Different Citrus Seed Oils (%).
Oil Class Citron Orange Mandarin Mixed
Triacylglycerols 66.8 65.4 68.4 68.0

Free fatty acids 14.5 13.4 11.7 12.8
Diacylglycerols 12.1 12.0 10.5 11.3
Monoacylglycerols 2.49 1.97 2.97 2.51
Sterols 2.18 3.52 3.27 3.14
Phospholipids 1.96 3.66 2.64 2.23
Hydrocarbons Tr. Tr. Tr. Tr.
Alcohols Tr. Tr. Tr. Tr.
*
Tr. Traces.
TABLE 7. Fatty Acids Composition of Different Citrus Seed Oils.
Fatty Acid [%] Citron Orange Mandarin Mixed
Lauric (C12:0) 0.39 0.36 0.65 0.37

Myristic (C14:0) 0.43 0.44 0.61 0.46
Palmitic (C16:0) 29.52 24.73 28.12 28.54
Stearic (C18:0) 4.32 5.27 4.34 4.37
Oleic (C18:1) 22.25 26.00 24.89 24.53
Linoleic (C18:2) 33.21 38.44 38.26 38.25
Linolenic (C18:3) 9.56 4.58 2.58 3.11
Arachidic (C20:0) 0.32 0.18 0.55 0.37
Total saturates 34.98 30.98 34.27 34.11
Total unsaturates 65.02 69.02 65.73 65.89
Total essentical fatty acids 42.77 43.02 40.84 41.36
at 2231% and 16%, respectively. Arachidic (C20:0), lauric (C12:0) and

myristic (C14:0) acids were found in trace amounts compared with other fatty acids
(9, 35, 40).
Table 7 shows the fatty acid composition of different citrus seed oils. The ratio of
unsaturated to saturated fatty acids is approximately 2:1 (9), although this ratio was
reported to be in the range of 35:1 by Nagy (35). Generally, different varieties,
cultivar, location, storage condition, and harvesting time of citrus fruit may lead
to this variation. Table 8 shows the content of the six major fatty acids in different
citrus seed oils; these are linoleic (C18:2), palmitic (C16:0), oleic (C18:1), linolenic
(C18:3), stearic (C18:0), and palmitoleic (C16:1) acids.
3.3. Chemical Composition of Citrus Essence

Distillation of citrus juices yields two volatile fractions, namely, aqueous essences
and essence oils that are separated from each other by condensation of the distillate
(7). Aqueous essence, the bottom layer of the condensate is comprised of organic
acids, alcohols, aldehydes, esters, hydrocarbons, ketones, hydrogen sulfide, and
oxides (10). Considering many components found in both cold-pressed peel oil and
aqueous essence, essence oil has a flavor similar to that of the combined peel oil
and aqueous essence (10). However, essence oil usually contains a larger amount of
TABLE 8. Major Fatty Acids in Citrus Seed Oils (%).
Seed Type Palmitic Palmitoleic Stearic Oleic Linoleic Linolenic
Oranges 2631 0.1 35 2428 3537 24

Grapefruit 2636 0.10.3 14 1825 3241 36
Mandarins 2230 0.11.0 25 2025 3745 35
Lemons 1824 0.10.3 24 2634 3138 612
Limes 2429 0.10.5 35 2022 3740 611
From, Nagy, 1977 (35).

aliphatic ethyl esters (e.g., ethyl butyrate in orange essence oil) compared with the
peel oil (41). Thus, its aroma resembles that of the corresponding juice more than
that of the peel oil (7). In general, citrus flavor results from a complex mixture of
components in the appropriate proportions, as described by Monshonas et al. (42).
4. STORAGE OF CITRUS OILS
4.1. Changes of Composition During Storage

Citrus peel oils have been used widely in beverages, cosmetics, pharmaceuticals,
and perfumery industry, whereas seed oils are used in cooking and for treatment
of leather and textile. The quality, freshness, and uniqueness of citrus oils are major
considerations pertaining to their value and applications (43). However, large
amounts of volatile components, as well as unsaturated compounds, render the
oils unstable and prone to change with time and storage conditions.
Most of the qualitative changes of citrus peel oil during storage occur due to
evaporation, oxidation, polymerization, rearrangement, and cyclization of some
labile constituents in the presence of heat, light, oxygen, moisture, and catalysts
resulting in the loss of volatile components and formation of off-flavor artifacts
(43, 44). This major deterioration in citrus peel oil occurs when stored at 20 C,
with a notable decrease in monoterpenes and an increase in oxygenated compounds
(33). Monoterpenes may undergo oxidation and, hence, formation of oxygenated
terpenes, rearrangement and cyclization into sesquiterpenes, polymerization into
artifacts, and evaporation that causes loss of these components as well. As reported
by Njoroge (33), the total percentage of monoterpenes decreased progressively
from 98.0 to 66.4 in Daidai (Citrus aurantium) peel oil after 12 months of storage
at 20 C. The prominently decreased components were b-myrcene, g-terpinene, and
limonene. However, the concentration of p-cymene was increased, indicating loss
of freshness of the oils, which might be explained by dehydrogenation and rearran-
gement of a- and g-terpinene, and hydrogenation and double-bond rearrangement
of limonene as well (43, 45).
The content of oxygenated compounds increased considerably during storage
(43). Monoterpene alcohols, as oxidized products of monoterpenes, showed the
most significant increase. In contrast, linalool (both cis and trans) was found to
decrease at 20 C after 12 months of storage (44, 45). Esters, such as linalyl and
geranyl acetates, and most oxides also increased significantly. Besides, artifacts
were formed in citrus peel oil upon prolonged storage at higher temperatures in
the forms of alcohol, carbonyl compound, ester, epoxide, and hydrocarbon (33).
Sesquiterpene alcohols were the dominant artefacts formed, as found in Yuzu
(Citrus junos Tanaka) peel oil, with spathulenol being the main constituent. Spathu-
lenol could be produced directly from bicyclogermacrene on air oxidation, or
through isomerization of bicyclogermacrene to aromadendrene and allo-aromaden-
drene, and the subsequent oxidation of these isomeric tricyclic sesquiterpenes (43).
The relative compositional change of Yuzu is shown in Table 9.
STORAGE OF CITRUS OILS 59
TABLE 9. Relative Compositional Changes in Yuzu (C. junos Tanaka) Cold-pressed Oil
During a 12-Day Storage in 20 C.
Relative Concentration (%)
Compound Fresh Oil 1 3 6 9 12
monoterpene hydrocarbons
a-pinene 1.84 1.95 1.94 1.61 1.45 0.66
b-pinene 0.69 0.74 0.74 0.75 0.59 0.3
sabinene 0.24 0.24 0.25 0.19 0.17 0.08
myrcene 2.14 2.02 1.84 1.36 1.05 0.56
a-terpinene 0.15 0.12 0.12 0.04 0.02 Tr.
limonene 78.13 77.06 76.54 66.7 51.87 31.49
g-terpinene 9.32 10.07 9.62 8.01 2.23 1.47
p-cymene 0.4 0.61 0.68 4.65 5.19 3.05
terpinolene 0.39 0.41 0.4 0.5 0.09 0.05
a-p-dimethylstyrene 0.04 0.04 0.04 0.08 0.16 0.06
p-mentha-1,4,8-triene
total 93.34 93.26 92.17 83.89 62.82 37.72
monoterpene alcohols
linalool 1.79 1.95 1.97 5.43 7.08 3.67
a-terpineol 0.1 0.13 0.13 0.49 1.24 1.36
(Z)-carveol 0.07 0.06 0.06 0.24 0.36 0.63
nerol 0.07 0.33 0.59
geraniol 0.07 0.32 0.45
perillyl alcohol 0.05 0.71 1.22
thymol 0.12 0.12 0.11 0.56 2.13 4.2
p-mentha-1,8-dien-10-ol Tr. 0.01 0.05 0.97 1.04
total 2.08 2.26 2.27 6.72 10.81 9.86
oxides and epoxides
(Z)-limonene oxide Tr. Tr. 0.01 0.07 0.15 0.07
(E)-limonene oxide 0.01 0.03 0.03 0.03 0.09 0.06
(Z)-caryophyllene epoxide 0.01 0.02 0.09 0.24 0.53
(E)-caryophyllene epoxide 0.01 0.01 0.03 0.16 0.48
total 0.01 0.05 0.07 0.22 0.64 1.14
sesquiterpene hydrocarbons
d-elemene 0.09 0.08 0.06 0.09 0.16 0.23
a-copane 0.01 0.03 0.03 0.03 0.15 0.15
b-elemene 0.04 0.04 0.04 0.13 0.18 0.16
caryophyllene 0.17 0.16 0.17 0.17 0.2 0.26
aroma dendrene 0.01 0.01 0.01 0.01 0.01 0.02
g-elemene 0.02 0.02 0.01 0.05 0.05 0.3
allo-aroma dendrene 0.01 0.01 0.01 0.05 0.05 0.1
(E)-b-farnesene 0.45 0.6 0.63 2.13 2.78 2.51
a-humulene 0.04 0.03 0.04 0.06 0.19 0.2
d-muurolene Tr. Tr. Tr. 0.05 0.06
germacrene D 0.2 0.19 0.17 0.19 0.16 0.16
a-muurolene 0.04 0.04 0.03 0.04 0.22 0.13
bicycloger-macrene 0.99 0.82 0.52 0.31 0.14 0.03
d-g-cadinene 0.04 0.04 0.04 0.07 0.15 0.32
sesquiphell-andrene 0.03 0.03 0.04 0.18 0.21 0.63
total 2.14 2.1 1.8 3.51 4.7 5.26
Relative Concentration (%)
Compound Fresh Oil 1 3 6 9 12
sesquiterpene alcohols
globulol 0.01 0.01 0.01 0.02 0.1 0.35
elemol Tr. 0.03 0.07 0.16
viridiflorol Tr. 0.02 0.35
spathulenol 0.03 0.26 2.64 14.05 28.25
a-cadinol Tr. 0.02 0.13 0.24 1.88
T-cadinol Tr. 0.21 0.52
b-eudesmol Tr. 0.02 0.18 0.47 1.91
a-eudesmol Tr. 0.11
(Z,E)-farnesol Tr. Tr. 2.55
total 0.01 0.04 0.31 3 15.16 36.08
total natural volatiles (%) 97.58 97.66 96.29 94.24 78.62 53.26
total artifact volatiles (%) 0.05 0.34 6.32 17.79 40.04
From Njoroge et al., 1996 (43).

Tr. Traces.
These compositional changes usually negatively influence the odor and flavor of
citrus peel oils by generating off-flavor products. It has been shown that nonvolatile
residues of citrus peel oil contain some compounds that exhibit antioxidative activ-
ities, among which permethoxylated flavones, dehydroabietic acid (46), coumarins,
and psoralens have been identified (33). In this respect, cold-pressed citrus peel oil
is more stable than distilled oil and essence oil, in which most of the natural anti-
oxidants present are left behind when the oil is distilled (1).
Citrus seed oil, another byproduct in citrus industry, is required to have high sta-
bility for cooking purposes. Citrus seed oil is subject to oxidative changes because
of the presence of a high percentage of unsaturated fatty acids. The oil is readily
oxidized in the presence of air, generating hydroperoxides, alcohols, aldehydes,
ketones, hydrocarbons, and carboxylic acids as the primary, secondary, and tertiary
oxidized products, respectively.
4.2. Storage Conditions

One of the most important reactions involved in chemical changes in citrus oils is
certainly the oxidation reaction, due to the high content of terpenes in citrus peel oil
and unsaturated fatty acids in citrus seed oil (3). Usually, quality deterioration in
oils may occur under autoxidation, photo-oxidation, lipoxygenase-assisted oxida-
tion, or thermal oxidation, all of which should be controlled in order to protect
the oils from deterioration and off-flavor development (47). To maintain the original
quality of citrus oils, undue exposure to air and contact with metals such as iron and
copper, which act as pro-oxidants, should be avoided during processing and subse-
quent storage (1). In some cases, citrus oils are stored refrigerated (05 C) under an
APPLICATIONS OF CITRUS OILS AND ESSENCES 61
inert gas (33). Meanwhile, the addition of antioxidants may help retard or control
oxidation of citrus oils. Tocopherols are present in citrus seed oil with a-tocopherol
predominating (1). They effectively protect the oil from oxidation, with phospho-
lipids remaining in the oil preventing their degradation (1, 47). Some phenolic anti-
oxidants, such as butylated hydroxyanisole (BHA), butylated hydroxytoluene
(BHT), propyl gallate (PG), and tert-butylhydroquinone (TBHQ), are also used
in limited dosage to inhibit the oxidation of edible oils such as citrus seed oil
(47). In addition, partial hydrogenation provides satisfactory shelf life to some
citrus seed oils, in which the high content of linolenic acid may pose stability pro-
blems (1). However, hydrogenation is not considered desirable as it leads to the loss
of essential fatty acids and may also affect the flavor quality of the product.
Unlike citrus seed oils, citrus peel oils are themselves very good antioxidants
capable of inhibiting free radical-mediated reactions (48). As reported by Song
et al. (49), abundant tocopherols were found in citrus peel oils, yet there was little
correlation between tocopherol content and antioxidative activity in the oils, sug-
gesting that the composition of terpenes present might be a major determinant of
the antioxidative status of citrus peel oils. The compounds b-pinene, myrcene,
a-terpinene, and g-terpinene were identified to have higher or similar antioxidative
activities compared with that of d-tocopherol. Encapsulation is a technique fre-
quently used in the storage of citrus peel oils and essences, which isolates the
oils from the atmospheric oxygen, moisture, temperature, and light, and hence
minimizes the oxidation of oil and reduces the release of volatile flavor compounds
(50). Citrus peel oil may be spray-dried and encapsulated in a double emulsion (50).
A water activity of 0.628 was found to result in good resistance to oxidation without
the occurence of caking or stickiness during the storage of spray-dried encapsulated
citrus peel oil (51).
5. APPLICATIONS OF CITRUS OILS AND ESSENCES
The applications of citrus oils are versatile and in many domains. As a result of
their freshness, lightness, and fine fruity aroma, citrus peel oils and essences are
widely used in the food and beverage industries as well as in some nonfood
applications (1).
The applications of cold-pressed peel oils in food and beverage are mainly in the
soft drinks, sherbet, confectionery, bakery, and household extracts (1). In addition,
they can act as reducing agents of peroxidase activity in leafy vegetables and anti-
oxidants for edible oils, such as olive oil, to improve their sensory properties
(52, 53). Moreover, they are effective inhibitors for the formation of N-nitrosodi-
methylamine (NDMA), a known carcinogen, that may occur during production and
storage of food (54). Citrus peel oils are also added as flavoring agents to pharma-
ceutical and drugs as well as herbal medicines in order to mask their unpleasant
tastes (14). Distilled peel oils, different from cold-pressed oils, have found their
place in the perfumery and cosmetic industries as well as in the manufacturing
of soap and paper (1).
Citrus peel oils may also be used for their antioxidative, antitumor, and radical-
scavenging activities. The radical-scavenging ability of citrus peel oil may help pre-
vent free radical-induced and various chronic diseases (48, 55, 56). Monoterpenes
from volatile components and polymethoxylated flavones from nonvolatile residues
have been reported to be effective inhibitors of tumor cell growth, implicating that
citrus peel oils may be good cancer preventive food additives (5759). Furthermore,
citrus peel oils are useful to alleviate pain from burnt skin (60). Demonstrating
anxiolytic and sedative effect, they could also be used in primary medical care
against insomnia, anxiety, and epilepsy (61).
The insecticidal property and antimicrobial activity of citrus peel oils have been
reported. The oil can repel moth, mosquito, cockroach, domestica, and housefly
(6266). It also inhibits the growth of microbes such as fungi and salmonellae,
with monoterpenes being the major compounds that account for pathogen fungi
inhibition (46, 67, 68).
Unlike citrus peel oil, citrus seed oil is mainly used after refining as edible oil
and a source of essential fatty acids. Refined citrus seed oil is widely used in mar-
garines, shortenings, salad dressings, and salad and cooking oils. Meanwhile, crude
citrus seed oils are useful in the preparation of fatty acid derivatives, soap and
detergent, and for the treatment of leather and textile (1).
6. CHALLENGES
Despite increasing applications of citrus oils, certain challenges related to potential

health-damaging properties and contamination of citrus oils should not be
ignored.
Citrus peel oil, such as bergamot oil, has been reported to show potential health
hazards. Bergamot oil has a pleasant refreshing scent, and had been used widely in
cosmetics until it was restricted in most countries a few years ago because of certain
adverse effects, primarily phototoxicity and Berloque dermatitis (69). More
recently, there seems to be increasing application of bergamot oil in aromatherapy.
However, as reported by Kaddu et al. (69), bergamot oil possesses potential photo-
toxic and photomutagenic properties, indicating that special attention should be
paid in using of psoralen-containing aromatherapy oils such as bergamot oil.
Aside from potential adverse health effects, contamination is another problem
that should be considered in production and application of citrus oils. Citrus oils
may be contaminated with plastic materials employed in production process or sto-
rage. Chloroparaffins, phthalate esters, and phosphorylated plasticizers are the
major contaminants extracted from plastic components by citrus oils as the oil/
water emulsions pass through various production phases (70). Phthalate esters
have a wide range of activities and may be hepatotoxic, carcinogenic, and possibly
damaging to the gastrointestinal tract (71). Chloroparaffins, another class of plasti-
cizers, are toxic by ingestion and subcutaneous route and are classified as being
potentially carnogenic to humans (72). It is suggested that contamination of citrus
peel oils by plasicizers does not depend on the nature of the oil, but probably
REFERENCES 63
correlated with the procedures used during production cycle or storage of the oils
(70, 72).
Another contamination of citrus peel oils comes from chlorine-treated water
used in the oil recovery process and sanitizers used in postharvest handling and pro-
cess equipment cleaning, which serve as a potential source of hypochlorous acid
(HOCl) (73). HOCl can react with a variety of terpenes similar to d-limonene in
structure, including limonene, a-pinene, and a-terpineol, resulting in the formation
of terpene chlorohydrins. The contamination of terpene chlorohydrins could be
reduced through reduction of the chlorine levels in the treatment water (74).
Pesticide residues may contaminate citrus peel oils as well. Cultivation of citrus
crops commonly involves the use of chemicals such as fertilizers and pesticides.
Regulations are increasingly stricter in terms of residual levels of pesticides
because of the application of citrus oils in food, pharmaceutical, and cosmetic
industries (75). Citrus peel oils, extracted from citrus peels, contain a higher con-
centration of pesticide residues than the fruits, due to the direct contact of the peels
with pesticides. Organophosphorus and organochlorine pesticide residues in citrus
peel oils have shown a steady decrease in recent years (76).
7. CONCLUSIONS
There is considerable interest in the chemical composition and properties of citrus

oils and essences as well as the role they play in food and nonfood industries. Citrus
peel oils and essences possess a pleasant aroma, with oxygenated compounds being
the major constituents that account for their characteristic odor. Terpenes, the most
abundant components in cold-pressed citrus peel oil, are removed in concentrated
oil production, usually by use of adsorbant and supercritical carbon dioxide, to
increase the concentration of oxygenated compounds and to enhance the qualifica-
tion of the oil. Meanwhile, citrus seed oils are composed largely of triacylglycerols
and are rich in oleic and linoleic acids.
Citrus oils and essences are unstable to heat, light, oxygen, and metals. Thus,
they should be stored under appropriate conditions in order to avoid compositional
changes that lead to their quality deterioration. Citrus oils and essences are widely
used in food and nonfood industries. Applications in the health-related areas are
forthcoming.
REFERENCES
1. H. E. Swisher and L. H. Swisher, Specialty citrus products, in Citrus Science and

Technology, vol. 2, S. Nagy, P. E. Shaw, and M. K. Veldhuis, ed., The AVI Publishing
Company, Westport, Connecticut, 1977, pp. 290345.
2. K. Ogawa, A. Kawasaki, T. Yoshida, H. Nesumi, M. Nakano, Y. Ikoma, and M. Yano,
J. Agric. Food Chem., 48, 17631769 (2000).
3. M. Usai, G. Arras, and F. Fronteddu, J. Agric. Food Chem., 40, 271275 (1992).
4. P. E. Shaw, Essential Oils, in Citrus Science and Technology, vol. 1, S. Nagy, P. E. Shaw,
and M. K. Veldhuis, eds., The AVI Publishing Company, Westport, Connecticut, 1977,
pp. 427462.
5. H-S. Choi and M. Sawamura, J. Agric. Food Chem., 48, 48684873 (2000).
6. H-S. Choi, J. Agric. Food Chem., 51, 26872693 (2003).
7. K. Bauer, D. Garbe, and H. Surburg, Natural raw materials in the flavor and fragrance
industry, in Common Fragrance and Flavor Materials, K. Bauer, D. Garbe, and
H. Surberg, eds., VCH Publishers, New York, 1990, pp. 146147.
8. R. W. Wolford, J. W. Kesterson, and J. A. Attaway, J. Agric. Food Chem., 19, 10971105
(1971).
9. T. A. El-Adawy, E. H. Rahma, A. A. El-Bedawy, and A. M. Gafar, Nahrung-Food, 43,
385391 (1999).
10. P. E. Shaw, Aqueous Essences, in Citrus Science and Technology, vol. 1, S. Nagy,
P. E. Shaw, and M. K. Veldhuis, eds., The AVI Publishing Company, Westport, Con-
necticut, 1977, pp. 463478.
11. R. W. Wolford, M. H. Dougherty, and D. R. Petrus, Int. Fruchtsaftunion. Ber. Wiss. Tech.
Komm., 9, 151169 (1969).
12. H. S. Song, M. Sawamura, T. Ito, A. Ido, and H. Ukeda, Flavor Fragr. J., 15, 323328
(2000).
13. L. Mondello, A. Casilli, P. Q. Tranchida, L. Cicero, P. Dugo, and G. Dugo, J. Agric. Food
Chem., 51, 56025606 (2003).
14. M-L. Lota, D. de R. Serra, F. Tomi, C. Jacquemond, and J. Casanova, J. Agric. Food
Chem., 50, 796805 (2002).
15. B. Steuer, H. Schulz, and E. Lager, Food Chem., 72, 113117 (2001).
16. F. Buiarelli, G. P. Cartoni, F. Coccioli, and T. Leone, J. Chromatogr. A., 730(1-2), 916
(1996).
17. A. Verzera, A. Trozzi, A. Cotroneo, D. Lorenzo, and E. Dellacassa, J. Agric. Food Chem.,
48, 29032909 (2000).
18. W. Feger, H. Brandauer, and H. Ziegler, J. Essential Oil Res., 15, 143147 (2003).
19. T. Veriotti and R. Sacks, Anal. Chem., 74, 56355640 (2002).
20. N. T. M. Tu, L. X. Thanh, A. Une, H. Ukeda, and M. Sawamura, Flavor Fragrance J., 17,
169174 (2002).
21. S. A. Vekiari, E. E. Protopadakis, P. Papadopoulou, D. Papanicolaou, C. Panou, and M.
Vamvakias, J. Agric. Food Chem., 50, 147153 (2001).
22. W. Feger, H. Brandauer, and H. Ziegler, J. Essential Oil Res., 13, 274277 (2001).
23. Z. Shen, V. Mishra, B. Imison, M. Palmer, and R. Fairclough, J. Agric. Food Chem., 50,
154160 (2002).
24. M. Sato, M. Boto, and T. Hirose, Ind. Eng. Chem. Res., 34, 39413946 (1995).
25. N. T. M. Tu, Y. Onishi, H-S. Choi, Y. Kondo, S. M. Bassore, H. Ukeda, and M. Sawamura,
J. Agric. Food Chem., 50, 29082913 (2002).
26. M. G. Chisholm, M. A. Wilson, and G. M. Gaskey, Flavor Fragrance J., 18, 106115
(2003).
27. H-S. Choi, Y. Kondo, and M. Sawamura, J. Agirc. Food Chem., 49, 24042408 (2001).
28. D. Sun and P. D. Petracek, J. Agric. Food Chem., 47, 20672069 (1999).
REFERENCES 65
29. A. Buettner, M. Montse, A. Fischer, J. Guasch, and P. Schieberle, Eur. Food. Res. Technol.,
216, 1114 (2003).
30. D. Chouchi, and D. Barth, J. Chromatogr. A., 672(1-2), 177183 (1994).
31. P. Dugo, L. Mondello, G. Dugo, D. M. Heaton, K. D. Bartle, A. A. Clifford, and P. Myers,
J. Agric. Food Chem., 44, 39003905 (1996).
32. P. Stremple, HRC-J. High Resolution Chromatogr., 21, 587591 (1998).
33. S. M. Njoroge, H. Ukeda, and M. Sawamura, J. Agric. Food Chem., 51, 40294035 (2003).
34. V. Andrea, N. Nedia, A. Andrea, and R. M. Teresa, J. Agric. Food Chem., 51, 49784983
(2003).
35. S. Nagy, Lipids: Identification, Distribution, and Importance, in Citrus Science and
Technology, vol. 1, S. Nagy, P. E. Shaw, and M. K. Veldhuis, eds., The AVI Publishing
Company, Westport, Connecticut, 1977, pp. 266301.
36. M. A. Habib, M. A. Hammam, A. A. Sakr, and Y. A. Ashoush, J. Amer. Oil Chem. Soc., 63,
11921197 (1986).
37. J. L. Weihrauch and B. B. Teter, Fruit and Vegetable By-product as Sources of Oil, in
Technological Advances in Improved and Alternative Sources of Lipids, B. S. Kamel and
Y. Kakuda, eds., Blackie Academic & Professional, London, 1994, pp. 178180.
38. K. Ajewole and A. Adeyeye, Food Chem., 47, 7778 (1993).
39. M. Saidani, W. Dhifi, and B. Marzouk, J. Food Lipids., 11, 242250 (2004).
40. A. Sattar, S. Mahmud, and S. A. Khan, Pakistan J. Sci. Indust. Res., 31, 743744 (1988).
41. A. J. Hognadottir and R. L. Rouseff, J. Chromatogr. A., 998, 201211 (2003).
42. M. G. Moshonas and P. E. Shaw, J. Agric. Food Chem., 31, 334336 (1983).
43. S. M. Njoroge, H. Ukeda, and M. Sawamura, J. Agric. Food Chem., 44, 550556 (1996).
44. H-S. Choi and M. Sawamura, Biosci. Biotechnol. Biochem., 66, 439443 (2002).
45. P. Schieberle and W. Grosch, Lebensm. Wissenschaft und Technologie, 21, 158162
(1988).
46. I. Vargas, I. Sanz, P. Moya, and E. Prima-Yufera, J. Food Prot., 62, 929932 (1999).
47. F. Shahidi, Nahrung-Food., 44, 158163 (2000).
48. H-S. Choi, H. S. Song, H. Ukeda, and M. Sawamura, J. Agric. Food Chem., 48, 41564161
(2000).
49. H. S. Song, H. Ukeda, and M. Sawamura, Food Sci. Technol. Res., 7, 5056 (2001).
50. A. Edris and B. Bergnstahl, Nahrung-Food, 45, 133137 (2001).
51. C. I. Beristain, E. Azuara, and E. J. V. Carter, J. Food Sci., 67, 206211 (2002).
52. M. Charai, M. Faid, and A. Chaouch, J. Essential Oil Res., 11, 517521 (1999).
53. A. G. Ponce, C. E. del Valle, and S. I. Roura, Lebensmittel. Wissenschaft. und Techno-
logie., 37, 199204 (2004).
54. M. Sawamura, S. H. Sun, K. Ozaki, J. Ishikawa, and H. Ukeda, J. Agric. Food Chem., 47,
48684872 (1999).
55. V. Calabrese, S. D. Randazzo, C. Catalano, and V. Rizza, Drugs. Exp. Clin. Res., 25, 219
225 (1999).
56. V. Calabrese, G. Scapagnini, S. D. Randazzo, G. Randazzo, C. Catalano, G. Geraci, and
P. Morganti, Drugs. Exp. Clin. Res., 25, 281287 (1999).
57. J. Chen, A. M. Montanari, and W. W. Widmer, J. Agric. Food Chem., 45, 364368 (1997).
58. P. L. Crowell, J. Nutr., 129, 775778 (1999).

59. Y. Takahashi, N. Inaba, S. Kuwahara, W. Kuki, K. Yamane, and A. Murakami, J. Agric.
Food Chem., 50, 31933196 (2002).
60. T. P. Brown, P. C. Hill, and F. B. Bailie, J. Int. Soc. Burn Injuries, 25, 678 (1999).
61. M. I. Carvalho-Freitas and M. Costa, Biol. Pharm. Bull., 25, 16291633 (2002).
62. H. C. Su, R. D. Speirs, and P. G. Mahany, J. Econ. Entomol., 65, 14381441 (1972).
63. E. A. Omer, A. A. Youssef, E. N. Abo-Zeid, and A. M. Sharaby, Egyptian J. Horticul., 24,
207218 (1997).
64. T. Bruno, Fitoterapia, 71, S29S37 (2000).
65. A. A. Shalaby, K. A. Allam, A. A. Mostafa, and S. M. Fahmy, J. Egypt. Soc. Parasitol., 28,
595606 (1998).
66. F. C. Ezeonu, G. I. Chidume, and S. C. Udedi, Bioresour. Technol., 76, 273274 (2001).
67. D. R. Caccioni, M. Guizzardi, A. Renda, and G. Ruberto, Int. J. Food Microbiol., 43(1-2),
7379 (1998).
68. M. E. Parish, D. Baum, R. Kryger, R. Goodrich, and R. Baum, J. Food Protec., 66, 1704
1707 (2003).
69. S. Kaddu, H. Kerl, and P. Wolf, J. Amer. Acad. Dermatol., 45, 458461 (2001).
70. G. D. Bella, M. Saitta, S. Locurto, F. Salvo, G. Licandro, and G. Dugo, J. Agric. Food
Chem., 49, 37053708 (2001).
71. G. D. Bella, M. Saitta, M. Pellegrino, F. Salvo, and G. Dugo, J. Agric. Food Chem., 47,
10091012 (1999).
72. G. D. Bella, M. Saitta, S. Locurto, V. L. Turco, A. Visco, and G. Dugo, J. Agric. Food
Chem., 48, 44604462 (2000).
73. E. R. Weiss, J. Pika, and R. J. Braddock, J. Agric. Food Chem., 51, 22772282 (2003).
74. E. R. Weiss, R. J. Braddock, R. M. Goodrich, J. F. Gregory, and J. Pika, J. Food Sci., 68,
21462149 (2003).
75. S. Barrek, O. Paisse, M. Florence, and G. Loustalot, Anal. Bioanal. Chem., 376, 157161
(2003).
76. M. Saitta, G. D. Bella, F. Salvo, S. Locurto, and G. Dugo, J. Agric. Food Chem., 48, 797
801 (2000).
4
Gamma Linolenic
Acid Oils
Rakesh Kapoor and Harikumar Nair
Bioriginal Food & Science Corp.
1. INTRODUCTION
Essential fatty acids (EFAs) can be defined by classic definition, which defines
EFAs as the fatty acids that are required for proper functioning of cells, but the
body cannot synthesize them and, therefore, must be supplied by diet. According
to this definition, there are only two EFAs: linolenic acid (LA, C18:2, n-6) and
alpha-linolenic acid (ALA, C18:3, n-3). The functional definition of EFAs includes
the fatty acids that can correct the symptoms produced by elimination of all EFAs
from the diet. According to this definition, LA, gamma linolenic acid (GLA, C18:3,
n-6), and arachidonic acid (AA, C20:4, n-6) are EFAs of n-6 family (1, 2).
Gamma linolenic acid (cis-6, cis-9, cis-12-octadecatrienoic acid) is an 18-carbon
polyunsaturated fatty acid containing three double bonds. It is produced in the body
from desaturation of LA by the reaction catalyzed by enzyme delta-6-desaturase
(D-6-D) (Figure 1). GLA is rapidly elongated to DGLA by elongase enzyme.
Cats do not have this enzyme; hence, they cannot synthesize GLA and subsequent
metabolites from LA (3). Therefore, cats must eat a meat-based diet to obtain longer
chain metabolites of LA (DGLA, AA). DGLA can be acetylated and incorporated
into membrane phospholipids. A small amount can be converted into AA, and this
67
68 GAMMA LINOLENIC ACID OILS
Figure 1. Metabolic pathway for linoleic acid. COX, Cycloozygenase, LOX, Lipoxygenase,
PGE1 , Prostaglandin E1 , PGE2 , Prostaglandin E2 , PGI2 , Prostaglandin I2 ,TXA2 , Thromboxane
A2 , 15-HETrE, 15-hydroxy eicosatrienoic acid, LTB4 , Leukotriene B4 .
reaction is catalyzed by delta-5-desaturase enzyme. Different animal species and

different tissues differ in their capacity to convert DGLA to AA. Rat metabolizes
DGLA to AA in significant amounts, whereas humans and other species have lim-
ited capacity to form AA from DGLA.
The reaction catalyzed by delta-6-desaturase enzyme is the slowest reaction in
the metabolic pathway of LA and is considered as a rate-limiting step (4, 5). Acti-
vity of this enzyme further decreases with age and in people suffering from various
diseases, including arthritis, diabetes, hypertension, eczema, psoriasis, and so on.
Lifestyle factors like stress, smoking, excessive consumption of alcohol, linoleic
acid (6), saturated and trans-fatty acids and nutritional deficiencies of Vitamin
B6, zinc (7), and magnesium inhibit this desaturase. As a result of limitations in
in vivo production of GLA, supplementation with preformed GLA is becoming
important. This has led to interest in development and commercialization of the
sources of GLA.
2. SOURCES OF GLA
GLA is present in small amounts in many plants belonging to the families Acera-
ceae, Boraginaceae, Cannabinaceae, Liliaceae, Onagraceae, Ranunculaceae,
SOURCES OF GLA 69
Saxifragaceae, and Scrophulariaceae. Kleiman et al. (8) investigated 29 species of

family Borageniceae for the presence of GLA and tetraenoic (stearidonic acid,
SDA) fatty acid. They observed 027% GLA, 056% ALA, and 017% SDA in
seed oils from different plants in Boraginaceae. Janick et al. (9) identified 32 plants
in which the content of GLA in seed oil can be more than 5% weight/weight (w/w)
of total fatty acids (Table 1). The important crops that have been commercialized as
sources of GLA-rich oils are discussed below.
TABLE 1. Selected Plant Species High in Gamma-Linolenic Acid.
Family Oil Content GLA Content (%)

Genus and Species of Seed (%) of Oil of Seed
Boraginaceae
Adelocaryum coelestinum 22.0 12.0 2.7
Alkanna orientalis 23.0 12.0 2.8
Anchusa azurea 21.0 13.0 2.7
Anchusa capensis 29.0 10.0 2.9
Anchusa hybrida 20.0 13.0 2.6
Borago offcinalis 2838 1725 5:08:4
Brunnera orientalis 27.0 15.0 4.2
Cerinthe minor 10.0 10.0 1.0
Cynoglossum amabile 23.0 11.0 2.5
Cynoglossum lanceolatum 25.0 13.0 3.3
Echium rubrum 15.0 14.0 2.1
Echium vulgare 22.0 11.0 2.4
Gastrocatyle hispida 28.0 16.0 4.5
Lithospermum arvense 17.0 14.0 2.4
Lithospermum purpureocaeruleum 14.0 18.0 2.5
Moltkia aurea 10.0 10.0 1.0
Moltkia coerules 10.0 11.0 1.1
Nonea macrosperma 39.0 13.0 5.1
Onosma sericeum 20.0 13.0 2.6
Onosmodium molle 17.0 20.0 3.4
Onosmodium occidentale 17.0 18.0 3.1
Paracarvum caelestinum 21.0 12.0 2.5
Pectocarva platycarpa 15.0 15.0 2.3
Symphaticum officinale 21.0 27.0 5.6
Cannabaceae
Cannabis sativa 38.0 36 1:12:3
Onagraceae
Oenothera biennis 1725 710 1:22:5
Oenothera grandifolia 4.0 9.3 0.3
Saxifragaceae
Ribes alpinum 19.0 9.0 1.7
Ribes nigrum 30.0 1519 4:65:8
Ribes rubrum 25.0 46 1:01:5
Ribes uva-crispa 18.0 1012 1:82:2
Scrophulariaceae
Scrophularia marilandica 38.0 10.0 3.6
2.1. Borage (Borago Officinalis L.)

Borage is also known as star flower because of the shape of its flowers. Borage is
the only member of the Boraginaceae family that is being grown commercially for
its seeds at present. It is an annual herb native to Europe, Asia Minor, and North
America. The morphology of this plant has been reviewed in many publications
(911). It is an erect, hispid plant that can grow up to 100 cm in height. It has inde-
terminate vegetative growth habits that pose a significant challenge in commercial
cultivation and harvesting of this plant. The plant has simple alternate leaves that
are obovate, ovate, or oblong with an obtuse apex and create margin. The upper
surface of leaves is dark-to-medium green, whereas the lower surface is light green.
The stem is cylindrical, hollow, succulent, and occasionally susceptible to lodging.
Stems, leaves, and calyx are covered with white, stiff, unicellular trichomes that can
cause contact dermatitis to susceptible people. Flowers are bright blue, violet, pink,
or white and are star shaped, hence, the other name for boragestar flower. Ovary
is four lobed, and as the flower matures, it develops into 34 ovoid or oblong seeds,
also known as nutlets. As they mature, they change color from green to brown to
black and abcise rapidly. Flowering continues over a long period of time, and at any
time, the plant bears the flowers and seeds in different stages of maturity. These
factors pose a significant challenge in commercial production and harvesting.
Borage is mainly grown in the United Kingdom, Holland, Canada, New Zealand,
and Poland, and it is estimated that about 95% of the world crop of borage is grown
in these five countries. Borage is still not a major crop as its cultivation is very labor
intensive. Most of the production of borage is under contract with oil producers
with buy-back arrangement for seeds, and the production data regarding tonnage
and acreage is not reported in any of the commercial publications. The acreage
for borage grew consistently between 1980 and 1999 from a few hundred acres
to about 7000 ha. There was an excellent harvest in 1999 in Canada, and the
demand flattened leading to an over supply. This situation led to very limited pro-
duction between 1999 and 2002 (unpublished information supplied by Bioriginal
Food & Science Corp.). In 2002, the cultivation of borage increased again, but
the severe drought in many of the borage-growing areas reduced the yield (unpub-
lished information, Bioriginal Food & Science Corp.). Foliage is not used commer-
cially any more, although in the folklore, it was used by early Europeans as a salad
or vegetable and as a decoction for the treatment of various disease conditions.
As a result of seed drop and the continuous and long period of flowering
and seed maturation, the seed yield is variable and a significant amount of potential
crop is lost. The yields vary between 100 and 300 kg/ha in Canada and New Zealand
and 300 and 500 kg/ha in the United Kingdom. The difference between Canada
and Europe is probably caused by climatic conditions as borage prefers cool, moist
conditions.
A limited amount of work has been done on optimization of cultivation, harvest-
ing conditions, and plant variety development for borage and is reviewed by Clough
(12) and Janik et al. (9) Because of the semiperennial nature of the crop and the
seed drop, swathing and combining is the preferred harvest method. Simpson
SOURCES OF GLA 71
(13) concluded from his studies that substantially higher yields could be obtained
by swathing borage than by desiccating it with diquat or glyphosate. El Hafid et al.
(14) studied the effect of seeding dates and nitrogen fertilization effects on borage
in Alberta, Canada. In this case, early planting resulted in significantly higher seed
yield and harvest index. Nitrogen fertility levels had no significant effect on the
seed yield. The major objectives of borage variety development are to improve
yield through development of varieties that are seed retentive on maturation and
mature evenly. Other traits of interest include high oil content and high GLA con-
tent and morphological and yield stability. There is significant variability between
the wild and cultivated borage accessions in terms of the amount of GLA and other
fatty acid components as a percentage of oil indicating the possibility of developing
lines with better quality parameters with additional breeding efforts (15). Hoffman
La Roche has registered two varieties of borage (Tyreman and Spruce), but neither
of these varieties are seed retentive. Bioriginal Food and Science Corp. has regis-
tered varieties of borage that have better seed retention compared with traditional
borage.
From studies done on the in vitro propagation and somatic embryology (1618)
in borage, it appears that it is possible to produce GLA from cotyledonary somatic
embryos. However, the authors observed that the system needs to be adapted to
liquid media to facilitate large-scale production.
2.1.1. Chemistry The average weight of seed is between 16.1 and 24.5 mg (18).
The oil content varies between 28% and 37% (w/w) and consists chiefly of linoleic
(C18:2, 3438%), GLA (C18:3, n-6; 1627%), and oleic (C18:1, 1418%) acids.
The variation in GLA content is a result of geographical location, length of light
period during growing season, average temperature, and diurnal temperature varia-
tions. There is no relationship between content of LA and GLA, although an inverse
relationship is reported for content of 18:1 and GLA.
The average fatty acid composition of several seed lots over years, as provided
by Bioriginal Food & Science Corp (unpublished data) and published information,
is reported in Table 2. It has been observed that geographical location only affects
the content of GLA and has no effect on the oil content of seeds, whereas plant
density has no effect on the content of GLA (19). Oil also contains minor compo-
nents that are characteristic of all vegetable oils and include sterols, tocopherols,
and pigments. Borage oil contains about 650 ppm of g- and 50 ppm of d-tocophe-
rols (20). The content of unsaponifiable fraction is dependent on the method of
extraction of oil and the degree of refining and varies between 0.8% and 1.5%.
The relative proportion of different sterols in borage oil is independent of the meth-
od of refining. The major proportion of sterols in borage oil belongs to 4-desmethyl-
sterols, although small amounts of 4-monomethyl and 4,4-dimethylsterols are also
present. Campsterol and sitosterol constitute more than 50% of 4-desmethylsterols
(21). Major sterols belonging to 4-monomethylsterol family include gramisterol,
obtusifoliol, and citrostadienol (21).
Like other members of the Boraginaceae family, the Borage plant also contains
pyrrolizidine alkaloids. Seven pyrrolizidine alkaloids have been identified so far in
TABLE 2. Average Composition of Commercial Borage Crops.
Crop Year 2002 1999 1998 1997 1996
Sample size 17 87 71 53 43
Moisture (%) 10.39 10.54 8.90 9.80 8.80
Oil (%) 31.16 31.40 30.60 30.70 32.40
Free Fatty acid 1.23 1.34 1.70 2.40 1.50
Fatty Acid Profile
C16:0 9.9 10.6 10.2 10.3 10.4
C 16:1 0.3 0.3 0.2 0.2 0.2
C18:0 4.1 4.3 3.3 3.5 3.6
C18:1 17.7 17.6 14.8 14.7 14.3
C18:2 36.7 35.7 37.9 37.6 37.3
C18:3 (n-3) 0.3 0.2 0.2 0.2 0.2
C18:3 (n-6) 22.4 21.6 24.6 24.1 24.9
C18:4 0.2 0.2 0.2 0.1 0.2
C20:0 0.3 0.3 0.2 0.1 0.2
C20:1 4.1 4.1 3.9 3.9 3.9
C20:2 0.2 0.2 0.2 0.2 0.2
C22:0 0.2 0.2 0.1 0.1 0.1
C22:1 2.6 2.7 2.6 2.5 2.5
C24:1 1.7 1.7 1.5 1.5 1.5
borage leaves, flowers, and seeds. Thesinine, a saturated pyrrolizidine alkaloid, is

the major alkaloid, and the six unsaturated pyrrolizidine alkaloids identified so far
include amabiline, supinine, lycopsamine, interemedine, acetyllycopsamine, and
acetylintermedine and are minor constituents. The total alkaloid content of the plant
was reported to be less than 0.001%, whereas mature seeds yield about 0.03% crude
alkaloids (22). Herrman et al. (23) reported the presence of thesinine as a glycoside
(Thesinine-40 -O-b-D-glucoside) in seeds. As borage oil is an item of commerce for
its content of GLA, there is a concern about the content of pyrrolizidine alkaloids in
the oil. Published research to date could not detect the presence of pyrrolizidine
alkaloids in borage oil. Dodson and Stermitz (22) used a method with a detection
limit of 5 ppm, whereas Parvais and Stricht (24) employed a method with a detec-
tion limit of 0.1 ppm. These authors reported the absence of pyrrolizidine alkaloids
in the oil samples tested. Mierendroff et al. (25) developed a method of detection of
pyrrolizidine alkaloids at a limit of 4 ppb. Employing Mierendroffs method at an
independent testing lab in Germany, Bioriginal tested several lots of borage oil over
several years and could never detect any traces of pyrrolizidine alkaloids at such a
low detection limit. The German Health Authority has limited the intake of unsa-
turated pyrrolizidne alkaloids to 1 mg per day. Based on the above results, and
assuming that borage oil contains PAs at a level of 4 ppb, one may have to consume
more than 250 capsules (1000 mg each) every day to get a total of 1-mg pyrrolizi-
dine alkaloids. Based on this analysis, there is no likelihood of toxicity from pyr-
rolizidine alkaloids from borage oil ingestion.
SOURCES OF GLA 73
2.2. Evening Primrose (Oenothera biennis L.)

Evening primrose of commerce consists of three species of family Onagraceae:
Oenothera biennis, O. lamarckiana, and O. parviflora. It is native to North America
and is now commonly found in many temperate zones around the world.
Evening primrose thrives in open areas, especially in dunes and sandy soil.
Evening primrose is grown commercially for its seed oil. China has become the
major supplier of evening primrose seed and oil to world markets because it has
the lowest cost of production and is estimated to supply about 90% of world
supply of oil. Until 1985, it was mainly collected from wildly growing plants;
however, increase in demand justified the commercial production. In the wild,
seeds exhibit physiological dormancy, whereby they may cause problems in
cultivation as the plant density depends on the conditions after seeds are sown
into the ground, of which the farmer has no control. Seeds may lay dormant for
a long period of time (up to several years in the wild). Seeds can be forced by
cold stratification. Germination is equal in light and dark. Plant also flowers
over a period of 23 months, with 24 flowers at any time, and seed maturation
takes place in pods that split when ripe and shatters seeds on the ground. New vari-
eties of evening primrose have been developed over the last two decades that ger-
minate within 3 weeks under test conditions and retain pods intact on the plant.
Seeds are small, angular, and dark brown, and 1000 seeds may weigh between
0.2 and 0.6 g and each pod may contain up to 180 seeds. Each plant can have up
to 250 pods.
Some of the species of evening primrose plant, including O. biennis and
O. lamarckiana, are very unique and are challenging for their unique genetics.
They do not obey Mendelian genetics and have been a subject of study for over
100 years. These species differ from common plants in that, during meiosis,
the chromosomes do not pair up but form a circle by joining end to end. This
prevents reshuffling of genes through translocation. The plants are self-pollinating,
but inbreeding does not occur because of the presence of a set of lethal
genes.
2.2.1. Chemistry Evening primrose plant is cultivated for its seeds that
contain about 16% protein and 2328% fat. The oil is rich in linoleic acid (70
75%) and gamma-linolenic acid (814%). The average fatty acid content of
commercial evening primrose oils is given in Table 3. The varieties containing
up to 14% GLA have been developed (26) but not yet commercialized on a
large scale, as the majority of evening primrose oil in trade contains between
9% and 10.5% GLA. The oil contains about 98% triacylglycerols, small amounts
of other lipids (free fatty acids, diacylglycerols), and 12% unsaponifiable
matter consisting chiefly of sterols. Sterols contain about 90% b-sitosterol, 5%
citrastadienol, 1.5% gramisterol, and 1% obtusifoliol. It also contains traces of
a-, g-, and d-tocopherols (27). Evening primrose also contains traces of phenolic
compounds consisting of catechin, (-)-epicatechin, and gallic acid (28).
TABLE 3. Average Fatty Acid Profile of Oils of Evening Primrose, Black Currant,
and Hemp.
Fatty Acid Evening Primrose Oil Black Currant Oil Hemp Oil
C16:0 5.98 7.10 5.60

C16:1 0.05 0.10 0.10
C18:0 1.84 1.50 2.60
C18:1 7.20 10.90 11.50
C18:2 73.87 45.20 56.60
C18:3 (n-3) 0.28 13.20 18.50
C18:3 (n-6) 9.74 16.90 1.60
C18:4 0.07 3.30 0.50
C20:0 0.30 0.10 0.90
C20:1 0.19 0.80 0.60
C22:0 0.10 0.10 0.30
C22:1 0.20
C24:0 0.04 0.10
2.3. Black Currant

Black currant (Ribes nigrum) belongs to the family Saxifragaceae. The seeds are
byproducts of the fruit processing industry that mainly uses black currant juice
for jam, jellies, and cordial. In the early 1900s, cultivation of any Ribes species
in the United States was prohibited by a federal ban, as they are the alternative
host of white pine blister rust, a problem for all five-needle pines and the lumber
industry. The ban was rescinded in 1966, but several states continue to ban the cul-
tivation of Ribes species. Limited commercial acreage, the short harvest season
(from mid-June until mid-July), and limited access to fruit for commercial process-
ing has prohibited widespread distribution in North American. For oil extraction,
seeds and pomace left as residue from food processing industry is used to extract
the oil. Black currant oil contains about 1317% GLA, 1014% ALA, and
24% SA. The average composition of black currant oil is given in Table 3.
2.4. Hemp
Hemp (Cannabis sativa L., family Cannabinaceae) is a fast-growing annual herbac-
eous plant that is well suited to the temperate climates. The largest supply of the
world hemp production comes from China and Eastern Europe. Hemp is mainly
used for fiber, although there is a long history of food use. As a source of GLA,
it has recently become popular. The hemp contains psychoactive substances
(cannabinoids), because of which its trade was restricted. Fiber hemp that is tradi-
tionally grown for fiber contains less than 0.3% delta-9 tetrahydrocannabinol
(THC) and is classified as low THC hemp. However, no THC is present in the
seed or seed oil. Seed contains 3035% oil, of which 2.5% to 3.6% is gamma-
linolenic acid and 1520% alpha-linolenic acid (29). The average composition of
the oil is given in Table 3.
SOURCES OF GLA 75
When grown as a seed crop, a considerably lower seeding rate of 1530 kg/ha is
used as compared with a fiber crop. As hemp is a dioecious crop with male and
female plants, for adequate seed set, approximately 35% of the plants should be
male to ensure pollination (30). A few popular French monoecious varieties are also
available in the market. Seed yields can vary greatly depending on the variety used
and growing conditions but averages around 500 kg/ha and can be as high as
1200 kg/ha. In the northern climates, soil temperature should be above 10 C
when planting. Although, the once growing vigorously hemp is a good competitor
and will suppress most weeds, early establishment is a must to reduce weed pres-
sure. Pesticides are generally considered not necessary to grow hemp. Crop length
is between 3 and 4 months, depending on variety. Breeding and development work
on this crop has been limited, and the movement of certified seed is highly regulated
and expensive. Traditional breeding has concentrated on developing varieties with
superior fiber yield. Thus, there is potential to improve seed yields domestically
through breeding programs.
2.5. Echium
Echium has not been commercialized to any significant level because of regulatory
requirements for registration. The genus Echium contains about 30 species distri-
buted across Europe, the Mediterranean region, Madeira, the Canaries, and the
Azores. The plants grow in the wild and are cultivated in home gardens as flowering
plants. Echium plantagineum L. has been discovered to contain significant amounts
of GLA, ALA, and stearidonic acid in seed lipids. Echium plantagineum is also
known by the common names of Purple Vipers Bugloss, Patersons Curse, and
Salvation Jane. Agricultural production of this species is largely limited to Eastern
and parts of Western Europe at present. Trials with Echium production in Canada
have had reasonable success.
Echium plantagineum is an erect biennial 2060 cm high, softly hairy, with one
or many flowering stems. The basal leaves are ovate with prominent lateral veins
and soft appressed setae. The cauline leaves are oblong to lanceolate, the uppermost
being more or less cordate at the base. Inflorescence is usually branched. Calyx is
710 mm at anthesis, and up to 15 mm in fruit. Corolla is 1830 mm long, infun-
dibuliform is blue becoming pink through purple, and is hairy on veins and margins
only. Two stamens are exserted from corolla tube, the remaining stamens are
included or only slightly exserted, and the stigmae are distinctly bifid.
As a crop, echium is a spring-sown annual with crop duration of 3 to 3.5 months.
It requires warm, sunny conditions for quick establishment, which helps in terms of
early weed competition. The best oil quality is maintained when crop-growing tem-
peratures are around 25 C. Well-worked medium-textured soils are preferable with
adequate moisture. Seed bed should be prepared to a firm, fine, moist tilth, and the
seed should be planted 12 cm deep at the seeding rate of 35 kg/ha. Weed control
is an issue, and as a minor crop, no herbicides are currently registered for this crop.
Pre-emergence weed control might be necessary in weedy fields to reduce the weed
pressure during establishment. No significant problems have been reported on this
crop yet. Yields are highly variable, however, and usually average 300 kg/ha. Very
little breeding work has been done to date on this crop, and no varietal information
is available at this time.
Echium plantagineum occurs over significant areas of farmland in Australia (31).
The young growth is eaten readily by livestock. The plant is considered a weed in
good pastures, whereas on the poor land, it is considered a reserve fodder (32). The
level of pyrrolizidine alkaloids is normally between 0.1% and 0.3% of the dry
weight of the whole plant, but levels as high as 0.9% have been reported (33). Field
evidence strongly indicates that horses, pigs, and, to a lesser extent, sheep are all
affected. They are mainly cultivated as ornamental flowers. However, John K.
Kings and Sons, Ltd. and Croda started a program on commercial cultivation of
E. platigenium because of the presence of GLA and stearidonic acid (SA) in the
seeds (unpublished information).
Echium oil is mainly composed of a-linolenic (3033%), linoleic (1418%),
g-linolenic (1013%), stearidonic (1315%), oleic (1417%), and palmitic (67%)
acids. Like other vegetable oils, echium oil contains between 0.6% and 1.8% unsa-
ponifiable matter. The oil samples analyzed by Bioriginal Food & Science Corp.
showed an average content of 0.91% total unsaponifiable matter, campesterol
was 15.71%, beta-sitosterol was 12.53%, stigmasterol was 0.55%, and others
were 33.52%. Tocopherols constituted 8.37% of total unsaponifiable matter and
consisted of alpha- (0.53%), gamma- (6.92%), and delta-(0.92%) tocopherols.
The fatty acid profile of echium oil is given in Table 3.
Echium seeds, like other members of family boraginaceae, contain pyrrolizidine
alkaloids. The seeds contain echimidine as a major alkaloid and many minor
alkaloids, including retronecine, lycopsamine, 7-acetyllycopsamine, and their
derivatives.
3. EXTRACTION OF OIL
Of the above discussed sources of GLA, the commercial oils are mainly produced
from borage and evening primrose. Black currant oil is limited as a result of avail-
ability of seeds for oil production, whereas hemp is still subject to trade restrictions
in many countries because of the potential tetrahydrocannabinoid (THC) content.
Echium is a new crop and has not been commercialized to any significant extent
so far. All of these seeds are processed following general methods of oil extraction
common in the vegetable oil industry as discussed here.
Borage, hemp, and black currant seeds have oil content in the range of 2640%
(on 8% moisture basis); they are best suited to mechanical pressing followed by
solvent extraction. In North America, mechanically expressed borage oil is a major
item of commerce, constituting about 8090% of total sales of borage oil, whereas
in Europe, solvent-extracted oil is the major item of commerce. Mechanically
expressed oil is sold to a limited extent in the European markets, although it is gain-
ing popularity. Evening primrose seeds are small and hard and contain 1820% oil.
Therefore, they are difficult to expeller press. Black currant seeds, being a byproduct
EXTRACTION OF OIL 77
of the food industry, are usually associated with pomace and must be separated. The
separation of pomace and seed can be done either by washing with alcohol or spe-
cial cleaning processes. Washing with alcohol is not very common because of cost
and environmental regulations.
3.1. Seed Cleaning

Seed cleaning is essential to protect the quality of the oil because seeds, as obtained
from the farm, are contaminated with weed seeds, other grain seeds, and extraneous
matter. These contaminating weed seeds may impart undesirable flavors and may
negatively affect the stability of the oil. Seeds can be cleaned on the farm, but com-
mercial seed-cleaning plants are used by most oil producers. The seed cleaning
process involves aspiration of dust and lighter materials, followed by two-stage
screening to remove larger and smaller sized particles. Cleaning is done to reduce
the extraneous matter to less than 1.0%.
3.2. Expeller Pressing

The clean seed is stored in silos from where it is either conveyed to the screw-press
or to a cooker, where they are conditioned by heating to about 5090 C. The con-
ditioning helps by improving the oil yield and inactivates the enzymes (lipase) that
can affect the quality of the oil. Preheated seeds are conveyed to a continuous
screw-press where they are crushed between a stationary cage (barrel) and rotating
screw as they move forward. The pressure built up during their forward move
causes oil to be released. The screw-press is similar to those used for other seeds,
including canola and soybeans. The major manufacturers of these screw-presses
include Anderson International, French Oil Mills, Krupp, and DeSmet. There are
many smaller manufacturers of screw-presses whose equipment is better suited
for small output plants common for GLA-containing oils. The oil is pumped to sto-
rage tanks and contains between 4% and 8% fines coming from the seed. The oil is
clarified either by decantation or filtration. The press-cake so obtained contains
between 12% and 18% oil and is either used as animal feed or subjected to solvent
extraction to recover the remaining oil.
3.3. Solvent Extraction

The majority of evening primrose, black currant, and some borage seeds are extrac-
ted using this method. Prior to solvent extraction, the seed must be crushed or
flaked to rupture the cell walls, enabling better extraction efficiency at a lower
energy cost. The combination of expeller pressing and solvent extraction is a com-
mon practice for GLA-rich oils. The press-cake obtained from expeller press may
be extruded or used as is. Flaking is commonly done for evening primrose seeds
using a smooth roller mill. Food-grade hexane is the solvent of choice, although
some work has been done replacing hexane with alcohol. Alcohol contains about
5% water. During the desolventization process, alcohol is removed first, leaving
water in the oil. Removal of water from the oil adds to the energy cost and adds
additional steps in the processing, further complicating the process. The majority
of the GLA oil producers are small and low-volume entities. They often use a batch
process employing either percolation or immersion. Continuous process is also used
but poses processing challenges in that the plant has to be optimized to produce
different oils, as processing one type of seed will not be able to sustain the plant
because of low-volume requirements. In either process, the cake or flaked seeds
first come in contact with a solvent rich in oil (miscella) followed by an oil-poor
solvent and the last stage, with a pure solvent.
3.4. Supercritical Extraction

Supercritical extraction using carbon dioxide under high pressure is also becoming
popular for GLA-rich oils; however, it is not being used to any major extent because
of the high cost of the plant and the oil obtained by this technology. Some work is
being done on optimization with respect to operating conditions (34). Major
emphasis is on the flow rate of carbon dioxide, pressure, and temperature to opti-
mize the yield. The particle size of the seed pieces and the moisture content also
play a role in extraction efficiency. It is reported that the lower the moisture content,
the better the yield. The supercritical extraction usually results in an oil with similar
fatty acid composition when compared with solvent extracted oil, but the oil is low
in sterol content and may be more prone to oxidation. The biggest advantage of
supercritical extraction is that it eliminates the need for further processing of oil
such as distillation for desolventization, degumming, and so on.
3.5. Desolventization
The oil-solvent mixture and the meal is stripped of the solvent to recover solvent-
free oil and meal. The solvent-enriched meal is conveyed to vertical desolventizer
where heat and vacuum facilitate removal of solvent vapors. Desolventizer contains
trays with sweeping arms to agitate the meal for improved efficiency. Some plants
purge the cake with steam to remove the solvent, whereas others use hot air,
although application of vacuum is most common. The solvent oil miscella are
stripped of solvent in a three-stage evaporator. The hexane is reused for the extrac-
tion of oil.
3.6. Further Processing

The desolventized as well as expeller-pressed oil is further processed to reduce/
remove the pigments and phosphatides (gums). Crude oils may contain 12% phos-
phatides, which are removed by the degumming process. The degumming process
is similar to that employed in the vegetable oil industry and uses water, citric acid,
phosphoric acid or a combination of acid, and water. After the oil is contacted with
these agents, the phosphatides settle as sludge and are removed by either filtration
or centrifugation.
EXTRACTION OF OIL 79
The water degumming process involves addition of about 2% water to the oil,
intensive mixing under vacuum at 80 C for 1030 min, and filtration/centrifugation.
Water degumming removes most of the hydratable phosphatides, leaving behind
between 50 and 200 mg/kg of phosphorous depending on the extraction conditions
employed. Acid degumming using a combination of citric acid or phosphoric acid
with water also removes nonhydratable phosphatides. In this process, oil is heated
to 6080 C, and 0.10.4% citric acid or phosphoric acid is added with intense mix-
ing for 15 min. To this mixture, 2% water is added and mixing is continued for 30
60 min. After contact with water, the oil is clarified of the precipitated gums by
centrifugation or filtration using clays.
Degummed oil is further purified by physical or chemical refining. Alkali refin-
ing is rarely used in oil for the health food/dietary supplement industry, although it
may be used for oil for cosmetic/pharmaceutical applications. In processing of oil
for the health food/dietary supplement industry, water degumming and bleaching
processes may be combined when the oil is heated with water and citric or phos-
phoric acid with activated bleaching clay to 80 C in a vacuum reactor. The mixture
is intensely agitated under vacuum for 3060 min. During this time, the gums are
precipitated and are adsorbed onto the bleaching clay along with pigments and
chloroplasts. The oil mixture is cooled and filtered to remove gums, and pigments.
The resulting bleached oil has a lighter color and a phosphorous content of less than
50 ppm. As these oils also contain wax esters and other compounds that may settle
with time at room temperature and are collectively called waxes, they are subjected
to the winterization process where they are chilled to 4 C and filtered to remove the
waxes. The oil for the dietary supplement industry is not winterized. Finally, the oil
may be subjected to steam stripping (deodorization). In this process, the oil is steam
distilled to remove free fatty acids and other volatile impurities. This is the last pro-
cess in the refining of oils, and the oil is then packed in drums or totes under nitro-
gen atmosphere. The oil must be stored in a cool dry place, tightly packed in the
container under nitrogen atmosphere to protect against oxidation.
3.7. Quality Control

The GLA-containing oils are used for nutritional and health-promoting or disease-
preventive actions. They must be of high quality and free from contaminants. The
quality of oil is dependent on many factors, including seed quality and purity, herbi-
cide and pesticide residues in seed, processing and storage conditions for seed and
oil, and so on. Improper storage and drying of seeds can raise the free fatty acid
levels in seed that can result in off flavors in the oil. Being a polyunsaturated fatty
acid, GLA is prone to oxidation. The oxidation process for GLA-rich oils involves
addition of an oxygen atom at the double bond in unsaturated fatty acids leading to
formation of hydroperoxides. These hydroperoxides are unstable and decompose
to form aldehydes and ketones. These oxidation products not only impart off flavors
to the oil, making it unacceptable organoleptically, but also may have adverse
health effects. The quality of the oil is tested by checking for peroxide value, an
indicator of primary oxidation product. Oil with a peroxide value of less than
TABLE 4. General Specifications of GLA-Rich Oils.
Parameter Units Value
Peroxide Value meq/kg <5

Anisidine Value <15
Acid Value (unrefined oils) mg KOH/g <4
Acid Value (refined oils) mg KOH/g <0.7
Unsaponifiable matter % <2
Pesticides/Herbicides mg/kg <0.05
Solvent residues ppm <1.0
Color (Lovibond 1 inch) <3 red
Heavy metals ppm <10
Lead ppm <0.1
Mercury ppm <0.1
Cadmium ppm <0.1
Arsenic ppm <0.1
Evening
Major Fatty Acid (% of total fatty acids) Primrose Oil Borage Oil Black Currant Oil
Oleic acid (C18:1) 69% 1422% 915%

Linoleic acid (C18:2) 7077% 3238% 4050%
Gamma Linilenic acid (C18:3, n-6) 812% 1825% 1519%
Alpha Linolenic acid (C18:3, n-3) 0.11.0% 0.12.0% 1215%
Stearidonic acid (C18:4) 0.10.3% 0.10.3% 25%
10 milliequivalent of KOH/kg oil is considered good for consumption. Peroxide

value alone is not a good indicator of oxidative stability of oil as it measures the
primary oxidation products, which degrade to secondary oxidation product, includ-
ing aldehydes and ketones. These secondary oxidation products can be measured by
several methods, including conjugated dienes, anisidine value, and so on. In addi-
tion to oxidative stability indices, the oils are tested for fatty acid profile to ensure
the quality and purity of the oil. The presence of free fatty acids is tested by acid
value. The free fatty acid content of the oil should be as low as possible. These oils
are also tested for heavy metal contamination. The total content of heavy metal
should be less than 10 ppm. The oils should be free of any herbicide or pesticide
residues and the solvent used in extraction of oil. The quality parameters for these
oils are listed in Table 4.
4. METABOLISM OF GLA
When GLA-rich oils are taken orally, GLA is rapidly absorbed. It first appears in
serum phospholipids, and with continuous administration, it is distributed in other
phospholipid fractions. Part of absorbed GLA is oxidized, and the rest is taken up
by various tissues/cells and is rapidly elongated to dihomogammalinolenic acid
(DGLA) (Figure 1). The oxidation rate of GLA was found to be 28% of that for
METABOLISM OF GLA 81
LA (35). DGLA can be acetylated and incorporated into membrane phospholipids,

or it can be desaturated to AA by delta-5-desaturase. DGLA competes with AA for
cyclooxygenase (COX) and lipoxygenase (LOX) enzymes. DGLA produced pros-
taglandins of series 1 (PGE1) and thromboxane A1 (TxA1) by the action of COX.
These products of COX action exert anti-inflammatory, vasodilatory, and anti-
aggregatory actions. DGLA produces 15-hydroxyeicosatrienoic acid (15-HETrE)
by the action of 15-lipoxygenase. 15-HETrE is a strong inhibitor of 5-lipoxygenase,
whereby it inhibits production of leukotriene B4 (LTB4) from inflammatory cells,
including neutrophils (36).
Hassam et al. (37) were the first to study the absorption and metabolism of GLA
using 14C-labeled GLA in rats. They observed accumulation of labeled DGLA and
AA in brain and liver after 22 hours of administration, suggesting that GLA is
rapidly metabolized to DGLA and AA. Leyton et al. (35) reported that DGLA is
preferentially incorporated in liver phosphoacylglycerols, mainly in choline and
inositol phosphoacylglycerols. Feeding a GLA-rich diet to rats caused accumula-
tion of DGLA in milk (38) and a rise in DGLA and AA in aorta and platelets
(39) and in immune cells, including macrophages, kupfer cells, and endothelial
cells (40, 41). Barre et al. (42) observed a rise in GLA and DGLA with no change
in AA levels in different platelet phospholipid fractions in human volunteers fol-
lowing daily administration of 5.23-g GLA from borage oil for 42 days. There
was a differential distribution of DGLA in various phospholipid fractions with
phosphatidylcholine had maximal (67.6%) followed by phosphatidylethanolamine
(16.7%), phosphatidylserine (12.9%), and phosphatidylinositol (2.6%). There was
no change in sphingomyelin. In all phospholipid fractions, the ratio of DGLA/
AA decreased significantly. In a later study, they observed a rise in GLA and
DGLA levels in phosphatidylcholine fraction of plasma HDL and cholesteryl
esters. AA levels increased only in phosphatidylcholine fraction of HDL (43). In
these studies, the dose of GLA employed is much higher than used in any of the
clinical trials. The difference in the observed rise in platelet AA levels after feeding
of GLA sources in above studies is caused by species difference. Rat platelets have
delta-5-desaturase enzyme required for conversion of DGLA to AA, whereas
human platelets lack this enzyme and obtain preformed AA from the circulation.
In one study on six healthy volunteers, time of administration of GLA-rich oil
was found to affect the peak serum levels of GLA (44). Administration of evening
primrose oil (equivalent to 240-mg GLA) in the evening caused a rapid peak in
serum levels (2.7 1.2 hours) compared with administration in the morning
(4.4 1.9 hours). There was a small but insignificant increase in serum DGLA
and AA levels. In this study, the second dose of evening primrose oil was given
12 hours after the morning dose. This might have contributed to the observed rapid
rise in peak serum GLA levels, or this could reflect faster absorption of GLA in the
evening. A rapid rise in plasma triacylglycerols level after the second meal has been
observed. Manku et al. (45) studied the effect of feeding evening primrose oil (con-
taining GLA) for a period of 10 days to 12 weeks on plasma fatty acid levels. In this
study, they collected the blood samples of 392 individuals who were part of 20 dif-
ferent studies. In all of these studies, DGLA levels in plasma phospholipids were
increased significantly. In 17 of these studies, there was a small but significant rise
in phospholipid AA levels, whereas in 3 studies, there was no rise, and in 2 of these
studies, there was a fall in AA levels. In these 3 studies, EPO was administered for
10 days only. In all of the studies, the ratio of AA/DGLA fell, suggesting a greater
rise in DGLA levels. These results indicate that, in humans, DGLA is slowly de-
saturated to AA. Feeding borage oil for 7 weeks to normotensive (WKY) and spon-
taneously hypertensive (SHR) rats resulted in an increase in GLA and DGLA levels
in plasma, liver, aorta, and renal artery in both strains of rats, although AA was
increased only in plasma and liver (46). These observations indicate that there is
a tissue-specific rise in AA after administration of GLA.
From the above discussion, it is clear that GLA is rapidly absorbed and elonga-
ted to DGLA. DGLA levels increase in most of the tissues after GLA administra-
tion, but the levels of AA rise to a smaller extent mainly in the liver. The capacity of
other tissues to desaturate DGLA is limited and depends on the species. Chilton
et al. (47) studied the effect of in-vivo administration of GLA and in vitro incuba-
tion of human neutrophils with GLA on metabolism of GLA. They observed that
in vivo administration of GLA to humans caused an increase in DGLA in the
neutrophils and no GLA was detected. Incubation of neutrophils with GLA resulted
in a rise in the DGLA concentration of neutrophils. Stimulation of these neutrophils
with ionophore A23187 caused a release of AA and DGLA from neutrophil phos-
pholipids. DGLA was metabolized to 15-HETrE that inhibited LTB4 production
with an IC50 of 5 mM.
4.1. Effect of Triacylglycerol Structure on Bioavailability of GLA

Major sources of GLA include borage oil, evening primrose oil, and fungal oils.
GLA is mainly distributed at sn-2 position in triacylglycerols in borage oil, at
sn-3 position in black currant and evening primrose oils, and at sn-1 and sn-3 posi-
tions in fungal oils (48). Evening primrose oil was reported to provide higher levels
of GLA compared with borage oil in rats, although the latter oil contains a higher
amount of GLA/g. This was a surprising finding and attributed to positional differ-
ences for GLA in the triacylglycerol structure and the inability of gastric and pan-
creatic lipases to hydrolyze fatty acids at sn-2 position. The fatty acid in the sn-2
position of triacylglycerols is preferentially absorbed as the 2-monoacylglycerol
and serves as the template for re-esterification by intestinal cells to reform triacyl-
glycerols. The sn-2 fatty acids are also preferentially preserved as components of
chylomicrons and very-low-density lipoprotein particles for ultimate incorporation
in tissue membranes. Subsequently, Raederstroff and Moser (49) repeated the stu-
dies in rats and failed to reproduce similar results. They observed that the levels of
GLA and DGLA in liver, aortic, and erythrocyte phospholipids reflected the amount
of GLA present in the source oil. This indicated that different oils were well
absorbed and that the amount of GLA absorbed was dose dependent, and the source
of GLA did not matter. To further resolve this matter, Chung et al. (50) studied the
efficacy of borage oil, evening primrose oil, or a combination of borage oil with
CARDIOVASCULAR EFFECTS 83
safflower oil to match the GLA content to evening primrose oil in reversing the epi-
dermal hyperproliferation induced by essential fatty acid deficient diets. In this
study, they observed that GLA-rich diets reversed epidermal hyperproliferation
caused by essential fatty acid deficiency and the potency order was borage oil great-
er than boragesafflower oil combination, greater than evening primrose oil. Final-
ly, two diets had similar amounts of GLA, but they differed in the structural
location of GLA on triacylglycerol molecule. There were higher levels of DGLA
in epidermal phospholipids and ceramides on the borage oil or the borage oil and
safflower oil diet than from the evening primrose oil diet. They proposed that
borage oil, being richest in GLA at sn-2 position, is more bioavailable; hence,
borage oil was more potent. Higher bioavailability of GLA at sn-2 position is
also supported by data from other laboratories studying the effect of triacylglycerol
structure on fat digestion and absorption. During digestion, gastric and pancreatic
lipases hydrolyze fatty acids at sn-1 and sn-3 position forming free fatty acids and
sn-2 monoacylglycerols. The absorption of free fatty acids is reduced in the pre-
sence of divalent ions (calcium and magnesium) because of soap formation,
whereas sn-2 monoacylglycerols are favorably absorbed. The differences in the
results obtained by these two groups [Chung et al. (50) and Raederstroff and Moser
(49)] could be caused by species differences, or by a difference in study design, as
Chung et al. (50) performed the studies in the essential-fatty-acid-deficient guinea
pigs, or by tissue differences.
5. CARDIOVASCULAR EFFECTS
Cardiovascular disease is a major cause of mortality and morbidity in industrialized

countries. Several risk factors have been linked to incidence of cardiovascular dis-
ease and include hypertension, lipid abnormalities (high plasma cholesterol and
triacylglycerol levels), atherosclerosis, obesity, diabetes, smoking, stress, heredity,
and diet. Dietary GLA affects many of these parameters and is discussed below.
5.1. Effect on Blood Pressure

Arterial blood pressure is regulated by the interaction of cardiac output and peri-
pheral vascular resistance. Several factors can influence these interactions, and
they can include renin-angiotensin system, local metabolic factors, stress hormones,
and so on. Interventions that interfere with these modulators can affect the blood
pressure regulation. In 1975, Rose et al. (51) observed a biphasic response of
intravenously administered DGLA on systemic arterial pressure in dogs that was
characterized by an initial fall in blood pressure followed by a sustained fall and
an increase in myocardial contractility. Only the sustained fall in blood pressure
was blocked by cyclooxygenase inhibition, whereas the early fall in blood pressure
and positive inotropic effects were not affected, suggesting that DGLA causes a
blood pressure-lowering effect directly and through PGE1 pathways. In 1982, even-
ing primrose oil was shown to inhibit the blood pressure-increasing activity of
intravenously administered renin and angiotensin II in rats given evening primrose

oil for 3 months (52). This observation suggested that GLA-rich oils may reduce the
blood pressure by interfering with the renin-angiotensin system in the body. GLA-
inhibited isolation (psychological) stress-induced rise in blood pressure in rats
when administered at a dose of 0.018 or 0.04 mg/hour via an osmotic pump
(53). In the unstressed rats, there was no effect of GLA on blood pressure. No effect
on heart rate, heart weight, or adrenal weight was observed in any animal. Mills
et al. (54) repeated the experiments on humans to observe if GLA has similar
actions on stress reactivity and performance. They selected 30 normotensive
male university students for the study and divided into various groups. One group
n 10 received olive oil capsules for 28 days, and another group received borage
oil capsules n 10 providing 1.3-g GLA per day. These volunteers were given
Stroop color word conflict test before commencement of supplement therapy and
after 28 days of supplementation. Borage oil supplementation significantly reduced
the stress-induced rise in systolic blood pressure and heart rate and did not affect
diastolic blood pressure or plasma norepinephrine levels. Borage oil treatment
increased the skin temperature and the performance as compared by number of cor-
rect responses. These data confirm the observations obtained earlier in rats and indi-
cate increased tissue perfusion by borage oil treatment. Leng et al. (55) also
observed a blood pressure-lowering effect in patients with peripheral arterial dis-
eases. In their study, they used a combination of GLA with EPA, so the probable
contribution of EPA to blood pressure-lowering effect cannot be entirely ruled out.
The exact mechanism of blood pressure-lowering effect is not very clear, and
GLA-rich oils appear to act via several mechanisms. Borage (56) and evening prim-
rose oils (57) were shown to reduce in vivo pressor responses to angiotensin-II and
norepinephrine without affecting in vitro contractile response of aorta to potassium
chloride and serotonin in rats. These observations suggest that GLA may be inter-
fering with agonist-receptor interactions without affecting the contractility of vas-
cular smooth muscles. Subsequent studies in spontaneously hypertensive rats
demonstrated the blood pressure-lowering effect of borage oil (58) without affect-
ing the pressor response to angiotensin and norepinephrine, suggesting the role of
other mechanisms. These findings suggest that there may be a species difference in
responsiveness to angiotensin II and norepinephrine, although the blood pressure-
lowering effect was similar in magnitude. GLA was shown to prevent development
of hypertension in SHR rats (59), which could have been mediated via the cyclo-
oxygenase pathway as an increase in aortic levels of 6-keto PGF1a was observed. In
hypertensive rats, GLA was shown to significantly reduce the ratio of plasma aldos-
terone to renin that was caused by a insignificant decrease in plasma aldosterone
levels and a small increase in plasma renin activity (60). There was no effect of
borage oil treatment on plasma cortisol levels compared with rats fed control
diet free of GLA. Borage oil treatments also reduced angiotensin receptor number
and affinity in SHR rats, suggesting a reduction in the responsiveness of adrenal
glomerulosa cells to angiotensin and interference with the reninangiotensinaldos-
terone axis might contribute to the hypotensive effects. These studies cannot alie-
nate the exact mechanism by which borage oil interferes with angiotensin receptors.
CARDIOVASCULAR EFFECTS 85
Mills et al. (61) studied the effects of dietary borage oil on baroreflexes in normo-
tensive, healthy males. These males were subjected to lower body negative pressure
of 10 and 40 mm Hg. A negative pressure of 10 mm Hg unloads cardiopul-
monary baroreceptors, whereas the negative pressure of 40 mm Hg unloads both
cardiopulmonary and arterial baroreceptors. They observed that borage oil treat-
ment augmented the baroreflex response to 40 mm Hg without affecting the
response to 10-mm Hg negative pressure, suggesting that GLA may be affecting
only high-pressure arterial baroreflex responses. This could be mediated either
by altering the sensitivity of baroreceptor stimulus-response relationship or by shift-
ing the operating point of the reflex to a much steeper point on the baroreceptor
stimulus-response relationship curve. In human hypertension, baroreceptor res-
ponses are decreased, which may be contributing to structural changes in hyperten-
sive patients.
5.2. Platelet Function and Plasma Lipids

Increased levels of plasma triacylglycerols and cholesterol and platelet dysfunction
(increased aggregation) are independent risk factors for cardiovascular disease.
The effects of GLA on blood lipids and platelet function are controversial.
Chaintreuil et al. observed a fall in serum triacylglycerols and cholesterol levels
in insulin-dependent diabetic patients administered 2 g/day GLA, but not with
500 mg daily dose for 6 weeks (62, 63). In hypertriglyceridemic patients, GLA
had no effect on plasma triacylglycerol levels or platelet function, although there
was an increase in GLA and DGLA levels in plasma and platelet phospholipids
(64). Viikari et al. (65) also failed to observe the lipid-lowering effect of evening
primrose oil in hyperlipidemic subjects in an open study. They continued adminis-
tration of evening primrose oil for 3 months but changed the dose every month from
2.4 ml (first month) to 7.2 ml (third month). They observed a rise in GLA levels in
serum cholesteryl esters, phospholipids, and triacylglycerols. The differences in the
results of above studies could be attributed to dose differences. Guivernau et al. (66)
fed GLA at a dose of 240 mg/day for 12 weeks to 12 hypertriacylglycerolmic pati-
ents and 12 rats. They observed a significant decrease in plasma triacylglycerols,
total cholesterol, and LDL cholesterol and an increase in HDL-cholesterol. Reac-
tivity of platelets to low doses of adenosine diphosphate and epinephrine was sig-
nificantly reduced. A reduction in plasma thromboxane B2 levels was also observed
in humans. In rats, a rise in plasma 6-keto-PGF1a levels was observed, suggesting
an increase in PGE1 production by GLA administration. Changes in eicosanoids
may contribute to the observed effects of GLA on platelet aggregation as throm-
boxane B2 is a potent platelet aggregator. GLA is rapidly metabolized to DGLA,
and DGLA has been shown to inhibit platelet aggregation in in vitro (67) and
in vivo studies (68, 69).
Ishikawa et al. (70), in a double-blind, cross-over trial in hypercholesterolemic
patients, demonstrated that GLA lowered low-density lipoprotein cholesterol
and apolipoprotein B in plasma and increased HDLC levels without affecting
the levels of total cholesterol. Jantti et al. (71) observed a decrease in plasma
apolipoprotein B concentrations in rheumatoid arthritis patients given evening

primrose oil at a dose of 20 ml (about 1.8 g GLA) per day for 12 weeks. In this
trial, no effect on plasma triacylglycerols or total or high-density lipoprotein
cholesterol was observed. Horrobin and Manku (72) found that evening primrose
oil exerted cholesterol-lowering effects in people with plasma cholesterol levels
above 5 mmol/l but had no effect in people having plasma cholesterol levels lower
than 5 mmol/l. Fukushima et al. (73) fed conventional diets enriched with 10%
borage oil, palm oil, perilla oil, evening primrose oil or mixed oils, and 0.5%
cholesterol for 15-week to 8-week-old rats. GLA-rich diets lowered plasma total
cholesterol and the sum of LDL, IDL, and VLDL cholesterol. Cholesterol-lowering
effects of a GLA-rich diet could be mediated by changes in membrane lipid com-
position affecting absorption of cholesterol. This observation is confirmed by Koba
et al. (74) in Cacao cells. When these cells were incubated with GLA, the absorp-
tion of cholesterol from the growth medium was attenuated and the cell membranes
were enriched with GLA, DGLA, and AA.
5.3. Atherosclerosis
Atherosclerosis is the most common cause of morbidity and mortality in pati-
ents with cardiovascular diseases. The exact cause of atherosclerosis is not clear.
Atherosclerosis is a culmination of several events, including vascular dysfunction,
which may be caused by an injury to vasculature, recruitment of inflammatory cells
including monocytes and neutrophils, activation of macrophages, vascular smooth
muscle cell proliferation, deposition of lipids, and synthesis of extracellular matrix.
Oxidized low-density lipoprotein cholesterol plays a role in initiation of atherogen-
esis. It stimulates monocytes with the resultant formation of foam cells. These cells
release mediators that stimulate expression of adhesion molecules like cadherin,
vcams, and so on. Macrophages, on stimulation, release eicosanoids and cytokines
that may stimulate proliferation of vascular smooth muscle cells. Proliferation of
vascular smooth muscle cells appears to be a central event in atherogenesis. Essen-
tial fatty acids are substrate for the production of eicosanoids, and the membrane
composition of inflammatory cells reflects dietary intake of various fatty acids. It
appears that dietary manipulation of the composition of cell membranes is the
easiest target to control atherogenesis. Renaud et al. (75) demonstrated that dietary
polyunsaturated fatty acids, including GLA, reduced severity of atherosclerotic
lesion in rabbits compared with saturated-fatty-acid-rich diets. In Japanese quail,
dietary primrose oil was shown to inhibit atherogenesis (76). Fan et al. (77) observ-
ed inhibitory action of dietary evening primrose oil either alone and in combination
with fish oil on aortic smooth muscle cell proliferative action of peritoneal macro-
phages from mice. The inhibitory action appeared to be mediated through cyclo-
oxygenase pathway as indomethacin (cyclooxygenase inhibitor)-inhibited PGE1
release and antiproliferative actions. Addition of 5-lipoxygenase inhibitor to the
culture medium had no effect on antiproliferative or DNA synthesis inhibitory
actions of primrose oil. In vitro incubation of endothelial cells with PUFAs,
including GLA, AA, ALA, EPA, or DHA, stimulated the oxidation of LDL and
CANCER 87
metabolism of oxidized LDL by macrophages (78). These interventions also

increased the release of superoxide anions by endothelial cells. These observations
suggest pro-atherosclerotic actions of PUFAs in humans. In apolipoprotein E
knockout mice, evening primrose oil inhibited aortic smooth muscle cell prolifera-
tion and reduced the aortic vessel wall medial layer thickness and the size of ather-
osclerotic lesion (79). This study confirms the beneficial effects of GLA in lowering
cardiovascular risks by inhibiting atherosclerotic plaque development.
5.4. Cardiac Arrhythmia

Several studies have demonstrated that LA (present in vegetable oils) exert antiar-
rhythmic activity in several models, including ischemic-reperfusion injury and
catechol-induced arrhythmias. Li et al. (80) observed that PGE1 and PGI2 exert
antiarrhythmic activity in cultured, spontaneously beating neonatal rat cardiac myo-
cytes, while PGD2, PGE2, PGF2a, and TXA2 exert proarrhtymic activity. Charnock
et al. (81) studied the effects of evening primrose oil and black currant oil on ven-
tricular fibrillation in rats induced by ischemia. They compared the effects of these
two oils to sunflower oil (a source of LA) and sheep fat (saturated fat). They
observed that, compared with the saturated fat group, all other dietary treatments
significantly reduced the number of premature ventricular beats, however, there
was no difference between the three PUFA groups. The effect on duration of ven-
tricular fibrillation was dependent on diet with saturated fat showing the longest
duration that was significantly reduced by the other three oils and the potency order
of these three oils was sunflower oil hevening primrose oilhblack currant oil. Eve-
ning primrose oil contains similar amounts of LA but additional amounts of GLA,
suggesting GLA might have been playing an additional protective role. As black
currant oil contains twice the amount of GLA and additional ALA, it is difficult
to assess from this study if additional protection provided by black currant oil
was caused by high amounts of GLA or synergistic action of GLA and ALA.
6. CANCER
Cancer is a collective term that defines a group of conditions caused by excessive

growth of cells in any organ/tissue. It can occur in any part of the body. It is a com-
plex phenomenon, the etiology of which is not very well understood. Risk of cancer
increases with age, and about 77% of cancers are diagnosed in people after 55 years
of age. Risk factors for cancer include lifestyle factors (diet, tobacco, excessive
alcohol use, and physical inactivity), radiations, chemicals, infections, heredity
(inherited mutations), immune conditions, obesity, and hormones. Heredity
increases the predisposition to cancer but in itself is not responsible for initiation
of cancer and requires interaction with other factors. About 510% of total cancers
are hereditary because of inheritance of mutated gene. According to the American
Cancer Society (http://www.cancer.org/downloads/PRO/12), about 1.37 million
new cases of cancer are expected to be diagnosed in 2004. This estimate excludes
basal and squamous cell carcinoma of skin and carcinoma in situ of any site except
urinary bladder. About 563,700 people are expected to die from cancer in the
United States in 2004. Of these deaths, 170,000 deaths will be related to tobacco,
and a similar number of deaths will be caused by nutrition, obesity, physical
inactivity, and other lifestyle factors.
Basic treatment for cancer includes chemotherapy, radiations, and surgery.
Strategies for prevention include modification of lifestyle factors and dietary
interventions. The role of dietary fat in cancer is controversial. Many prospective
studies found an increase in cancer risk (8284), whereas others reported no
association between fat intake and cancer (8587).
GLA has been studied in several studies for its effects on various cancer cell
lines in vitro. It has been observed to exert cytotoxic activities against several tumor
cell lines in vitro and tumor implants in experimental animal models. There are
limited studies on the effect of GLA on tumors in humans. In cell lines, the effect
of GLA appears to depend on the cell line, dose, and incubation time. In a study by
Dippenaar et al. (88), GLA caused significant (up to 70%) growth inhibitory effects
on mouse BL6 melanoma cells in vitro at a dose of 20 mg/ml. At this dose, GLA did
not affect the growth of normal bovine kidney epithelial MDBK cells, suggesting
that GLA acts as an anticancer agent and inhibits the growth of cancer cells without
affecting the normal cells. Human hepatoma cell lines differ in sensitivity to GLA
as they require continuous presence of GLA in culture media for 4 days to observe
growth inhibitory effects (88); withdrawal of GLA from the growth media after
5-day treatment suppressed the growth for 5 more days (89). This observation
suggests that cancer cells may lack delta-6-desaturase and, hence, cannot make
GLA and, subsequently, DGLA. Cancer cells incorporate GLA and DGLA in their
cell membranes and DGLA may be acting via a cyclooxygenase pathway in inhi-
biting cancer cell growth as PGE1 stimulates cyclic-AMP formation and induces
cell death in cancer cell lines (90). In 1985, Begin et al. (91) confirmed that
GLA has growth inhibitory actions against human prostate, breast, and lung cancer
cells with no effect on normal cells.
Experiments were conducted to study the effects of GLA treatment on
carcinogen-induced cancers in animals. Lee and Sugano (92) failed to observe
any tumor inhibitory action of evening primrose oil in pathogen-free female Spra-
gue Dawley rats in whom the tumor was induced by intragastric administration of
10 mg of 7,12-dimethylbenz(a)anthracene (DMBA) one week after animals were on
experimental diets containing 5% evening primrose oil, sunflower oil, or palm oil.
In another study, 50-day old female rats (Sprague Dawley) were given either 5 mg
or 10 mg 7,12-dimethylbenz(a)anthracene (DMBA) intragastrically to induce
mammary tumors (93). On 14-(5 mg DMBA rats) or 21-(10 mg DMBA rats) day
post-DMBA adminstration, rats were divided into two groups and were fed a high-
fat diet containing either 20% evening primrose oil or 20% corn oil (93). The group
of rats on the evening primrose oil diet had significantly lower number of rats bear-
ing tumors, and malignant tumors. Linoleic acid content of the primrose oil diet
was higher than that of corn oil diet and linoleic acid has been linked to promote
mammary tumorigenesis in rats and mice. The two diets differ in GLA only, which
CANCER 89
suggests that GLA may be responsible for the tumor-inhibiting effects of the even-
ing primrose oil diet. The different results of the above two studies (92, 93) can be
due to differences in the dose of GLA given to rats. The other difference could be in
the immune status of the rats, as Lee and Sugano conducted their studies on patho-
gen-free rats. Gonzalez et al. (94) performed a case controlled study in 4 regions of
Spain investigating the association of dietary factors and risk of gastric cancer.
Zaragova is an area in Spain where people eat borage leaves and stem, usually
cooked by boiling in water. After adjusting for intake of fruits and vegetables
and caloric intake, a strong negative association was observed between risk of gas-
tric cancer and borage intake. The negative association showed a strong dose-
response effect, when the population was subdivided into quartiles. On analysis,
they found that boiled borage leaves contained about 4.4% GLA, while boiled
stems contained 14.6% GLA. This is the first study on association between dietary
borage consumption and risk of gastric cancer. As very few populations are habitual
borage eaters, it is difficult to repeat the studies and also this study cannot definitely
link GLA as an anticancer agent in borage leaves and stems.
To confirm if the cytotoxic effects of GLA are, in fact, mediated by prostaglan-
din pathway, Botha et al. (95) cultured human breast carcinoma cell line NUB1
with 50 ml GLA or DGLA and studied the effects on prostaglandin production
and cell growth. They observed that GLA had inhibitory actions on NUB1 cell
growth that were accompanied by an increase in production of prostaglandin
E and F. On the other hand, DGLA caused a significantly higher increase in the
level of these prostaglandins but had no effect on cell growth, indicating GLA
exerts cancer-cell growth-inhibitory actions by some other mechanisms. Kenny
et al. (96) co-administered 2.8 g GLA with 20 mg tamoxifen to 38 breast cancer
patients. The control group consisted of 47 breast cancer patients on 20 mg tamox-
ifen only. They observed that GLA acted synergistically with tamoxifen in reducing
the expression of estrogen receptors in tumor cells and enhanced the efficacy of
tamoxifen. GLA tamoxifen group of patients showed early response to therapy
and had significantly better quality of life by 6 weeks on therapy. GLA treatment
was well tolerated with 42% of patients reporting no side effect and a general
feeling of well being, 34% of patients reported alterations in the bowel habits
with a tendency towards loose stool (many elderly patients found this beneficial).
In early responders, the GLA group had a much higher reduction in expression of
estrogen receptors (ER) than tamoxifen alone. The GLA group also had downregu-
lation of expression of bcl-2 gene at 6 weeks, compared with no effect or transient
increase in bcl-2 protein in the tamoxifen group. As bcl-2 plays a role in prevention
of apoptotic cell death, this observation suggests that, by reducing the expression of
antiapoptotic protein, GLA stimulates apoptotic cell death in cancer cells, which
may have contributed to faster response at 6 weeks.
GLA has been shown in experimental model of cancer to inhibit metastasis of
cancer. Urokinase concentration is increased in malignant cancer cells, and it is
reported to play a role in invasiveness and metastasis of cancer. du Toit et al.
(97) studied the effect of GLA on urokinase activity. They observed that GLA is
a competitive inhibitor of urokinase activity with a Ki value of 120 mM. In a
subsequent study, they observed that GLA inhibited production of urokinase activ-
ity in human prostate tumor (DU-145) cells (98). These observations suggest that
GLA, by inhibiting urokinase activity, may be playing a role in preventing metas-
tasis of cancers. Jiang et al. (99) studied the effect of GLA on motility and invasive-
ness of three colon cancer cell lines (HT115, HT29, and HRT18) induced by
hepatocyte growth factor. They observed GLA and its lithium salt reduced metas-
tasis and invasiveness of all the cancer cell lines by upregulating expression of
E-cadherin and inhibiting attachment of cancer cells to fibronectin without affecting
fibronectin receptors. Dissociation of tumor cells from the main tumor is the first
requirement for metastasis. By increasing the expression of E-cadherin, GLA
increases the adhesiveness of tumor cells, so the incidence of metastasis is reduced.
In subsequent studies, they further demonstrated reduced metastasis and increased
adhesion of tumor cells that are E-cadherin negative (HT115 and MDA-MB 231)
suggesting that other mechanisms play a role in reducing the invasiveness of cancer
cells. They reported increased formation of desmosomes by increasing the expres-
sion of desmoglein. As desmosomes play a role in cellcell adhesion, this observa-
tion indicates a role of GLA in preventing metastasis by increasing the adhesiveness
of tumor cells so they fail to dissociate and, hence, metastasize. At the same time,
GLA inhibits cell-matrix interaction and the exact mechanism is not clear. Integrins
play a major role in cell-matrix interactions. GLA has been shown to inhibit this
interaction at several stages by inhibiting focal adhesion kinase activation and paxi-
lin activation. Both of these molecules are activated by tyrosine phosphorylation,
which is inhibited by GLA in tumor cells. GLA also upregulates expression of
metastasis suppressor nm-23 gene (100). A reduction in the level of nm-23 gene
expression has been reported in various cancers, including colorectal, breast, liver,
ovarian, and bladdeer cancers. These studies indicate that GLA may act on different
targets at the gene level to reduce metastasis and invasiveness of cancers. Jiang et al.
(101) demonstrated that GLA may be acting through activation of peroxisome
proliferator activated receptor-gamma (PPAR-g) through increased phosphorylation
of these receptors. On phosphorylation, these receptors are translocated to the
nuclear membranes and regulate the expression of various genes. They demon-
strated that removal of PPAR-g with antisens oligos abolished the effect of GLA
on expression of adhesion molecules and tumor-suppressor genes.
6.1. Prostate Cancer

GLA has been shown to inhibit 5a-reductase activity in androgen-sensitive
(LNCaP) and androgen-insensitive (PC3) human-prostate cancer-cell lines (102).
This observation may suggest that GLA could be acting as an anticancer agent
against androgen-dependent prostate and skin cancers.
6.2. Glioma
Patient suffering from malignant cerebral glioma are treated aggressively with radi-
ation, chemotherapy, and surgery, although surgery is the first option combined with
CANCER 91
the other two treatments. The median survival time after aggressive treatment is
about one year (103, 104). Naidu et al. (105) treated six patients suffering from
histochemically confirmed malignant glioma with GLA. Of these patients, four
patients received 1 mg GLA daily for 10 days, whereas the other two patients
were treated only on alternate days. Treatment started 10 days after surgery; all
these patients demonstrated marked necrosis of tumor immediately after the ther-
apy. Of these six patients, three were alive after two years, whereas two were lost to
follow-up and one died. No side effect of therapy was observed during or after
treatment. During subsequent follow-up, authors did not observe any increase in
size of residual tumor or recurrence of tumor. Based on the results of this study,
authors extended the treatment to 15 more patients and found increased survival
by one and one-half to two years. This study also confirmed necrosis of tumor cells
and safety of GLA. They also injected GLA to normal dogs intracerebrally and
found no cytotoxic effects (106). These studies demonstrated that GLA injected
directly into tumor mass may potentially be useful treatment for malignant glioma.
6.3. Liver Cancer

Merve et al. (107) conducted a double-blind placebo-controlled trial of evening
primrose oil in patients suffering from primary liver cancer, a fatal disease. The pati-
ents were randomly assigned to the GLA or placebo group. The GLA group patients
received 36 capsules per day supplying 18 g evening primrose oil containing 1.44 g
GLA. The control group received the same amount of olive oil. They observed a
mean survival time of 58 days in the treatment group compared with 42 days in
the placebo group, although the difference was not statistically significant. Gamma
glutamyl transaminase enzyme activity was decreased in seven patients in the treat-
ment group compared with two patients in the placebo group. This difference was
statistically significant, suggesting that evening primrose oil may have some effect
on tumor. In this study, patients had up to 3-kg tumor weight, suggesting an
advanced stage of cancer. Probably, the dose of GLA was not sufficient to obtain
a statistically significant effect on survival time. A major finding was that the qua-
lity of life was better for the evening primrose oil group as indicated by the patients
self assessment. Falconer et al. (108) studied the effect of lithium salt of GLA on
pancreatic cancer in 18 patients who had unresectable pancreatic cancer and had
undergone either surgical bypass or had pancreas endoscopically stented. These
patients were administered GLA intravenously for 10 days and then were switched
to oral GLA therapy. During the infusion period, the dose of GLA was gradually
increased for the first five days and then continued at maximal tolerated dose for
a subsequent 5 days. Patients received a mean dose of 5.7g lithium GLA for the
last 5 days and mean oral dose of 3 g afterwards. They observed a median survival
of 8 months and 4 patients were still alive compared with normal life expectancy of
36 months for these patients. GLA treatment increased T-cell function and reduced
TNF production. In this report, the study design was not well defined; therefore, it
was difficult to assess if the protocol had any beneficial effect on patient survival,
though the treatment was reported to be well tolerated.
6.4. Mechanism of Anticancer Effects of GLA

The exact mechanism of anticancer effects of GLA is not very clear and may
depend on the cancer type. Many cancer cells have been shown to lack phospho-
lipase A2 (PLA2) activity and delta-6-desaturase activity. PLA2 is essential in
releasing free fatty acids from the membrane phospholipids. Released free fatty
acids, like DGLA, AA, EPA, etc, act as a substrate for cyclooxygenase and lipoxy-
genase enzyme to produce prostaglandins and leukotrienes. Delta-6-desaturase is
essential for the conversion of dietary LA to GLA. Therefore, administration of
GLA can bypass these metabolic steps and show anticancer effects. GLA, being
a polyunsaturated fatty acid, can increase lipid peroxidation in the cancer cells.
Free radicals have been implicated in cytotoxic actions of several anticancer drugs.
It is possible that GLA may be showing its anticancer effects through oxidative
mechanisms (109). Leaver et al. (110) analyzed the effect of GLA and AA on
free radical production and cell death by necrosis and apoptosis in 30 human glioma
types. The brain samples were obtained from patients undergoing surgery. Patients
had grade 1 to grade 4 tumors. They observed that tumor cells in general produced
less free radicals than normal cells and amongst the tumor cells, total free radical
production was higher for advanced tumors (Grade 4). GLA and AA, both increas-
ed the production of free radicals in normal and tumor cells; however, tumor cells
responded with a much higher increase in the production of free radicals and GLA
was more potent than AA in increasing the free radical production in glioma cells.
In this study, the necrotic cells produced less free radicals than nonnecrotic tumor
cells and they showed a lower degree of rise in free radical production when incu-
bated with GLA or AA. As necrotic cells are rich in phagocytic cells, this observa-
tion suggests that GLA or AA increase the production of free radicals in tumor cells
mainly and the phagocytic cells are not the major source of free radical in gliomas
incubated with GLA or AA. This observation also indicates that GLA is free from
toxic effects on healthy cells in contrast to cancer therapeutics probably because it
does not promote formation of free radicals from phagocytic cells that may release
free radicals at several sites and damage healthy cells.
GLA could show its cancer-cell growth-inhibitory action by inhibiting cell proli-
feration or by increasing apoptotic cell death. de Kock et al. (111) demonstrated
that GLA acts differently on human osteogenic sarcoma cells (MG-63 cells) and
human epithelial cervix carcinoma cells (HeLa cells). In MG-63 cells, GLA-
induced inhibition of mitosis was associated with abnormal metaphase cell spindle
formation and inhibition of protein synthesis in G1 and S-phase. HeLa cells respond
differently, showing increased hypercondensation of chromosome, suggesting
increased apoptotic cell death that was associated with increased protein synthesis
for all of the G1 proteins and selective S-phase proteins. In a subsequent study on
HeLa cells, they further demonstrated that GLA inhibits MAP-kinase pathway and
c-Jun expression. As c-jun is the transcription factor involved in cell proliferation
and is activated by MAP-kinases, GLA is interfering with nuclear processes in in-
ducing apoptosis in HeLa cells (112). Jiang et al. (113) observed a decrease in phos-
phorylation of p27kip1 and p57kip2 that are inhibitors of cyclin-dependent kinases
IMMUNE FUNCTION AND AUTOIMMUNE DISEASES 93
and play a role in progression of mitotic growth (progression from G1 to S phase).

Decreased phosphorylation resulted in increased binding of these proteins to
cyclin-dependent kinases including CDK4, cyclin E, and CDC2. Seegars et al.
(114) studied the involvement of p53 protein in apoptotic cell death induction by
GLA and AA in skin fibroblasts and lymphoblast cells containing wild type and
mutant p53. They confirmed the earlier observations that normal cells are not
affected by GLA to any appreciable extent. They also observed that AA was
more toxic to normal cells than GLA, as GLA at much higher doses induced
apoptosis in normal cells. Transformed cells were more susceptible to apoptotic
cell death induction by GLA. The p53 does not appear to play a role in apoptosis
induction by GLA as transformed cells containing wild type and mutant p53
responded to apoptosis induction by GLA.
7. IMMUNE FUNCTION AND AUTOIMMUNE DISEASES
Immune function is a very complex function that involves interplay of several cell
types and humoral and cellular factors. Immune cells, including lymphocytes, poly-
morphonuclear leukocytes, monocytes, splenocytes, kuppfer cells, etc, have a high
content of polyunsaturated fatty acids in their membrane phospholipids. The com-
position of PUFAs in membrane phospholipids can be altered by dietary interven-
tions. GLA is taken up by inflammatory cells and is rapidly elongated to DGLA. In
some species, it can be desaturated to AA but, in human immune cells, it is not
desaturated probably because of, very limited to no delta-5-desaturase in immune
cells. By the action of enzyme phospholipase A2, free DGLA is released from the
membrane phospholipids and competes with AA for cyclooxygenases and lipoxy-
genases. DGLA produces PGE1 and thromboxane A1 (TxA1). The actions of PGE1
have been reviewed in detail by Horrobin (115). It mainly exerts anti-inflammatory
and vasodilatory properties. DGLA produces 15-hydroxyeicosatrienoic acid (15-
HETrE) by the action of 15-lipoxygenase. This metabolite of DGLA is a strong
inhibitor of 5-lipoxygenase whereby it inhibits production of leukotriene B4
(LTB4) from neutrophils (116). LTB4 has a diverse array of inflammatory actions:
It is a very potent chemotactic factor that attracts neutrophils at the site of inflam-
mation, increases adherence of leukocytes to endothelial cells, enhances migration
of T-lymphocytes in vitro stimulates release of interferon gamma and IL-2 produc-
tion by T cells, and promotes the biosynthesis of IL-1 from monocytes. Thus,
dietary administration of GLA-rich oils has a potential in modulating immune
function. Several in vitro and in vivo studies have investigated the effect of GLA
on immune functions.
Ziboh and Fletcher (117) observed a dose-dependent inhibition of calcium iono-
phore stimulated release of LTB4 by human neutrophils obtained from healthy
human volunteers fed either 0.48 or 1.5 g GLA per day for 6 weeks from borage
oil. A linear relationship between rise in polymorphonuclear neutrophil (PMN)
phospholipid DGLA and inhibition of LTB4 production was not observed. Kaku
et al. (118) observed inhibitory effects of dietary GLA on LTB4 production by
rat peritoneal exudates cells. They also reported stimulation of immunoglobin

production from mesenteric lymph node leukocytes by GLA. This action may
suggest that GLA may strengthen gut immune responses and may prevent allergic
reactions.
Santoli and Zurier (119) studied the effect DGLA, AA, and EPA on mitogen-
induced production of interleukin 2 (IL-2) by human peripheral blood mononuclear
cells (PBMCs). They observed inhibition of IL-2 production by AA or DGLA in a
dose-dependent manner. EPA showed inhibitory action in some donors only. Indo-
methacin, a cyclooxygenase inhibitor, caused an increase in IL-2 release and sup-
pressed PGE release from PBMCs. It inhibited PGE release from fatty-acid-
incubated PBMCs but did not attenuate IL-2 inhibitory action of fatty acids, sug-
gesting that the suppressive effect of AA and DGLA on IL-2 release is not mediated
through prostaglandin pathway. The inhibition of IL-2 release could be mediated by
the effect of GLA and DGLA on early response genes, as both these fatty acids
have been shown to reduce a rise in c-fos and a fall in c-myc oncogenes in T cells
(120). DeMarco et al. (121) observed a reduction in IL-2-dependent proliferation of
T-lymphocytes isolated from synovial tissue and synovial fluid from arthritic
patients. Rotondo et al. (122) studied the effect of GLA, DGLA, AA, and EPA
on IL-1-induced proliferation of thymic lymphocytes and observed that GLA was
less potent than DGLA in inhibiting the IL-1-induced proliferation of lymphocytes.
The actions of these fatty acids were not mediated through the prostaglandin path-
way, as cyclooxygenase inhibitors had no effect on the actions of these fatty acids,
which might exert a direct effect on lymphocytes. Rothman et al. (123) observed
stimulation of production of IL-1b in human peripheral mononuclear cells by
DGLA. Incubation with LPS further stimulated the production of IL-1b. Intracel-
lular IL-1b was entirely pro-IL-1b. Incubation with DGLA also stimulated release
of pro-IL-1b and small amounts of mature IL-1b. LPS failed to stimulate the further
release of IL-1b from PBMC. This could be due to the maturation of monocytes to
macrophages during 16 hours of incubation. Mature macrophages are reported to
release decreased amounts of IL-1b in response to LPS stimulation. The observa-
tions of Rothman et al. (123) are contrary to expectations, as GLA and DGLA exert
anti-inflammatory actions, one would expect a decrease in production and a release
of IL-1b. It could be caused by experimental design as they incubated the cells for
24 hours followed by incubation with LPS for 16 hours. DeLuca et al. (124) stimu-
lated the PBMCs for 30 minutes followed by stimulation with LPS for 16 hours.
They observed a dose-dependent decrease in LPS-induced release of IL-1b and
TNFa by GLA and DGLA. EPA also inhibited mediator release but required twice
the amount. They observed a similar reduction in the release of these mediators
when 2.4 g GLA was administered to human volunteers as a single dose.
A recent study by Furse et al. (125) demonstrated that LPS-stimulated IL-1b
release is further increased by IL-1, and this process is known as auto-induction.
GLA inhibits IL-1b release from LPS-stimulated monocytes mainly by inhibiting
the auto-induction process. This information may suggest that GLA may be inhibit-
ing excessive release of IL-1b to prevent inflammation but may not interfere with
basal release of IL-1b, which plays a role in host defense.
GLA and DGLA inhibit protein kinase C (PKC) activity in PMA-stimulated

T-lymphocytes. However, only GLA inhibited basal PKC activity. Both fatty acids
stimulated translocation of PKC from cytosol to membrane (126). GLA and DGLA
inhibited anti-CD3 monoclonal antibody induced early and late rise in intracellular
calcium in T cells and also inhibited a rise in inositol-1,4,5-triphosphate (IP3) pro-
duction (127). Stimulation of T cells resulted in the formation of IP3 and diacyl
glycerol (DAG). IP3 stimulates the early rise in intracellular calcium by releasing
the calcium ions from intracellular stores (SR). DAG was found to stimulate the
formation of PKC, which phosphorylates several proteins in the cells and plays a
role in late rise in intracellular calcium. GLA and DGLA promote translocation of
PKC to cellular membranes, whereby they may be inhibiting phosphatidylinositol
turnover. These studies provide a strong support to the hypothesis that GLA and
DGLA interfere with signal transduction pathways and exert antiproliferative
actions on T cells, and these actions may mediate immune modulating and anti-
inflammatory actions of these fatty acids.
Wu et al. (128) studied the effect of supplementation of black currant oil on
immune function in healthy elderly volunteers. They isolated the mononuclear
lymphocytes pre- and post-supplementation and studied the release of IL-1b,
IL-2, and PGE2, and proliferation of lymphocytes in response to mitogens, includ-
ing conclavin A and phytohemoagglutinin A (PHA). No effect of black currant oil
administration was observed on lymphocyte proliferation in response to conclavin
A, but it increased in response to PHA. There was no effect on the release of IL-1b
and IL-2, while PGE2 release was significantly decreased. Black currant oil supple-
mentation also increased delayed type hypersensitivity (DTH) response as shown
by the increase in the total diameter of induration at 24 hours and response to speci-
fic antigens (tetanus toxoid and T. mentagrophides). DTH response is depressed in
aged populations and may contribute to increased mortality and morbidity. In this
study, volunteers consumed 675-mg GLA and 653-mg ALA per day for 2 months.
Therefore, it was not possible to ascribe the results to GLA only. Nerad et al. (129)
demonstrated that administration of 2 g GLA for 12 weeks from borage oil to
healthy volunteers caused an increase in total score of indurations, suggesting
that the increase in induration observed by Wu et al. (128) may be contributed
by GLA content of black currant oil. Immune enhancing activity observed in these
studies could be contributed by a reduction in PGE2, as it is well known inhibitor of
lymphocyte proliferation and T-cell function. Zurier et al. (130) observed in vitro
suppression of T-lymphocytes proliferation by GLA, DGLA, AA, and EPA. Of all
fatty acids examined, GLA and DGLA were more potent than AA and EPA. In this
study, preincubation of lymphocytes with fatty acids was required, but the contin-
uous presence of GLA or DGLA was not needed for inhibition of proliferation sug-
gesting that these fatty acids are incorporated in the membrane phospholipids of the
cells and exert inhibitory actions on proliferation. Addition of these fatty acids
along or after the addition of a stimulant has no effect on T-cell proliferation indi-
cating that they interact at earlier stages of signal transduction, which leads to inhi-
bition of proliferation. Thus, the fatty acids tested may reduce the stimulant-
induced rise in cytosolic calcium that is required for proliferation.
7.1. Rheumatoid Arthritis

Rheumatoid arthritis is an autoimmune disease characterized by inflammation of
the joints and cartilage destruction. Several studies discussed above have demon-
strated anti-inflammatory potential of GLA-containing oils. These studies suggest
that GLA-rich oils can be used to treat inflammatory conditions.
Tate et al. (131) demonstrated in rats that a GLA-rich diet can reduce the inflam-
mation induced by injection of monosodium urate. GLA inhibited polymorpho-
nuclear leukocyte recruitment, crystal phagocytosis, and lysosomal enzyme
release. In a subsequent study, the authors demonstrated anti-inflammatory effects
of GLA in Freunds adjuvant induced arthritis in rats (132). The anti-inflammatory
effect of GLA was associated with inhibition of proliferation of pouch-lining cells
and maintenance of architecture of these cells. Hansen et al. (133) administered 4 g
of evening primrose oil supplying 360-mg GLA per day along with zinc, ascorbic
acid, niacin, and pyridoxine to a group of 20 arthritis patients for 12 weeks. They
did not observe any effect of treatment on several parameters of arthritis (number of
tender and swollen joints, pain, erythrocyte sedimentation rate, and duration of
morning stiffness). The failure of GLA to exert any beneficial effect could be
caused by low dosage short duration of treatment. Belch et al. (134) studied the
effect of 540-mg GLA or 450-mg GLA and 240-mg EPA per day on symptoms
of arthritis and NSAID requirement. They continued the treatment for 12 months
and observed that a significant number of patients had reduced requirements for
NSAIDs at the end of 12 months. After 12 months, the treatment with GLA was
stopped and three months after stopping the treatment, all the patients needed a
full dose of NSAIDs, indicating that GLA or EPA had NSAID sparing effects
and were not disease-modifying agents. In an open label clinical study, 1.1-g
GLA given for 12 weeks reduced inflammation in arthritic patients and also
reduced release of PGE2, LTB4, and LTC4 (135). Laventhal et al. (136), in a rando-
mized, placebo-controlled trial, observed that 1.4-g GLA given as borage oil for
6 months resulted in significant reduction in swollen joint count and score, tender
joint count and score, and platelet counts. They also observed a 33% reduction in
duration of morning stiffness. The only side effects of GLA treatment were belch-
ing, flatulence, and soft stools. Zurier et al. (137) repeated the trial by increasing the
dose to 2.8 g of GLA. In this trial, the patients were randomized to receive either
2.8-g GLA/day from borage oil or placebo for 6 months and, after 6 months, all the
patients were switched to GLA arm. Patients on GLA group at the end of 6 months
showed reductions in swollen joint count and score, morning stiffness, and tender
joint count and score. At the end of 12 months, the patients who started with GLA
from the beginning continued to show improvement in their symptoms. Patients
who started GLA after 6 months on placebo also started to improve. None of the
patients in the GLA group experienced deterioration of condition in first 6 months,
but, at the end of 12 months, two patients (out of 21) reported deterioration in con-
dition. Seven of these 21 patients required a reduction in the dosage of nonsteroidal
anti-inflammatory drugs or prednisone. Three months after stopping the treatment
with borage oil, most of the patients showed exacerbation of disease condition sug-
gesting that borage oil must be continued for relief of symptoms.
Overall, research with GLA-containing oils has shown that GLA provides bene-
fit in the reduction of morning stiffness by about 73 minutes and exerts a NSAID-
sparing effect. However, the dosage of GLA required for the treatment of arthritis is
not well established as, in various studies, from 340-mg to 2.8-g GLA per day has
been used.
7.2. Acute Respiratory Distress Syndrome

Acute respiratory distress syndrome (ARDS) is an acute, severe injury to the lungs.
Patients with ARDS suffer from severe shortness of breath, requiring mechanical
ventilation. It is associated with increased pulmonary capillary permeability, pul-
monary edema, increased pulmonary vascular resistance, and progressive hypoxe-
mia. ARDS can also lead to damage and failure of other organs. The exact cause of
ARDS is not known but several factors can contribute, including chest trauma,
sepsis, bacterial infections, and so on. At the cellular level, oxygen-free radicals,
cytokines, and prostaglandins can play a role. Recent research is focused on dietary
manipulations that help in reducing the inflammation and generation of pro-
inflammatory mediators. Oxidative metabolism of AA results in the formation of
pro-inflammatory mediators, including PGE2, TXB2, LTB4, etc. LTB4 is a potent
chemotactic factor and attracts neutrophils and exacerbates the damage to tissues.
GLA and its metabolic product DGLA counter the effects of AA by forming anti-
inflammatory mediators, such as PGE1 and 15-HETrE, and reduce the formation of
AA-derived inflammatory mediators. Kumar et al. (138) observed a significant
reduction in plasma phospholipid levels of GLA, DGLA, ALA, and EPA in patients
at risk of developing ARDS while patients with established ARDS additionally had
lower amounts of AA. This study suggests that treatment with GLA and EPA is
warranted in these patients. Gadek et al. (139) conducted a multicenter double-
blind, placebo-controlled clinical trial on patients with ARDS. The treatment
group (51 patients) was administered a mixture of borage oil, fish oil (providing
5.8 0.3-g GLA, 6.9 0.3-g EPA, and 2.9 0.1-g DHA per day), and antioxidants
via gastric or jejunal tube. They observed a significant reduction in the number of
total cells and neutrophils in brachioalveolar fluid by day 4 compared with the fluid
obtained from the control group. This was associated with improved arterial oxy-
genation. Patients in the treatment group had a lesser requirement for ventilator
support, supplemental oxygen, and lower number of days of stay in ICU compared
with patients in the control group. Significantly fewer patients in the treatment
group developed new organ failure, and there was about a 17% reduction in the total
number of infections in the treatment group. As this study used a combination of
EPA and GLA with antioxidants, it is difficult to differentiate the effects of GLA
alone, although this study provides a strong support for using a combination of
EPA and GLA. Murray et al. (140) studied the effect of fish oil alone or in combi-
nation with borage oil on cardiac function in pigs during acute lung injury induced
by infusion of E.coli endotoxin. They observed that fish oil or a fish oil and borage
oil combination attenuated lung injury induced depression of cardiac function. A
combination of fish oil and borage oil acted synergistically compared with fish
Saline control
Fish oil + Borage oil (EPA + GLA)
8 Fish oil (EPA)
Corn oil (LA)
7
Cardiac Index (L/min/M2)
6
+
3
0 1 2 3 4 5
Pulmonary Vascular Resistance (dynes.sec/cm5)
1000
800
600
*
400
+ +#
*
200
#
0
0 1 2 3 4 5
Time (hours)
Figure 2. Effect of various treatments on cardiac index (upper panel) and pulmonary vascular
resistance (lower panel). Data adapted from (140).
oil alone in attenuating the cardiac depression (Figure 2, Upper panel). A fish oil
and borage oil combination had lower pulmonary vascular resistance during the
4-hour experiment duration than either the control group or the group given fish
oil alone (Figure 2, Lower panel). The other interesting finding of this study was
that GLA in combination with EPA prevented the loss of platelets from circulation,
whereas EPA alone did not exert this effect. This observation indicates that GLA
decreased the aggregatory and adhesive properties of platelets in vivo. A significant
SKIN CONDITIONS 99
reduction in the amount of TXB2 (EPA, and EPA GLA groups) and 6-keto pros-
taglandin F1a in the alveolar fluid in EPA GLA group was also observed,
suggesting that the beneficial effects of the treatment may be mediated by a reduc-
tion in the formation of pro-inflammatory and vasoconstrictor metabolites of AA. In
a subsequent study, the administration of EPA or EPAGLA to pigs altered the
composition of pulmonary surfactant by reducing the concentration of oleic acid
and increasing the concentration of DGLA, EPA, and DHA. However, there was
no effect on the pulmonary compliance or surfactant function. Mancuso et al.
(141) observed that a combination of fish oil and borage oil attenuated endotoxin
induced rise in pulmonary microvascular protein permeability in rats, which was
associated with a decrease in LTB4, TxA2, and PGE2 production by pulmonary
alveolar macrophages. Additionally, this treatment also attenuated endotoxin
induced early and late hypotension.
8. SKIN CONDITIONS
Skin is the largest organ of the body and provides a barrier function, whereby it
protects the inner organs from environmental toxins and bacteria. It also plays an
important role in temperature regulation, sensory perception, and excretion. It
undergoes constant renewal. Burr and Burr (142) observed that rats on a fat-free
diet developed dried, scaly skin suggesting the role of fats in normal physiology
of skin. Subsequent studies have demonstrated that skin contains essential fatty
acids and is metabolically an active organ. It has the capability to elongate the
fatty acids but lacks the capacity to desaturate. This information suggests that der-
mal cells take up preformed long-chain metabolites of LA (GLA, DGLA, and AA)
and ALA (EPA, DPA, and DHA).
During EFA deficiency, the levels of LA, DGLA, and AA are reduced in the skin
and may contribute to dry and scaly appearance of the skin, with the increase in
epidermal water loss. In studies with EFA-deficient rats, mice, and guinea pigs, it
has been demonstrated that skin undergoes hyperproliferation (acanthosis, hyper-
granulosis, and hyperkeratosis) with increased DNA synthesis. LA levels were sig-
nificantly decreased with an increase in mead acid (20:3 n-9, abnormal fatty acid
characteristic of EFA deficiency). Supplementing diets with a large dose of saf-
flower oil (rich in LA) or much smaller dose of evening primrose oil (rich in LA
and GLA) reversed the signs of EFA deficiency on skin, whereas fish oil failed to
reverse these symptoms (143). In this study, a rise in EPA, DPA, and DHA levels in
skin phospholipids was observed, but the levels of LA did not increase. This study
also confirmed, by labeled fatty acid incubation, that skin lacks delta-6 and delta-5
desaturase activities, indicating that skin cannot metabolize LA or ALA.
8.1. Atopic Dermatitis

Atopic dermatitis, or eczema, is a skin disorder characterized by dry, itchy, and
hypersensitive skin. It is common in children but can occur at any time and age.
The exact cause of eczema is not well understood and it can be hereditary. Soon
after the discovery of essential fatty acids, it was observed that infant patients suf-
fering from atopic eczema had low levels of LA and AA (144) and responded well
to supplemental lard containing LA and AA (144). Very high doses of LA (2050 g)
provided partial relief from symptoms of eczema but failed to raise the levels of
metabolites of LA in the blood. Later on, it was confirmed that the plasma phos-
pholipids of adults suffering from atopic dermatitis had higher concentrations of
LA and a lower concentration of GLA, DGLA, and AA (145), suggesting that ato-
pic patients may suffer from defective delta-6-desaturation. These patients also
failed to show flushing response to topically applied niacin, suggesting that they
have defects in prostaglandin pathways and fail to produce vasodilatory prostaglan-
dins. Subsequent studies have shown lower levels of DGLA in breast milk of atopic
mothers than in the normal mothers (146, 147). As the breast-fed infants get their
nutrient requirements from breast milk, they do not receive sufficient quantities of
DGLA and may be prone to dermatitis.
Based on these observations, it seems logical that dietary GLA or DGLA should
help prevent/treat atopic dermatitis. Wright and Burton (148) conducted a double-
blind, placebo-controlled clinical trial of evening primrose oil. They recruited
60 adults and 39 children suffering from moderate to severe atopic dermatitis in
the study. Adult patient groups received 4, 8, or 12 capsules daily, whereas children
were given 2 or 4 capsules daily providing 45 mg GLA per capsule, and the placebo
was liquid paraffin. Treatment was continued for 12 weeks. All of the patients had
moderate to severe eczema. They observed that the lower dose of GLA only pro-
vided relief from itch, while the other two groups of adult patients on higher doses
of GLA showed better improvements in itch, scaling, and the general impression of
severity as assessed by the physician and the patient. Children in this study did not
perform as well as the adults, possibly caused by either insufficient dose of GLA or
high placebo effects in children. Manku et al. (149) analyzed the blood samples of
adult patients from the above study for plasma phospholipid fatty acids. They
observed that LA levels were higher in the atopic patients, and the scatter of values
for LA was also very high. Levels of DGLA and AA were lower in these patients.
Treatment with 4 capsules per day did not affect blood GLA or DGLA or plasma
PGE1 levels, whereas 8 and 12 capsules per day caused a significant elevation in the
levels of DGLA and PGE1. Schafer and Kragballe (150) observed that neutrophils
and epidermis of atopic dermatitis patients have high levels of monounsaturated
fatty acids (MUFAs), which correlated positively with the severity of disease,
and lower ratios of n-6 PUFAs/MUFA. Feeding 6 g of evening primrose oil for
10 weeks increased the ratio of n-6 PUFA/MUFA and increased the levels of
DGLA in neutrophils and epidermal phospholipids. They did not evaluate the effect
of the treatment on dermatitis as the patients were allowed to use emollients; this
study cannot shed any light on the efficacy of GLA in dermatitis. In another large
multicenter study, 179 patients with eczema were treated with 4.0 g of evening
primrose oil per day, and they demonstrated clinical improvements as evaluated
by a dermatologist. Scarff and Lloyd (151) studied the effect of treatment with
evening primrose oil in dogs suffering from dermatitis. In this study, the dogs
SKIN CONDITIONS 101
were on olive oil placebo for 3 weeks followed by either olive oil or evening prim-
rose oil for 9 weeks. At the end of 9 weeks, the treatments were switched over with-
out any wash out period. They observed a deterioration in condition during first
3 weeks on olive oil in all of the dogs; however, during the first treatment period,
all dogs showed improvement that could be ascribed to placebo effect in the olive
oil group. In the second treatment period, dogs on olive oil worsened, whereas those
on evening primrose oil improved. They observed an interaction in the order of
treatment with the evening primrose oil that could be caused by a change in treat-
ment between active and placebo without any washout period. Fiocchi et al. (152)
evaluated safety and efficacy in children suffering from atopic eczema. They treat-
ed these children (average age 11.4 months) with 3.0 g/day of GLA for 28 days.
None of the children showed complete recovery, although gradual improvement
in erythema, excoriations, and lichenification was reported. They also reported a
significant reduction in itching and the use of antihistamines without observing
any side effect because of the treatment. Borrek et al. (153) compared the effect
of borage oil with corn oil in 24 subjects suffering from atopic eczema in
a double-blind cross-over trial. The subjects were between 3 and 17 years old
and received 360-mg GLA daily for 1014 weeks. They did not observe any differ-
ence between the two groups, and the placebo treatment also showed improve-
ments. In this study, 10 patients on borage oil treatment showed improvements
but they did not differ from nonresponders in any of the characteristics (age, sex,
symptom severity, etc.). As there was a large placebo effect, the effectiveness of
borage oil may have been masked due to a small number of subjects in this study.
Eczematous skin also has high transepidermal water loss compared with normal
skin. Hartop and Protty (154) observed that application of pure GLA triacylglycerol
to rats with dry skin due to essential fatty acid deficiency reduced the transepider-
mal water loss. In essential fatty acid deficiency, there was a loss of LA, ALA, and
their long-chain metabolites in plasma and other organs. However, the eczema dif-
fered from essential fatty acid deficiency in which there is usually an excess of LA
with imbalance of longer chain metabolites. Tollesson and Frithz (155) studied the
effects of topically applied borage oil on the transepidermal water loss from the
skin of infants suffering from seborrhoeic dermatitis. They observed that topically
applied borage oil relieved the symptoms of dermatitis within 34 weeks and also
normalized the elevated transepidermal water loss. Topically applied borage oil also
caused a rise in serum LA content, suggesting transdermal absorption of LA from
borage oil. The site of application of borage oil was not important as borage oil in
the napkin area of the infants also relieved the symptoms at other sites.
Henz et al. (156) evaluated the efficacy of borage oil in the treatment of atopic
dermatitis in a double-blind, placebo-controlled, multicenter clinical trial. In this
study, 160 patients with moderate eczema (Costa score between 20 and 36 points)
were divided into two groups. The active group received 3.0 g borage oil (690 mg
GLA) daily for 24 weeks and the placebo group received migliol, an oil containing
no GLA. Patients were allowed to use a steroid cream during the trial. Some
patients did not follow the guidelines and violated the conditions of protocol and
included poor compliance (less than 70% of dose consumed; 7 patients on placebo,
6 on borage oil), excessive use of steroid cream (three times above median dose;
1 patient on placebo and 4 on borage oil), and less than 11 weeks of treatment
(6 patients on each treatment), and patients with unstable disease (Costa score of
less than 18 at week 2; 32 patients on placebo and 21 on borage). When all the
patients, including those who did not follow the protocol, were included in the
data analysis, no significant differences in Costa scores between the two groups
was observed, although borage oil treatment improved erythema, vesiculation,
crusting, excoriation, lichenification, and insomnia scores over placebo group. A
marked reduction in serum IgE levels was observed, but the difference was statis-
tically insignificant due to large intersubject variations. Borage oil treatment also
increased plasma and erythrocyte levels of GLA and DGLA in the majority of
patients. When the subgroup of patients who did not follow the protocol was
excluded from the analysis, the borage oil treatment showed significant improve-
ment on the reduction of steroid cream use. Borage oil was well tolerated with
minor side effects (headache, nausea, vomiting, and diarrhea). The frequency of
side effects was not different from that observed with the placebo treatment.
Takwale et al. (157) conducted a single-center, double-blind clinical trial to study
the efficacy and tolerability of borage oil in the treatment of atopic eczema in chil-
dren and adults. Adult patients were given 8 capsules of borage oil (supplying 920-
mg GLA) and children were given 4 capsules (supplying 460 mg of GLA) daily for
12 weeks. Patients were allowed to continue using a steroid ointment for sympto-
matic relief. The efficacy of the treatment was evaluated from the change in total
symptom score measured with the six-area, six-sign, atopic dermatitis (SASSAD)
score as the primary endpoint. Secondary end points included symptom score
assessment on visual analogue scales, topical corticosteroid requirement, and global
assessment of response by participants. This study failed to observe any effect of
borage oil treatment on eczema, although the treatment was safe, well tolerated, and
was free from major side effects. This study suffers from several major limitations.
They recruited 151 patients, of which 11 were lost at week two of the 12-week
study. An additional 16 participants withdrew during the trial, leaving only 124 sub-
jects who completed the trial. However, they analyzed the data for 140 patients,
including those who did not complete the protocol. Good clinical trial demands
inclusion of data only from those patients that follow the protocol. Noncompliance
with the treatment protocol is the single most important reason for failure of treat-
ment in dermatological practice and was evident in the study by Henz et al. (158).
They used two different placebo treatments: Liquid paraffin for adults and olive oil
for children. Liquid paraffin is an inert material for its effect on atopic dermatitis,
whereas olive oil is not as inert because it can modify the cellular fatty acid profile.
It has been reported to increase tissue levels of DGLA (159, 160). By increasing
tissue levels of DGLA, olive oil may increase the dermal levels of lipoxygenase
and cyclooxygenase metabolites of DGLA that are reported to exert anti-inflamma-
tory actions (161, 162). Therefore, olive oil may show some beneficial effects
because of the above-mentioned biochemical pathways, and hence, may not be a
true placebo and dampen the effect of treatment. Therefore, separate analysis in
adults and children was highly desirable to avoid the potential variations in
DIABETES 103
outcome induced by different placebo. The scoring system (SASSAD) used in this
study as a primary outcome parameter is reported to have a very high interobserver
variation (730, median 15.5, out of a possible score of 108) (163).
8.2. Other Skin Conditions

Radiation-induced damage: Skin is sensitive to radiations. Hopewell et al. (164)
studied the effect of GLA or GLA and EPA combinations on radiation-induced
skin damage in pigs. They treated female pigs for 4 weeks prior to and 1016 weeks
after irradiation of skin. Control pigs were treated with a placebo oil devoid of GLA
and EPA. The pigs were irradiated with a single or fractionated (20 F/28 days) dose
of b-rays. They observed that prior administration of GLA or GLA EPA had no
protective effect, while given before and after irradiation, both of these interventions
reduced the development of early (bright red erythema or moist desquamation) and
late (dusky/mauve erythema and dermal necrosis) reactions of radiations. This
observation suggests that GLA alone or in combination with EPA may help im-
prove the efficacy of radiation treatment by reducing the side effects.
9. DIABETES
Diabetes is a metabolic syndrome resulting in disturbed glucose homeostasis. This

can be mediated because of decreased production of insulin (because of damage to
insulin-producing pancreatic b-cells), as in type 1 diabetes, or increase in tissue
resistance to the action of insulin (type 2 diabetes). According to the American Dia-
betes Association, approximately 17 million people in the United States, or 6.2% of
the population, have diabetes. Although an estimated 11.1 million have been
diagnosed, unfortunately, 5.9 million people (or one-third) are unaware that they
have the disease. It affects people of all ages and races, although it is more common
in African Americans, Latinos, Native Americans, Asian Americans, and Pacific
Islanders. Incidence of diabetes increases with age in all populations. Diabetics
are also at high risk of other complications, including cardiovascular (atherosclero-
sis, heart attack, stroke, peripheral vascular disease), neurological (neuropathies),
renal failure, skin diseases (dry, itchy skin, skin infections, dermopathy), slow
wound healing, retinopathy, and impotence. A combination of neuropathy and vas-
cular disease in diabetics leads to more amputations. The mechanism of complica-
tions of diabetes is not well understood. High blood-glucose levels may be
contributing to various complications, although several pathways, including increa-
sed oxidative stress, modification of proteins by glucosylation, reduction in produc-
tion of vasodilator mediators including nitric oxide, prostacyclin, altered cytokine
production, and so on, may be involved. Reduced tissue perfusion and resulting
cellular damage contribute to several pathophysiological situations. The manage-
ment of diabetes involves controlling the blood-glucose levels by diet, exercise,
or drugs.
Several studies have confirmed that diabetes inhibits the activity of delta-6-desa-
turase (165167), which is the first enzyme in the metabolism of LA and ALA. As a
result of inhibition of this enzyme, diabetics have a lower content of DGLA and AA
in various tissues (168). This may lead to imbalance in different eicosanoids in dia-
betics and may contribute to various complications common in diabetics. Based on
these observations, it was hypothesized that GLA may help correct many of the
complications of diabetes. To test this hypothesis, many studies were conducted
in animals and humans. Some of these studies are discussed here.
Diabetic patients and animals show reduced nerve conduction velocity. Julu
(169) from the University College of London performed a detailed study investigat-
ing the effect of GLA alone or in combination with EPA on nerve conduction velo-
city. During diabetes, nerve conduction velocity was reduced. Rats made diabetic
by injection of streptozotocin (a pancreatic toxin) suffered about a 20% decline in
motor nerve-fiber conduction velocity. Supplementation with GLA alone signi-
ficantly attenuated the diabetes-induced deficit in nerve conduction velocity. Com-
bined treatment with GLA and EPA completely prevented diabetes-induced
reduction in motor nerve conduction velocity. In this study, they observed no effect
of GLA or EPA on diabetes-induced weight loss, increase in blood-glucose, and
glycosylated hemoglobin levels. In the subsequent study, Julu and Mutamba
(170) studied the comparative effect of GLA and insulin for 3 or 5 days of treatment
after induction of diabetes with streptozotocin in rats. They again observed a reduc-
tion in conduction velocity in myelinated sensory and motor nerve fibers. Unmye-
linated sensory fiber also showed a trend for reduction in conduction velocity, but
because of high intersubject variability, the fall could not reach statistical signi-
ficance. Treatment with insulin for 3 days partially corrected the deficit in sen-
sory-nerve conduction velocity, whereas motor-nerve conduction velocity was
brought back to normal levels. Treatment for 5 days with insulin returned the
motor-nerve conduction velocity to normal levels. GLA treatment for 3 days
overcorrected the sensory-nerve conduction velocity, whereas the motor-nerve con-
duction velocity was brought back to normal levels. Treatment with GLA for 5 days
brought sensory-nerve conduction velocity to normal levels. The overcorrection in
sensory-nerve conduction velocity by three-day treatment with GLA cannot be
explained. GLA treatment had no effect on blood-glucose levels or diabetes-
induced weight loss, whereas insulin corrected both of these parameters. Cameron
et al. (171) studied the effect of GLA alone or in combination with fish oil on dia-
betes-induced reduction in nerve conduction velocity and resistance to conduction
block and found that diabetes increased the resistance of nerves to hypoxic conduc-
tion block and reduced the nerve conduction velocity. Treatment with GLA pre-
vented these changes, whereas a combination with fish oil was less effective.
Their results on conduction velocity are different from those reported by Julu
and Mutamba (170) who observed better efficacy of GLA and fish oil combination.
This difference is difficult to explain except for the sex differences in the rats in the
two studies. Julu and Mutamba (170) used female rats, whereas Camerons (171)
group used male rats. The role of female sex hormones in differential observation of
two groups cannot be discounted as the polyol pathway is differentially affected in
DIABETES 105
males and females. This pathway may be mediating a greater role in conduction
velocity reduction in males than in females because of reduction of perfusion of
vasa nervosum. In addition, Cameron et al. (171) observed that GLA treatment
reduced the resistance to ischemic conduction block that was increased by diabetes.
This may be caused by improved perfusion of vasa nervosum. During diabetes, the
perfusion to vasa nervosum is reduced, which may cause ischemic preconditioning
of nerves leading to increased resistance to subsequent ischemic conduction block.
They also demonstrated increased capillary density in the sciatic nerve of GLA-
treated diabetic rats leading to improved perfusion. The fatigue index of skeletal
muscle was increased in diabetes, which increase was attenuated significantly by
GLA treatment. This was also associated with increased capillary vascularization
of muscle. In the subsequent study, Cameron and Cotter (172) observed that strep-
tozotocin-induced diabetes reduced endoneural blood flow and oxygen tension that
may be causing the hypoxic injury to the nerve cells leading to reduced nerve con-
duction and neuropathy. Treatment with evening primrose oil prevented the
decrease in endoneural blood flow that was primarily caused by increased perfusion
through capillaries leading to normal endoneural oxygen tension. However, in non-
diabetic rats, evening primrose oil treatment increased the bulk flow (flow through
major arteries, arteriols, veins, and arterio-venous shunts) without affecting capil-
lary blood flow. The chronic increased blood flow through major vessels may be a
stimulus to increased angiogenesis observed in an earlier study. Treatment with
vasodilator drugs similarly caused an increase in blood flow and stimulated angio-
genesis. After confirming the beneficial effects of GLA treatment in amelioration of
diabetic nerve-conduction deficit, Cameron and Cotter (173) studied the effect of
GLA and antioxidant treatment on nerve conduction velocity in diabetes. They
selected a dose of GLA that would only correct the nerve conduction velocity
reduction by about 20%. When the diabetic rats were given 20 mg/kg/day of
GLA alone or in combination with an antioxidant (BM15.0639), the combination
had a synergistic effect in improving the velocity. This synergistic effect on
nerve conduction velocity was mediated by synergistic improvement in sciatic nerve
capillary endoneural blood flow by the combined treatment. Several later studies
confirmed the beneficial effects of antioxidants and GLA given in combination or
as conjugates. In these studies, investigators studied the effects of GLA conjugates
with ascorbic acid or a-lipoic acid or in combination with various antioxidants,
including Vitamin E, ascorbic acid, a-lipoic acid, n-acetylcystein, butylated hydr-
oxyltoluene, and so on. These observations confirm previously reported observa-
tions that diabetic complications may be mediated by a combination of increased
oxidative stress and abnormalities in essential fatty acid metabolism.
Jamal and Carmichael (174) conducted a double-blind, placebo-controlled trial
of evening primrose oil in patients with type 1 and 2 diabetes with established neu-
ropathy for at least 6 months. The patients were given 380 mg of GLA per day for
6 months and were evaluated before and after the treatment for neurological symp-
toms (pain, parasthesia, numbness, weakness, and abnormal sensation to heat and
cold) as well as nerve conduction. At the end of the 6-month treatment, a significant
improvement in 9 out of 12 variables was observed on GLA treatment. Treatment
with GLA also increased plasma phospholipid content of GLA, DGLA, and AA
that are reduced in diabetes. The GLA treatment had no effect on glycosylated
hemoglobin levels. These observations indicate that GLA-induced improvements
may be mediated by improved perfusion of nerves rather than correction in meta-
bolic derangements. By improving the tissue perfusion, treatment may have pre-
vented hypoxic insult and related injury to the nerves, whereby improving the
nerve conduction and reducing the associated pain and numbness. Keen et al.
(175) conducted a double-blind, placebo-controlled, multicenter clinical trial of
GLA in 111 diabetic patients with mild neuropathy. They studied the effect of treat-
ment over 1 year on 13 symptoms, including motor-nerve conduction velocity, mus-
cle strength, hot and cold thresholds, sensation, and tendon reflexes. GLA treatment
at a dose of 480 mg per day was demonstrated to render benefits on functions,
and the effects were more pronounced in well-controlled diabetics than in poorly
controlled subjects. Keen et al. (175) used a higher dose of GLA than Jamal
(174), and the study was of longer duration.
Diabetes is also known to impair immune response in humans (176) and animals,
which may contribute to slow wound healing in diabetics. In streptozocin-induced
diabetes in rats, a number of circulating T and B lymphocytes decreased with no
effect on the number of circulating monocytes and neutrophils. In a study by
Oon et al. (177), it was demonstrated that feeding evening primrose oil prevents
a diabetes-induced fall in lymphocyte number that may have been mediated by
increased production of PGE1 in this group.
10. PREMENSTRUAL SYNDROME (PMS)
Premenstrual syndrome (PMS) is a recurrent cyclic disorder associated with the

cyclic hormonal rhythms of the menstrual cycle. A large number of symptoms
have been associated with PMS that are divided into physical, behavioral, and emo-
tional symptoms. PMS may be associated with dysmenorrhea and other menstrual
irregularities. Physical symptoms include bloating, abdominal and back cramps and
discomfort, change in appetite, weight gain, breast tenderness and pain, and head-
ache. Behavioral changes include anxiety, depression, lethargy, hypersomnia or
insomnia, moodiness, irritability, anger, and social withdrawal. These symptoms
vary in intensity from mild to severe and affect up to 90% of women some time
in their child-bearing age. About 40% of women in industrialized countries suffer
from mild to moderate symptoms of PMS, whereas about 10% of North American
women suffer from moderate to severe symptoms affecting their daily life activities
(178).
The objective of treatment is to prevent the symptoms of PMS. Several therapeu-
tic agents are available to control the symptoms of PMS, but none of the agents are
effective in controlling more than a few symptoms. In a study on 42 women suffer-
ing from PMS, a deficiency of long-chain metabolites of LA was reported with
above-normal LA levels in plasma phospholipids, suggesting inhibition of delta-
6-desaturase enzyme (179). As prostaglandins play an important role in regulation
INFANT NUTRITION AND DEVELOPMENT 107
of reproductive function, it was suggested that an imbalance between different pros-

taglandins may contribute to the symptoms of PMS. GLA, being the only natural
source of DGLA that can be supplemented, may benefit these women by improving
the ratio of prostaglandins of series 1 and 2. Puolakka et al. (180) performed a
placebo-controlled trial of evening primrose oil in 30 women suffering from severe
PMS. They observed that evening primrose oil treatment reduced the symptoms of
PMS, although it was more effective in reducing depression. Treatment with even-
ing primrose oil also reduced the production of thromboxane B2 by the platelets
during clotting. The effectiveness of GLA oils in treatment of PMS is not equivo-
cally proven. Some studies reported no effect of evening primrose oil treatment
over placebo (181).
11. INFANT NUTRITION AND DEVELOPMENT
The role of GLA in infant nutrition and development as such is not very clear. The
body weight of infants at birth was positively associated with the proportions of AA
and DGLA in plasma triacylglycerols and choline phosphoacylglycerols in prema-
ture infants (182) and infants born at full term (182, 183). The positive association
of DGLA with birth weight is more consistent than that observed for AA or DHA.
This information suggests that DGLA is playing an important role in fetal develop-
ment. The exact role, however, is not clear. Careful analysis of breast milk compo-
sition from women of different geographical areas revealed that women who had a
higher amount of DHA in their breast milk lipids also had a higher amounts of
DGLA (Figure 3) (184). The role of DHA is well established in infant development,
1.2 Japanese
Chinese
1
Canadian
0.8
0.6
0.4
0.2
0
ALA DPA DHA DGLA AA
Figure 3. Breast milk fatty acid composition from Japanese, Chinese, and Canadian women.
while the role of GLA and DGLA is not clear. This observation suggests that
DGLA may play an important role in infant development. Earlier discussion in
this chapter revealed that GLA and DGLA may be playing a role in atopy in infants
as the breast milk of mothers of atopic children had lower GLA and DGLA levels in
breast milk. Recently popularized, the Barker hypothesis (185) (the fetal origins
hypothesis) states that, during fetal development, the environmental exposures set
the limit to metabolic capacity. When this capacity is exceeded later in life, it sets as
an overt disease. According to this theory, the effect of nutrient deficiency may be
more marked than that during later life and may cause predisposition to many
chronic diseases, such as heart disease, diabetes, metabolic syndrome, etc. (186,
187). In support of this hypothesis, it was observed that children who had higher
levels of GLA in umbilical cord plasma phospholipids at birth did not develop insu-
lin resistance at the age of 7 years, while those with lower levels of DGLA had
increased insulin resistance, body fat, insulin, proinsulin, and leptin concentrations
(188). The exact mechanism of how DGLA in infancy and fetal development can
affect the health conditions in later life is not clear. It may be possible that GLA, as
a ligand for peroxisome proliferator activated receptors (PPARs), affects the tran-
scriptional regulation of glucose and lipid homeostasis.
12. DRUG/NUTRIENT INTERACTIONS
GLA modulates the second messengers at the cellular levels. Many drugs act by
interfering with these second messengers. This suggests that GLA may interact
with drugs to affect their actions. These actions could include alterations in the ther-
apeutic potential or side effect profiles. Cyclosporin is a potent immunosuppressant
drug that is commonly used in the prevention of graft rejection in transplant reci-
pients and in the treatment of severe psoriasis. The major side effects of cyclosporin
include hypertension and renal toxicity (189, 190). In borderline hypertensive rats,
GLA was shown to inhibit hypertensive and glomerular filtration-rate-reducing
actions of cyclosporin (191). In Wistar rats, GLA was shown to attenuate nephro-
toxic effects of cyclosporin (192). These actions could be mediated by increased
production of PGE1, as demonstrated by the increased ratio of 6-keto-PGF1a/
TXB2. GLA has been shown to enhance the sensitivity of 36B10 astrocytoma cells
to radiation (193). The enhanced sensitivity could be mediated by increased free
radical production as it was blocked by Vitamin E. GLA has been shown in vitro
to enhance the cytotoxicity of paclitaxel to various breast cancer cell lines, includ-
ing MDA-MB-231, MCF-7, SK Br3, and T47D (194). In this study, the authors
observed the synergistic action of GLA when the cells were co-incubated with
paclitaxel, whereas preincubation of cancer cell lines with GLA, followed by treat-
ment with paclitaxel, resulted in only additive effects. These actions of GLA were
only partly inhibited by Vitamin E, suggesting that increased oxidative stress may
partly be contributing to the cytotoxic actions of GLA against breast cancer cell
lines. Thus, GLA was found to be most potent in enhancing cytotoxic actions of
paclitaxel followed by ALA, EPA, DHA, and OA, whereas LA had no effect.
SAFETY OF GLA-CONTAINING OILS 109
Similar results were obtained for vinorelbine and GLA in breast cancer cells
(MDA-MB-231, T47D, and SK-Br3) (195).
Recently, it was shown that GLA acts synergistically with tamoxifen in enhanc-
ing the antitumor activity (196). This study was conducted in nude mice implanted
with estrogen receptor positive breast cancer cells (MCF-7 B1M). GLA acted
synergistically with tamoxifen in inhibiting tumor growth and expression of estro-
gen receptors.
Ikushima et al. (197) studied the interaction of GLA on cytotoxicity of various
anticancer drugs in human nueroblastoma cell lines in culture. They observed that
GLA enhanced absorption and cytotoxicity of vinca alkaloids (vincristine, vinblas-
tin, and vindesine) 22.5-fold. This was associated with increased lipid peroxida-
tion of cancer cells. In the same cell lines, GLA inhibited the cytotoxic action of
platinating agents like cisplatin and carboplatin. This study suggests that GLA may
react differently with various anticancer drugs. Liu and Tan (198) observed that
GLA and DHA increase the absorption of doxorubicin into doxorubicin-sensitive
and-resistant lymphoma cancer cells. The resistant cells became sensitive to doxo-
rubicin toxicity, which was associated with increased superoxide dismutase activity
with no effect on catalase activity and p-glycoprotein levels. This observation sug-
gests that GLA has no effect on p-glycoprotein, which plays a role in multidrug
resistance development. However, by increasing the levels of only superoxide dis-
mutase and not catalase activity, GLA may stimulate the formation of hydrogen
peroxide in the cancer cells, which may contribute to oxidative toxicity of doxo-
rubicin. Hydrogen peroxide can form hydroxyl radicals, which are highly toxic
to adjacent molecules.
Kaku et al. (199) studied the interaction of GLA with soy protein and casein in
mediating immune response and LTB4 production by rat peritoneal exudates cells.
They observed that dietary borage oil reduced production of LTB4 from the perito-
neal exudates cells and the effect was stimulated by soy protein but not by casein.
This study suggests that soy protein, but not casein, may stimulate anti-inflamma-
tory action of GLA-rich oils.
13. SAFETY OF GLA-CONTAINING OILS
GLA-containing oils have been studied in several clinical trials on humans, in addi-
tion to laboratory animal studies. All of these studies have revealed that these oils
are safe and are devoid of serious side effects. The more commonly observed side
effects with these oils include gastric upsets (nausea, vomiting, diarrhea, belching),
and headache. Evening primrose oil was associated with reducing sensitivity to sei-
zure threshold in patients suffering from temporal lobe epilepsy (200). Although
similar effects with borage or black currant oils have not been reported, it is advi-
sable to observe caution when giving these oils to epileptic patients. Potential of
hepatotoxicity from pyrrolizidine alkaloids in borage is nonexistent (as discussed
in section 2.1.1) as the content of pyrrolizidine alkaloids in borage oil is less
than 4 ppb. At this level, one may have to consume more than 250 g of oil per day to
expose to toxic levels of alkaloids.
14. CURRENT RESEARCH FOCUS
Current focus of research is on increasing the concentration of GLA in oils and to

find new sources of GLA for commercial use. The strategies include genetic mani-
pulations, variety development, and concentrations of existing GLA-rich oils like
borage and evening primrose. GLA-containing oils can be concentrated to higher
GLA levels by employing common techniques such as hydrolysis of oil to form
free fatty acids followed by urea complexation to remove saturated and monounsa-
turated fatty acids (201). Employing this technique, the oil can be concentrated to
4080% GLA (202, 203). The resultant oil contains GLA as a free fatty acid or can
be converted to ethyl ester or triacylglycerol form by chemical/enzymatic esteri-
fication. The triacylglycerol form produced in this way contains about 5070% tri-
acylglycerols, 1025% diacylglycerols, and 510% monoacylglycerols (Bioriginal
Food and Science Corp). The enzymatic process involves the use of microbial
lipases (from Pseudomonas sp.). Other areas of research include increasing the con-
tent of GLA in foods and alternative crops by genetic engineering. Cook et al. (204)
inserted the delta-6-desaturase gene from borage into tomato. This strategy resulted
in an increase in the content of GLA in tomato fruit along with a reduction in LA
content. Although this variety has not been commercialized, there is a potential in
optimizing the variety. Similar efforts have been made on other plants, including
tobacco (205, 206) and canola (207). GLA levels in tobacco plants could be in-
creased to about 14% of total fatty acids. At this level, it is not an economical
source for production of GLA. Recently, Ross Labs has been successful in produc-
ing a transgenic variety of canola plant that contains up to 40% GLA in its seed oil
(207). The transgenic canola oil containing GLA was compared with borage oil for
its pharmacological actions and was found to be similar to borage oil, demonstrat-
ing that the transgenic canola oil could become an economically viable source of
GLA.
Another area of current research is development of structured lipids where GLA
is combined with a fatty acid of omega-3 family, preferably EPA or DHA, into
one triacylglycerol molecule. Structured lipids can be produced by interesterifying
a mixture of conventional fats and oils of interest using chemical or enzymatic
methods. Chemical methods provide random distribution of different fatty acids
on the glycerol backbone, whereas enzymatic reactions could be position specific,
affording controlled production of triacylglycerols with desired configuration (208).
Interesterification using the chemical method usually involves a reaction between
two oils using metal alkoxide (sodium methoxide) as a catalyst. The unreacted fatty
acids are removed by vacuum distillation. The alternative and more researched pro-
cess involves acidolysis using lipases. In this process, either pure fatty acid is reac-
ted with a triacylglycerol molecule or relatively rich fraction of fatty acid of interest
is taken/prepared before acidolysis reaction.
REFERENCES 111
The structured lipids have unique chemical, physical, or physiologic properties

that are not observed by simply blending mixtures of the starting fats and oils (209).
At present, the enzymatic process is under development but has not been widely
commercialized so far due to the economy of the process. Laboratory research is
in progress with the objective to develop a process that can be economically scaled
up. The current emphasis is on optimization of lipases, reaction conditions includ-
ing water activity of the reaction mixture, mole ratios of fatty acids to triacylgly-
cerol, amount of enzyme, reaction temperature, and duration. The nonspecific
lipases can be obtained from Candida rugosa, Pseudomonas sp, while 1, 3-position
specific lipase are obtained from Aspergillus niger, Mucor javanicus, Rhizomucor
miehei, Rhizopus sp., Geotrichum candidum, Candida cylindracea, Candida para-
lipolytica, Rhizopus delemar, etc. One can utilize either pure EPA or DHA as free
acid or ethyl esters for incorporation into borage or evening primrose oil or GLA
can be added to fish oils containing EPA and DHA. In these approaches, a struc-
tured lipid containing EPA, DHA, and GLA in one triacylglycerol molecule is pro-
duced. Spurvey et al. (210) studied the effect of reaction conditions on the
incorporation of GLA into menhaden and seal blubber oils. They observed that
the best conditions include a mole ratio of 3:1 for GLA to triacylglycerol, enzyme
concentration of 500 units/g of oil, reaction temperature of 40 C, and time of
24 hours for incorporation of GLA into fish oils. They utilized concentrated
GLA from borage oil for acidolytic reaction. The GLA of 91% concentration
was prepared by chemical hydrolysis followed by urea complexation (201). This
concentrate was reacted with fish oils using lipase from Pseudomonas sp (non-
specific enzyme) and Mucor miehei (sn 1,3-specific enzyme). Lipase form
Pseudomonas sp. gave higher incorporation of GLA into fish oil. Using a similar
approach, production of EPA-rich borage oil and evening primrose oil has been
produced (208).
REFERENCES
1. A. G. Hassam, J. P. Rivers, and M. A. Crawford, J. Nutr., 107(4), 519524 (1977).

2. A. G. Hassam, J. P. Rivers, and M. A. Crawford, Nutr. Metab., 21(Suppl 1): 190192
(1977).
3. J. P. Rivers, A. J. Sinclair, and M. A. Crawford, Nature, 258(5531), 171173 (1975).
4. H. Sprecher, Prog. Lipid Res., 20, 1322 (1981).
5. R. R. Brenner, Adv. Exp. Med. Biol., 83, 85101 (1977).
6. G. D. de, I. M. J. de Alaniz, and R. R. Brenner, Lipids, 19(2), 9195 (1984).
7. S. Ayala and R. R. Brenner, Acta Physiol. Lat. Am., 33(3), 193204 (1983).
8. R. Kleiman, F. R. Earle, I. A. Wolff, and Q. Jones, J. Am. Oil Chem. Soc., 41(7), 459460
(1964).
9. J. Janick, J. Simon, J. Quinn, and N. Beaubaire, Herbs, Spices and Med. Plants, 4, 147
167 (1989).
10. K. R. Fell and J. M. Peck, Planta Medica, 16, 2942 (1968).
11. M. L. Fernald, Greys Manual of Botany, 8th ed., D. Van Nostrand Co., New York, NY,
1970, 1198.
12. P. Clough, in Structured and Modified Lipids, F. D. Gunstone, ed., Marcell Dekker, Inc.,
New York, NY, 2000, pp. 75117.
13. M. J. A. Simpson, Ann. Appl. Biol., 123, 105108 (1993).
14. R. El Hafid, S. F. Blade, and Y. Hoyano, Industrial Crops and Products, 16, 193199
(2002).
15. A. De Haro, V. Dominguez, and M. del Rio, J. Herbs, Spices and Med. Plants, 9(4), 297
304 (2002).
16. J. Janick, J. E. Simon, and A. Whipkey, Hort. Sci., 22, 493495 (1987).
17. J. Quinn, J. E. Simon, and J. Janick, J. Am. Soc. Hort. Sci., 114(3), 511515 (1989).
18. A. Whipkey, J. Simon, and J. Janick, J. Am. Oil Chem. Soc., 65(6), 979984 (1988).
19. N. A. Beabaire, Masters Thesis, Purdue Univsersity, West Lafayette, Indiana, 1987.
20. A. Sensidoni, G. Bortolussi, C. Orlando, G. Lognay, P. Fantozzi, and M. Paquot,
Lebensm. Wiss. Technol., 28(3), 343346 (1995).
21. I. Wretnesjo and B. Karlberg, J. Amer. Oil Chem. Soc., 79(11), 10691074 (2002).
22. C. Dodson and F. R. Stermitz, J. Nat. Prod., 49(4), 727728 (1986).
23. M. Herrmann, H. Joppe, and G. Schmaus, Phytochemistry, 60(4), 399402 (2002).
24. O. Parvais and V. Stricht, J. Planar Chromat., 7(1/2), 8082 (1994).
25. H. J. Mierendorff, Fett. Wissenschaft Technologie., 97(1), 3337 (1995).
26. H. W. Hulan, I. Hall, and D. M. Nash, Crop. Res., 27, 19 (1987).
27. W. W. Christie, Industrial Crops and Products, 10, 7383 (1999).
28. M. Wettasinghe, F. Shahidi, and R. Amarowicz, J. Agric. Food Chem., 50(5), 12671271
(2002).
29. B. D. Oomah, M. Busson, D. V. Godfrey, and J. C. G. Drover, Food Chem., 76(1), 3343
(2002).
30. Hemp Oil Canada Inc. (2004) Nutritional Composition. Available: http://hempoilcana-
da.com/compo.html.
31. C. C. J. Culvenor, Trends Pharmacol. Sci., 6(1), 1822 (1985).
32. C. C. J. Culvenor, Aust. J. Chem., 9, 512 (1956).
33. A. A. Seawright, M. P. Hegarty, L. F. James, and R. F. Keeler, in Plant Toxiciology
Proceedings of the Australia / USA Poisonous Plants Symposium Yeerongpilly, Bris-
bane, Australia, 1985.
34. A. Molero Gomez and E. Martinez de la Ossa, Chem. Eng. J., 88(13), 103109
(2002).
35. J. Leyton, P. J. Drury, and M. A. Crawford, Lipids, 22(8), 553558 (1987).
36. L. Chilton, M. E. Surette, D. D. Swan, A. N. Fonteh, M. M. Johnson, and F. H. Chilton, J.
Immunol., 156(8), 29412947 (1996).
37. A. G. Hassam, A. J. Sinclair, and M. A. Crawford, Lipids, 10(7), 417420 (1975).
38. A. G. Hassam and M. A. Crawford, Nutr. Metab., 20(2), 112116 (1976).
39. M. M. Engler, J. W. Karaian, and N. Salem, Jr., Nutr. Res., 11, 753763 (1991).
40. J. D. Palombo, S. J. DeMichele, E. E. Lydon, T. J. Gregory, P. L. Banks, R. A. Forse, and
B. R. Bistrian, Am. J. Clin. Nutr., 63(2), 208219 (1996).
REFERENCES 113
41. J. D. Palombo, S. J. DeMichele, E. Lydon, and B. R. Bistrian, J. Parenter. Enteral Nutr.,

21(3), 123132 (1997).
42. D. E. Barre and B. J. Holub, Lipids, 27(5), 315320 (1992).
43. D. E. Barre and B. Holub, J. Nutr. Res., 12(10), 11811194 (1992).
44. J. Martens-Lobenhoffer and F. P. Meyer, Int. J. Clin. Pharmacol. Ther., 36(7), 363366
(1998).
45. M. S. Manku, N. Morse-Fisher, and D. F. Horrobin, Eur. J. Clin. Nutr., 42(1), 5560
(1988).
46. M. M. Engler and M. B. Engler, Prostaglandins Leukot. Essent. Fatty Acids, 59(1), 1115
(1998).
47. L. Chilton, M. E. Surette, D. D. Swan, A. N. Fonteh, M. M. Johnson, and F. H. Chilton, J.
Immunol., 156(8), 29412947 (1996).
48. L. D. Lawson and B. Hughes, Lipids, 23(4), 313317 (1988).
49. D. Raederstorff and U. Moser, Lipids, 27(12), 10181023 (1992).
50. S. Chung, S. Kong, K. Seong, and Y. Cho, J. Nutr., 132(10), 30903097 (2002).
51. J. C. Rose, M. Johnson, P. W. Ramwell, and P. A. Kot, Proc. Soc. Exp. Biol. Med., 148(4),
12521256 (1975).
52. B. A. Scholkens, D. Gehring, V. Schlotte, and U. Weithmann, Prostaglandins Leukot.
Med., 8(3), 273285 (1982).
53. D. E. Mills and R. Ward, Proc. Soc. Exp. Biol. Med., 176(1), 3237 (1984).
54. D. E. Mills, K. M. Prkachin, K. Harvey, and R. Ward, J. Hypertens., 3(2), 111116
(1989).
55. G. C. Leng, A. J. Lee, F. G. R. Fowkes, R. G. Jepson, and G. D. O. Low, Clin. Nutr., 17(6),
265271 (1998).
56. M. M. Engler, M. B. Engler, S. Erickson, and S. M. Paul, J. Hypertens., 10(10), 1197
1204 (1992).
57. B. A. Scholkens, D. Gehring, V. Schlotte, and U. Weithmann, Prostaglandins Leukot.
Med., 8(3), 273285 (1982).
58. M. M. Engler, M. B. Engler, S. K. Erickson, and S. M. Paul, J. Hypertens., 10(10), 1197
1204 (1992).
59. P. Hoffmann, H. U. Block, J. Beitz, C. Taube, W. Forster, P. Wortha, P. Singer,
E. Naumann, and H. Heine, Lipids, 21(12), 733737 (1986).
60. M. M. Engler, M. Schambelan, M. B. Engler, D. L. Ball, and T. L. Goodfriend, Proc. Soc.
Exp. Biol. Med., 218(3), 234237 (1998).
61. D. E. Mills, M. Mah, R. P. Ward, B. L. Morris, and J. S. Floras, Am. J. Physiol.,
259(6 Pt 2), R1164R1171 (1990).
62. J. Chaintreuil, L. Monnier, C. Colette, P. Crastes de Paulet, A. Orsetti, D. Spielmann,
F. Mendy, and A. Crastes de Paulet, Hum. Nutr. Clin. Nutr., 38(2), 121130 (1984).
63. P. Singer, P. Hoffmann, J. Beitz, W. Forster, M. Wirth, and W. Godicke, Prostaglandins
Leukot. Med., 22(2), 173177 (1986).
64. M. Boberg, B. Vessby, and I. Selinus, Acta. Med. Scand., 220(2), 153160 (1986).
65. J. Viikari and A. Lehtonen, Int. J. Clin. Pharmacol. Ther. Toxicol., 24(12), 668670 (1986).
66. M. Guivernau, N. Meza, P. Barja, and O. Roman, Prostaglandins Leukot. Essent. Fatty
Acids, 51(5), 311316 (1994).
67. M. Lagarde, M. Burtin, M. Dechavanne, B. Sicard, and B. Coiffier, Prostaglandins Med.,

4(3), 177183 (1980).
68. P. B. A. Kernoff, A. L. Willis, K. J. Stone, J. A. Davies, and G. P. McNicol, Br. Med. J.,
2(6100), 14411444 (1977).
69. J. C. Rose, M. Johnson, P. W. Ramwell, and P. A. Kot, Proc. Soc. Exp. Biol. Med., 148(4),
12521256 (1975).
70. T. Ishikawa, Y. Fujiyama, O. Igarashi, M. Morino, N. Tada, A. Kagami, T. Sakamoto,
M. Nagano, and H. Nakamura, Atherosclerosis, 75(23), 95104 (1989).
71. J. Jantti, T. Nikkari, T. Solakivi, H. Vapaatalo, and H. Isomaki, Ann. Rheum. Dis., 48(2),
124127 (1989).
72. D. F. Horrobin and M. S. Manku, Lipids, 18(8), 558562 (1983).
73. M. Fukushima, K. Shimada, E. Ohashi, H. Saitoh, K. Sonoyama, M. Sekikawa, and
M. Nakano, J. Nutr. Sci. Vitaminol. (Tokyo), 47(3), 228235 (2001).
74. K. Koba, J. W. Liu, L. T. Chuang, S. N. Anderson, T. Bowman, E. Bobik, Jr., M. Sugano,
and Y. S. Huang, Biosci. Biotechnol. Biochem., 64(12), 25382542 (2000).
75. S. Renaud, L. Mcgregor, R. Morazain, C. Thevenon, C. Benoit, E. Dumont, and
F. Mendy, Atherosclerosis, 45(1), 4351 (1982).
76. A. M. Sadi, T. Toda, H. Oku, and S. Hokama, Exp. Anim., 45(1), 5562 (1996).
77. Y. Y. Fan, K. S. Ramos, and R. S. Chapkin, Arterioscler. Thromb. Vasc. Biol., 15(9),
13971403 (1995).
78. C. Maziere, F. Dantin, M. A. Conte, J. Degonville, D. Ali, F. Dubois, and J. C. Maziere,
Biochem. J., 336(Pt 1) 5762 (1998).
79. Y. Y. Fan, K. S. Ramos, and R. S. Chapkin, J. Nutr., 131(6), 16751681 (2001).
80. Y. Li, J. X. Kang, and A. Leaf, Prostaglandins, 54(2), 511530 (1997).
81. J. S. Charnock, P. L. McLennan, M. Y. Abeywardena, and W. F. Dryden, Ann. Nutr.
Metab., 29(5), 306318 (1985).
82. A. R. Kristal, J. H. Cohen, P. Qu, and J. L. Stanford, Cancer Epidemiol. Biomarkers
Prev., 11(8), 719725 (2002).
83. F. Levi, C. Pasche, F. Lucchini, and C. La Vecchia, Ann. Oncol., 13(3), 369373 (2002).
84. A. J. Littman, S. A. Beresford, and E. White, Cancer Causes Control, 12(8), 691702
(2001).
85. C. Byrne, H. Rockett, and M. D. Holmes, Cancer Epidemiol. Biomarkers Prev., 11(3),
261265 (2002).
86. R. Jarvinen, P. Knekt, T. Hakulinen, H. Rissanen, and M. Heliovaara, Br. J. Cancer,
85(3), 357361 (2001).
87. S. A. Smith-Warner, D. Spiegelman, H. O. Adami, W. L. Beeson, P. A. van den Brandt,
A. R. Folsom, G. E. Fraser, J. L. Freudenheim, R. A. Goldbohm, S. Graham, L. H. Kushi,
A. B. Miller, T. E. Rohan, F. E. Speizer, P. Toniolo, W. C. Willett, A. Wolk, A. Zeleniuch-
Jacquotte, and D. J. Hunter, Int. J. Cancer, 92(5), 767774 (2001).
88. N. Dippenaar, J. Booyens, D. Fabbri, P. Engelbrecht, and I. E. Katzeff, S. Afr. Med. J.,
62(19), 683685 (1982).
89. J. Booyens, N. Dippenaar, D. Fabbri, P. Engelbrecht, C. C. Louwrens, and I. E. Katzeff,
S. Afr. Med. J., 65(15), 607612 (1984).
90. P. Coffino, H. R. Bourne, and G. M. Tomkins, Am. J. Pathol., 81(1), 199204 (1975).
REFERENCES 115
91. M. E. Begin, U. N. Das, G. Ells, and D. F. Horrobin, Prostaglandins Leukot. Essent. Fatty
Acids, 19(2), 177186 (1985).
92. J. H. Lee and M. Sugano, Nutr. Rep. Int., 34(6), 10411049 (1986).
93. S. Abou El-Ela, K. W. Prasse, R. Carroll, and O. R. Bunce, Lipids, 22(12), 10411044
(1987).
94. C. A. Gonzalez, J. M. Sanz, G. Marcos, S. Pita, E. Brullet, E. Saigi, A. Badia, A. Agudo,
and E. Riboli, Cancer Epidemiol. Biomarkers Prev., 2(2), 157158 (1993).
95. J. H. Botha, K. M. Robinson, N. Ramchurren, and R. J. Norman, Prostaglandins Leukot.
Essent. Fatty Acids, 35(2), 119123 (1989).
96. F. S. Kenny, S. E. Pinder, I. O. Ellis, J. M. Gee, R. I. Nicholson, R. P. Bryce, and J. F.
Robertson, Int. J. Cancer, 85(5), 643648 (2000).
97. P. J. du Toit, C. H. van Aswegen, and D. J. du Plessis, Prostaglandins Leukot. Essent.
Fatty Acids, 51(2), 121124 (1994).
98. P. J. du Toit, C. H. van Aswegen, and D. J. du Plessis, Prostaglandins Leukot. Essent.
Fatty Acids, 55(3), 173177 (1996).
99. W. G. Jiang, S. Hiscox, M. B. Hallett, C. Scott, D. F. Horrobin, and M. C. Puntis, Br. J.
Cancer, 71(4), 744752 (1995).
100. W. G. Jiang, S. Hiscox, R. P. Bryce, D. F. Horrobin, and R. E. Mansel, Br. J. Cancer,
77(5), 731738 (1998).
101. W. G. Jiang, A. Redfern, R. P. Bryce, and R. E. Mansel, Prostaglandins Leukot. Essent.
Fatty Acids, 62(2), 119127 (2000).
102. T. Liang and S. Liao, Biochem. J., 285(Pt 2), 557562 (1992).
103. L. Kleinberg, S. A. Grossman, K. Carson, G. Lesser, A. ONeill, J. Pearlman, P. Phillips,
T. Herman, and M. Gerber, J. Clin. Oncol., 20(14), 31493155 (2002).
104. A. A. Brandes, U. Basso, and L. M. Pasetto, Expert. Rev. Anticancer Ther., 1(3), 357370
(2001).
105. M. R. C. Naidu, U. N. Das, and A. Kishan, Prostaglandins Leukot. Essent. Fatty Acids,
45, 181184 (1992).
106. U. N. Das, W. S. K. Prasad, and D. R. Reddy, Cancer Let., 94(2), 147155
(1995).
107. C. F. van der Merwe, J. Booyens, H. F. Joubert, and C. A. van der Merwe, Prostaglandins
Leukot. Essent. Fatty Acids, 40(3), 199202 (1990).
108. J. S. Falconer, K. C. Fearon, J. A. Ross, and D. C. Carter, World Rev. Nutr. Diet., 76, 74
76 (1994).
109. P. Sangeetha, U. N. Das, R. Koratkar, G. Ramesh, M. Padma, and G. Sravan Kumar,
Cancer Let., 63, 189198 (1992).
110. H. A. Leaver, J. R. Williams, J. W. Ironside, E. P. Miller, A. Gregor, B. H. Su, R. J.
Prescott, and I. R. Whittle, Eur. J. Clin. Invest., 29(3), 220231 (1999).
111. M. de Kock, M. L. Lottering, and J. C. Seegers, Prostaglandins Leukot. Essent. Fatty
Acids, 51(2), 109120 (1994).
112. M. de Kock, M. L. Lottering, C. J. Grobler, T. C. Viljoen, M. le Roux, J. C. Seegers,
Prostaglandins Leukot. Essent. Fatty Acids, 55(6), 403411 (1996).
113. W. G. Jiang, R. P. Bryce, D. F. Horrobin, and R. E. Mansel, Int. J. Oncol., 13(3), 611617
(1998).
114. J. C. Seegers, M. de Kock, M. L. Lottering, C. J. Grobler, D. H. van Papendorp, Y. Shou,

R. Habberset, and B. E. Lehnert, Prostaglandins Leukot. Essent. Fatty Acids, 56(4), 271
280 (1997).
115. D. F. Horrobin, Wien. Klin. Wochenschr., 100(14), 471477 (1988).
116. L. Chilton, M. E. Surette, D. D. Swan, A. N. Fonteh, M. M. Johnson, and F. H. Chilton,
J. Immunol., 156(8), 29412947 (1996).
117. V. A. Ziboh and M. P. Fletcher, Am. J. Clin. Nutr., 55(1), 3945 (1992).
118. S. Kaku, K. Ohkura, S. Yunoki, M. Nonaka, H. Tachibana, M. Sugano, and K. Yamada,
Prostaglandins Leukot. Essent. Fatty Acids, 65(4), 205210 (2001).
119. D. Santoli and R. B. Zurier, J. Immunol., 143(4), 13031309 (1989).
120. W. V. Williams, H. Rosenbaum, and R. B. Zurier, Pathobiology, 64(1), 2731 (1996).
121. D. M. DeMarco, D. Santoli, and R. B. Zurier, J. Leukocyte Biol., 56(5), 612615
(1994).
122. D. Rotondo, C. R. Earl, K. J. Laing, and D. Kaimakamis, Biochim. Biophys. Acta,
1223(2), 185194 (1994).
123. D. Rothman, H. Allen, L. Herzog, A. Pilapil, C. M. Seiler, and R. B. Zurier, Cytokine,
9(12), 10081012 (1997).
124. P. DeLuca, R. G. Rossetti, C. Alavian, P. Karim, and R. B. Zurier, J. Investig. Med., 47(5),
246250 (1999).
125. R. K. Furse, R. G. Rossetti, and R. B. Zurier, J. Immunol., 167(1), 490496 (2001).
126. R. G. Rossetti, C. M. Seiler, M. Laposata, and R. B. Zurier, Clin. Immunol. Immuno-
pathol., 76(3), 220224 (1995).
127. D. Vassilopoulos, R. B. Zurier, R. G. Rossetti, and G. C. Tsokos, Clin. Immunol.
Immunopathol., 83(3), 237244 (1997).
128. D. Wu, M. Meydani, L. Leka, Z. Nightingale, G. Handelman, and J. Blumberg, Am. J.
Clin. Nutr., 70(4), 536543 (1999).
129. J. L. Nerad, S. N. Meydani, and C. A. Dinarello, Cytokine, 3, 513 (1991).
130. R. B. Zurier, R. G. Rossetti, C. M. Seiler, and M. Laposata, Prostaglandins Leukot.
Essent. Fatty Acids, 60(56), 371375 (1999).
131. G. A. Tate, B. F. Mandell, R. A. Karmali, M. Laposata, D. G. Baker, H. R. Schumacher,
Jr., and R. B. Zurier, Arthritis Rheum., 31(12), 15431551 (1988).
132. G. A. Tate and R. B. Zurier, Agents Actions, 43(12), 3538 (1994).
133. T. M. Hansen, A. Lerche, V. Kassis, I. Lorenzen, and J. Sondergaard, Scand. J.
Rheumatol., 12(2), 8588 (1983).
134. J. J. Belch, D. Ansell, R. Madhok, A. ODowd, and R. D. Sturrock, Ann. Rheum. Dis.,
47(2), 96104 (1988).
135. S. Pullman-Mooar, M. Laposata, D. Lem, R. T. Holman, L. J. Leventhal, D. DeMarco,
and R. B. Zurier, Arthritis Rheum., 33(10), 15261533 (1990).
136. L. J. Leventhal, E. G. Boyce, and R. B. Zurier, Ann. Intern. Med., 119(9), 867873
(1993).
137. R. B. Zurier, R. G. Rossetti, E. W. Jacobson, D. M. DeMarco, N. Y. Liu, J. E. Temming,
B. M. White, and M. Laposata, Arthritis Rheum., 39(11), 18081817 (1996).
138. K. V. Kumar, S. M. Rao, R. Gayani, I. K. Mohan, and M. U. Naidu, Clin. Chim. Acta,
298(12), 111120 (2000).
REFERENCES 117
139. J. E. Gadek, S. J. DeMichele, M. D. Karlstad, E. R. Pacht, M. Donahoe, T. E. Albertson,

C. Van Hoozen, A. K. Wennberg, J. L. Nelson, and M. Noursalehi, Crit. Care Med.,
27(8), 14091420 (1999).
140. M. M. Murray, M. Kumar, T. Greogory, P. Banks, H. D. Tazelaar, and S. J. DeMichele,
Am. J. Physiol., 269(6 pt 2), H2090H2099 (1995).
141. P. Mancuso, J. Whelan, S. J. DeMichele, C. C. Snider, J. A. Guszcza, K. J. Claycombe,
G. T. Smith, T. J. Gregory, and M. D. Karlstad, Crit. Care Med., 25(3), 523532 (1997).
142. G. O. Burr and M. M. Burr, J. Biol. Chem., 82(2), 345367 (1929).
143. R. S. Chapkin, V. A. Ziboh, and J. L. McCullough, J. Nutr., 117(8), 13601370 (1987).
144. A. E. Hansen, Proc. Soc. Exp. Biol. Med., 31, 160161 (1933).
145. M. S. Manku, D. F. Horrobin, N. Morse, V. Kyte, K. Jenkins, S. Wright, and J. L. Burton,
Prostaglandins Leukot. Med., 9(6), 615628 (1982).
146. S. Wright and C. Bolton, Br. J. Nutr., 62(3), 693697 (1989).
147. L. Businco, M. Ioppi, N. L. Morse, R. Nisini, and S. Wright, J. Allergy Clin. Immunol.,
91(6), 11341139 (1993).
148. S. Wright and J. L. Burton, The Lancet, 2(8308), 11201122 (1982).
149. M. S. Manku, D. F. Horrobin, N. L. Morse, S. Wright, and J. L. Burton, Br. J. Dermatol.,
110(6), 643648 (1984).
150. L. Schafer and K. Kragballe, Lipids, 26(7), 557560 (1991).
151. D. H. Scarff and D. H. Lloyd, Vet. Record, 131(5), 9799 (1992).
152. A. Fiocchi, M. Sala, P. Signoroni, G. Banderali, C. Agostoni, and E. Riva, J. Int. Med.
Res., 22(1), 2432 (1994).
153. S. Borrek, A. Hildebrandt, and J. Forster, Klin. Padiatr., 209(3), 100104 (1997).
154. P. J. Hartop and C. Prottey, Br. J. Dermatol., 95(3), 255264 (1976).
155. A. Tollesson and A. Frithz, Acta Derm. Venereol., 73(1), 1820 (1993).
156. B. M. Henz, S. Jablonska, P. C. van de Kerkhof, G. Stingl, M. Blaszczyk, P. G.
Vandervalk, R. Veenhuizen, R. Muggli, and D. Raederstorff, Br. J. Dermatol., 140(4),
685688 (1999).
157. A. Takwale, E. Tan, S. Agarwal, G. Barclay, I. Ahmed, K. Hotchkiss, J. R. Thompson, T.
Chapman, and J. Berth-Jones, Br. Med. J., 327(7428), 1385 (2003).
158. B. M. Henz, S. Jablonska, P. C. van de Kerkhof, G. Stingl, M. Blaszczyk, P. G.
Vandervalk, R. Veenhuizen, R. Muggli, and D. Raederstorff, Br. J. Dermatol.,
140(4), 685688 (1999).
159. M. D. Giron, F. J. Mataix, and M. D. Suarez, Biochim. Biophys. Acta., 1045, 6973
(1990).
160. K. L. Campbell and G. P. Dorn, Res. Vet. Sci., 53, 172178 (1992).
161. C. C. Miller and V. A. Ziboh, Biochem. Biophys. Res. Commun., 154(3), 967974 (1988).
162. V. A. Ziboh, C. C. Miller, and Y. Cho, Am. J. Clin. Nutr., 71(1 Suppl), 361S366S (2000).
163. C. R. Charman, A. J. Venn, and H. C. Williams, Br. J. Dermatol., 146(6), 10571060
(2002).
164. J. W. Hopewell, G. J. van den Aardweg, G. M. Morris, M. Rezvani, M. E. Robbins, G. A.
Ross, E. M. Whitehouse, C. A. Scott, and D. F. Horrobin, Int. J. Radiat. Oncol. Biol.
Phys., 30(5), 11191125 (1994).
165. S. Ayala and R. R. Brenner, Acta Physiol. Lat. Am., 25(5), 371378 (1975).
166. J. P. Poisson, Enzyme, 34(1), 114 (1985).

167. O. Mercuri, R. O. Peluffo, and R. R. Brenner, Biochim. Biophys. Acta, 116(2), 409411
(1966).
168. M. Arisaka, O. Arisaka, and Y. Yamashiro, Prostaglandins Leukot. Essent. Fatty Acids,
43(3), 197201 (1991).
169. P. O. Julu, J. Diabet. Complications, 2(4), 185188 (1988).
170. P. O. Julu and A. Mutamba, J. Neurol. Sci., 106(1), 5659 (1991).
171. N. E. Cameron, M. A. Cotter, and S. Robertson, Diabetes, 40(5), 532539 (1991).
172. N. E. Cameron and M. A. Cotter, Acta Diabetologica, 31(4), 220225 (1994).
173. N. E. Cameron and M. A. Cotter, Am. J. Physiol., 271(3 Pt 1), E471E476 (1996).
174. G. A. Jamal and H. Carmichael, Diabetic Medicine, 7(4), 319323 (1990).
175. H. Keen, J. Payan, J. Allawi, J. Walker, G. A. Jamal, A. I. Weir, L. M. Henderson, E. A.
Bissessar, P. J. Watkins and M. Sampson, Diabetes Care, 16(1), 815 (1993).
176. B. Singh, J. Lauzon, J. Venkatraman, A. B. Thomson, R. V. Rajotte, and M. T. Clandinin,
Diabetes Res., 8(3), 129134 (1988).
177. B. B. Oon, D. Muggleston, and A. Warley, Exp. Physiol., 77(1), 185190 (1992).
178. D. N. Ugarriza, S. Klingner, and S. OBrien, Nurse Pract., 23(9), 4052 (1998).
179. M. G. Brush, S. J. Watson, D. F. Horrobin, and M. S. Manku, Am. J. Obstet. Gynecol.,
150(4), 363366 (1984).
180. J. Puolakka, L. Makarainen, L. Viinikka, and O. Ylikorkala, J. Reprod. Med., 30(3), 149
153 (1985).
181. S. K. Khoo, C. Munro, and D. Battistutta, Med. J. Aust., 153(4), 189192 (1990).
182. A. A. Leaf, M. J. Leighfield, K. L. Costeloe, and M. A. Crawford, Early Hum. Dev.,
30(3), 183191 (1992).
183. P. Rump, R. P. Mensink, A. D. M. Kester, and G. Hornstra, Am. J. Clin. Nutr., 73(4), 797
806 (2001).
184. L. Wang, Y. Shimizu, S. Kaneko, S. Hanaka, T. Abe, H. Shimasaki, H. Hisaki, and
H. Nakajima, Pediatrics. Int., 42(1), 1420 (2000).
185. D. J. Barker, P. D. Winter, C. Osmond, B. Margetts, and S. J. Simmonds, Lancet, 2(8663),
577580 (1989).
186. D. W. Lamont, L. Parker, M. A. Cohen, M. White, S. M. Bennett, N. C. Unwin, A. W. Craft,
and K. G. Alberti, Public Health, 112(2), 8593 (1998).
187. A. A. Jackson, Adv. Exp. Med. Biol., 478, 4155 (2000).
188. P. Rump, C. Popp-Snijders, R. J. Heine, and G. Hornstra, Diabetologia, 45(3), 349355
(2002).
189. J. M. Rabkin, C. L. Corless, H. R. Rosen, and A. J. Olyaei, Am. J. Surg., 183(5), 595599
(2002).
190. T. Markham, A. Watson, and S. Rogers, Clin. Exp. Dermatol., 27(2), 111114
(2002).
191. D. E. Mills, M. Mah, R. P. Ward, B. Morris, and J. Floras, Am. J. Physiol., 259(6 Pt 2),
R1164R1171 (1990).
192. P. Morphake, J. Bariety, I. Darlametsos, and G. Tsipas, Prostaglandins Leukot. Essent.
Fatty Acids, 50(1), 2935 (1993).
193. S. Vartak, M. E. Robbins, and A. A.Spector, Lipids, 32(3), 283292 (1997).
REFERENCES 119
194. J. A. Menendez, M. D. Barbacid, S. Montero, E. Sevilla, E. Escrich, M. Solanas,

H. Cortes-Funes, and R. Colomer, Eur. J. Cancer, 37(3), 402413 (2001).
195. J. A. Menendez, S. Ropero, M. M. del Barbacid, S. Montero, M. Solanas, E. Escrich,
H. Cortes-Funes, and R. Colomer, Breast Cancer Res. Treat., 72(3), 203219 (2002).
196. F. S. Kenny, J. M. Gee, R. I. Nicholson, I. O. Ellis, T. M. Morris, S. A. Watson, R. P. Bryce,
and J. F. Robertson, Int. J. Cancer, 92(3), 342347 (2001).
197. S. Ikushima, F. Fujiwara, S. Todo, and S. Imashuku, Anticancer Res., 10(4), 10551059
(1990).
198. Q. Y. Liu and B. K. Tan, Life Sci., 67(10), 12071218 (2000).
199. S. Kaku, S. Yunoki, K. Ohkura, M. Sugano, M. Nonaka, H. Tachibana, and K. Yamada,
Biosci. Biotechnol. Biochem., 65(2), 315321 (2001).
200. K. S. Vaddadi, Prostaglandins Med., 6(4), 375379 (1981).
201. S. P. Senanayake and Shahidi F, J. Food Lipids, 7, 5161 (2000).
202. J. C. Philips and Y. S. Huang, in Gamma-linolenic acid: Metabolism and its roles in
Nutrition and Medicine, Y. S. Huang and D. E. Mills, eds. AOCS Press, Champaign, IL.
1996, pp. 113.
203. F. C. Huang, Y. H. Ju, and J. C. Chiang, J. Am. Oil Chem. Soc., 76(7), 833837 (1999).
204. D. Cook, D. Grierson, C. Jones, A. Wallace, G. West, and G. Tucker, Mol. Biotechnol.,
21(2), 123128 (2002).
205. A. S. Reddy and T. L. Thomas, Nat. Biotechnol., 14(5), 639642 (1996).
206. O. Sayanova, M. Smith, P. Lapinskas, and A. K. Stobart, Proc. Natl. Acad. Sci. USA,
94(8), 42114216 (1997).
207. P. E. Wainwright, Y. S. Huang, S. J. DeMichele, H. Xing, J. W. Liu, L. T. Chuang, and J.
Biederman, Lipids, 38(2), 171178 (2003).
208. S. P. Senanayake and F. Shahidi, J. Agric. Food Chem., 50(3), 477483 (2002).
209. K. T. Mok, A. Maiz, K. Yamazaki, J. Sobrado, V. K. Babayan, L. L. Moldawer, B. R.
Bistrian, and G. L. Blackburn, Metabolism, 33(10), 910915 (1984).
210. S. A. Spurvey, S. P. Senanayake, and F. Shahidi, J. Am. Oil Chem. Soc., 78(11), 1105
1112 (2001).
5
Oils from Microorganisms
James P. Wynn1 and Colin Ratledge2
1
Martek Biosciences Corporation
Columbia, Maryland
2
Lipid Research Centre, University of Hull,
Hull, United Kingdom
1. GENERAL INTRODUCTION AND BACKGROUND

INFORMATION
As this chapter is the only one in this new edition that is solely devoted to the poten-
tial use of microorganisms as sources of oils and fats, we have felt it appropriate to
preface our review by including a brief outline of the importance of microbial bio-
technology in general so that the extent to which microorganisms can now be grown
and the amounts of products that are produced can be appreciated by the general
reader who may not be familiar with the field of microbial biotechnology.
1.1. Microorganisms and Biotechnology

Microorganisms range from bacteria, which are known as the prokaryotic microbes
as they do not contain a defined nucleus, although they do, of course, contain DNA,
to the eukaryotic microorganisms that do have a nucleus and range from the simple
yeasts to fungi, which may show more complex structural variations and be capable
of some limited differentiation. Some fungi can differentiate into macromolecular
forms and give rise to, for example, the mushrooms and toadstools. There are also a
range of microorganisms in the aquatic environments, and these will include both
121
122 OILS FROM MICROORGANISMS
prokaryotic and eukaryotic forms. Many of these will only grow in the presence of
light (phototrophs), although a few will grow in the dark if some form of fixed car-
bon (e.g., sugar) is supplied (i.e., they will grow heterotrophically like other micro-
organsisms). Thus, the term microbe applies to a very wide range of living cells,
although all would be regarded, at least in the first instance, as being unicellular
and, as such, could only be seen as individual cells when examined down a micro-
scope.
The advantages of microorganisms, which are used extensively in biotechnology
systems, are as follows:
The range of products that they can producefrom simple molecules like
ethanol and citric acid, to complex proteins, carbohydrates, etc.
They can be grown phenomenally fast: some bacteria can divide once every
10 minutes, although yeasts and fungi may take several hours to accomplish
this doubling of cell numbers;
They can be randomly mutated, by chemicals and other mutagens, so that
products can be produced in vastly increased quantities. Penicillin, for
example, when first produced, was at a few milligrams per liter; now the
current mutated strains will produce up to 100-g penicillin/L.
Microorganisms can also be modified by genetic manipulation where genes
(that is, specific sections of DNA coding for the synthesis of particular
proteins) taken from other living cells, i.e., other microorganisms, plants, or
even animals cells, can be easily incorporated into a microorganism so that it
will now produce not only its own proteins, but also the foreign protein.
Microorganisms can therefore produce, at least in theory, any product that we

currently can identify in any living cell. They will produce them faster, more safely
(you do not need to spray them with pesticides, herbicides, etc.) than any other
system. Moreover, the products can be produced to a guaranteed quality on a
year-round basis as productions do not depend on the vagaries of the weather or
climate. They are the ultimate controlled living system. This, then, is the world
of microbial biotechnology (1). The scale of cultivation can be in small fermenters
(about 15 m3) for production of very highvalue, lowvolume materials, such as
some of the current healthcare products, up to fermenters of 100500 m3, which can
be used to produce bulk materials such as citric acid, ethanol, and even whole cells
destined for animal or human foodthe so-called single cell proteins.
1.2. Microbial Oils

Microorganisms, like every other living cell system, produce lipids. All cells are
surrounded by membranes that require the synthesis of fatty acids (except in the
case of the most ancient of bacteriathe Archaeawhere long-chain, branched
terpenoid structures are used), which are then attached to glycerol 3-phosphate giv-
ing rise to the phospholipids and triacylglycerols. This system of fatty acid and lipid
GENERAL INTRODUCTION AND BACKGROUND INFORMATION 123
Figure 1. (a) Cells of an oleaginous yeast, Cryptococcus curvatus (formerly Candida curvata), in
which the lipid droplets constituting about 70% of the cell weight of the cells can be clearly seen.
(b) Electron micrograph of extracted oil from the same yeast shows a possible boundary layer
around the oil droplets. The oil is virtually pure triacylglycerol. (This figure is available in full color
at http://www.mrw.interscience.wiley.com/biofp.)
biosynthesis is then used, in some cases, for the overproduction of triacylglycerols

that then serve as reserve storage materials within the microbial cell (see Figure 1).
These are the so-called oleaginous species, and the oils are then not only useful to
the microorgansism to reuse during any subsequent period of starvation, but also
may be considered as sources of these commodities. The oils are referred to as
single cell oils (SCO), which is a euphemism similar to the term single cell pro-
tein used to indicate protein derived from microbial (single) cells. Single cell oils
is now a generally accepted term and is used in preference to microbial oil, which
may often create an undesirable impression with an otherwise unappreciative public.
Microorganisms have been considered as potential sources of oils since the early
decades of the twentieth century when scientists, particularly in Germany, began to
explore them as an alternative to plant oils, which were increasingly in short supply
because of the advent of two world wars. Indeed, during the Second World War,
processes had been developed that produced small amounts of the oils and fats,
although, because of the lack of large-scale fermentation technology, this never pro-
ceeded beyond a demonstration scale (2, 3). None of the microbial fat seems to
have been used for human consumption, however. Nevertheless, there was consid-
erable interest in microbial oils and fats and in exploring their potential as alterna-
tives to plant seed oils, which were poorly developed at this time.
Work on understanding which were the best oil-producing strains of microorgan-
ism to use and how they had to be cultivated to induce the highest possible lipid
contents was carried out during both war-time and peace-time in the first half of
the twentieth century, so that by the late 1950s, a shrewd understanding of the range
of oils that could be produced had been gained as well as knowing how these organ-
isms had to be grown to produce the highest yields (2). However, the economic
value of the microbial oils was always the Achilles heel to their commercialization.
Advances in agriculture, particularly after the end of the Second World War in
1945, meant that all agricultural commodities, and not just oils and fats, became
cheaper and more plentiful than ever before. As the microorganisms being consid-
ered for oil production had to be grown on sugar, which was an agricultural product,
it was clear that it was nonsensical to grow one field of sugar cane (or sugar beet)
for it to be converted at the ratio of 5 tons of sugar to 1 ton of oil in an expensive
bioreactor (i.e., the fermenter) to produce oil that could be produced just as easily
using the self-same field to grow a plant oilseed crop.
Thus, by the 1960s, there seemed to be no arguable case in favor of microbial
oils being an economic proposition. And yet today, in the first decade of the twenty-
first century, we have at least four microbial oils that are, or have been, in full-scale
production, with the prospect that others may soon follow.
This change in attitude toward microbial oils has originated mainly, but not
exclusively, by the appreciation of the need for specific polyunsaturated fatty acids
(PUFAs) to be included in our diets, both for our infants and babies as well as our-
selves in later life. Such polyunsaturated fatty acids, which have been known to
exist in microorganisms for many years, are now in demand. They cannot be easily
produced by plants and, in some cases, cannot be produced at all by plants. The
sources of these PUFAs then are from animals and, in particular, fish. Fish, though,
are a dwindling resource and the oils, which are mainly obtained from their livers
or the bodies of fatty fish, can be contaminated with various pollutants, including
organo-mercury compounds, dioxanes, and other materials that we would do best to
avoid in our own diets. Thus, the commercial exploitation of microbial oils has
originated by them being destined for human consumption and by them not being
readily available from traditional sources, either plants or animals.
Coupled with the realization that microorganisms can be a unique source of cer-
tain desirable oils, has been the advent of large-scale fermentation technology
throughout the latter half of the last century. This has culminated in the ability to
design, build, and operate purpose-designed fermenters in excess of 1,000,000-L
capacity for the production of microorganisms to provide a variety of products. A
simple calculation tells us that a 1000-m3 fermenter that produces 100-g cells/L in 4
days would yield a 100 tons of biomass with, perhaps, 40 tons of oil: an annual
Acid Base Antifoam

CENTRIFUGE
cryovial Waste
125 ml
To All Fermentors
SPRAY
DRYER
1L
20 L 200 L PACKAGING
2,000 L 12,000 L 120,000 L DRY BIOMASS
Figure 2. A diagrammatic presentation of a fermentation system for the production of a single

cell oil. The main fermenter is inoculated from a smaller vessel at about 10% (v/v). Additional
nutrients, including glucose, may be added to the microbial culture within the fermenter during
growth. The process operates in a batch mode so that when the cells reach their highest lipid
contents (see Figure 3), the fermenter is emptied, the cells are removed from the broth, and then
finally they are spray-dried. In this form, the oil within the cells is stable and can be extracted by
solvents whenever, and wherever, is convenient. (From 50, with kind permission of the author,
Dr. David J. Kyle, and the publishers.)
yield of over 3000 tons of oil. In agriculture, about 4 square miles (9 km2) of land
would need to be sown with sunflowers or oilseed rape to give the same yield of oil.
A typical large-scale process for the production of SCO is shown diagrammatically
in Figure 2. This would be typical of the current fermentation processes being used
for these types of products.
In this review, we aim to provide details of microbial oils that have been con-
sidered of commercial potential and to describe in more detail those microbial sys-
tems that are in commercial operation to provide key fatty acids for the expanding
nutraceuticals industry.
1.3. Microbial Lipid Production Systems

Microorganisms that accumulate more than 20% of their biomass are known as the
oleaginous species. This value, though, is arbitrary and has no precise numerical
definition. There are probably far more nonoleaginous species than oleaginous
ones; that is, most microorganisms will accumulate, even under the most propitious
conditions, only a few percent of their biomass as lipid. The lipid that is produced in
the oleaginous species is usually in the form of triacylglycerols (see Figure 1) and is
an intracellular reserve supply of both carbon and energy (and perhaps water) to
be used in times of nutrient starvation. The extent of lipid accumulation may range
from the lower limit of about 20% up to 70% of the cell weight being extractable
oil. The range of oil contents of a selected range of microorganisms is given in
Table 1 together with a profile of the constituent fatty acids. More detailed lists
TABLE 1. Lipid Contents and Fatty Acid Profiles of Selected Oleaginous Microorganisms (Compiled from Lists Given in (4)(6), Which
Should be Consulted, if Needed, for Further Information).
Major Fatty Acyl Residues

(Relative % w/w)
Lipid
Content 14:0 16:0 16:1 18:0 18:1 18:2 18:3 18:3 20:4 20:5 22:6
Organism (% w/w) (n-7) (n-9) (n-6) (n-3) (n-6) (n-6) (n-3) (n-3) Others (%)
A. Yeasts
Candida sp. 107 42 Trace 44 5 8 31 9 1
Cryptococcus albidus 65 Trace 12 1 3 73 12
C. curvatus D1 58 Trace 32 15 44 8
Waltomyces lipofer2 64 Trace 37 4 7 48 3
Lipomyces starkeyi 63 Trace 34 6 5 51 3
Rhodosporidium toruloides 66 Trace 18 3 3 66 23:0 (3%)
24:0 (6%)
Rhodotorula glutinis 72 Trace 37 1 3 47 8
Trichosporon beigelii 45 Trace 12 22 50 12
Yarrowia lipolytica 36 Trace 11 6 1 28 51 1
B. Fungi
Zygomycetes
Conidiobolus nanodes 26 1 23 15 25 1 4 4 20:1 (13%)
22:1 (8%)
Entomophthora coronata 43 31 9 2 14 2 1 12:0 (40%)
Cunninghamella japonica 60 Trace 16 14 48 14 8
Notes : Mortierella isabellina 86 1 29 3 55 3 3
Rhizopus arrhizus 57 19 18 6 22 10 12
Mucor alpine-peyron 38 10 15 7 30 9 1 5 20:0 (8%)
20:3 (6%)
Ascomycetes
Aspergillus terreus 57 2 23 Trace 14 40 21
Fusarium oxysporum 34 Trace 17 8 20 46 5
Pellicularia practicolo 39 Trace 8 2 11 72 2
Hyphomycetes
Cladosporium herbarum 49 Trace 31 12 35 18 1
Clavicipitaceae
Claviceps purpurea 60 Trace 23 2 19 8 12-HO-
18:1 (42%)
C. Microalgae and thraustochytrids3
Continued
Prokaryota
Spirulina maxima 22 8 63 2 4 9 12
Spirulina platensis 25 1 26 5 23 10 21
Eukaryota
Chlorella minutissima 28 12 13 21 1 2 3 45
Chlorella vulgaris 52 16 2 58 9 14
Crypthecodium cohnii 3 50 16 16 1 21 1 40
Isochrysis galbana 23 12 10 11 3 2 <1 25 11 18:4 (11%)
Monodus subterraneus 20 19 10 5 2 <1 14 34
Nannochloropsis oculata 45 4 15 22 3 1 4 38 14:1 (13%)
Phaeodactylum tricornutum 24 10 21 1 4 1 1 33 4
Porphyridium cruentum 5 30 5 <1 5 1 16 38
Thraustochytrids
Schizochytrium sp.3 40 17 32 8 5 1 28 22.5 (n-5)
(8%)
Thraustochytrium aureum3 25 3 8 16 2 2 3 52
1. This yeast was initially known as Candida curvata, then was renamed Apiotrichum curvatum, and now is regarded as Cryptococcus curvatus.
2. Formerly known as Lipomyces lipofer.
3. Microalgae cultivated phototrophically except for C. cohnii, Schizochytridium sp., and T. aureum, which are nonphotosynthetic and are grown heterotrophically.
are provided in other reviews (46) and in an earlier comprehensive monograph on

microbial lipids (7).
As SCO are usually triacylglycerols that can account for over 90% of the total
lipid in the microbial cell, this makes them potential substitutes for plant oils and
animal fats, although clearly their economic potential will rest on their intrinsic
value. It is to be emphasized, however, that an SCO will only complete with other
commercial sources if it can be shown to be better in some respect or cheaper than
the traditional source.
The biochemical procedure by which microorganisms accumulate lipid is now
understood in some detail and has been recently described (8). Readers wishing
more detailed information should consult this article. In brief, oleaginicity, which
is not shared by all microorganisms, depends on the presence of just a few key
enzymes.
The process of lipid accumulation begins by the cells being grown in a culture
medium in which the supply of a nutrient; usually it is nitrogen in the form of an
ammonium salt, becoming exhausted (this is shown schematically in Figure 3). At
the same time, there remains in the medium a surfeit of carbon, usually glucose or
some other assimilatable carbon source. The cells, as a consequence of this nutrient
exhaustion, can no longer grow and multiply, but they do continue to take up the
sugar present in the medium. It is this surplus sugar that then becomes the source of
carbon for lipid biosynthesis.
Figure 3. Stylistic presentation of the course of lipid accumulation by an oleaginous

microorganism. The concentration of nitrogen (NH3) in the medium is adjusted so that it
becomes exhausted after the first 24 hours growth; after this point, the cells enter the lipid
accumulation phase in which the excess carbon (e.g., glucose) continues to be assimilated by
the cells, and because there is no new cell synthesis because of the lack of nitrogen, the surplus
carbon is converted into lipid, which functions as a reserve of carbon and energy for the cells.
Under the conditions of nitrogen-limited growth, the first requirement is for the
cells to cease generating energy (i.e., ATP), which is no longer needed for the
synthesis of new macromolecules, (e.g., proteins and nucleic acids) as the cells
are unable to grow and divide because of the lack of nitrogen (required for protein
biosynthesis). A key enzyme in the citric acid cycle (Krebs cycle or tricarboxylic
acid cycle), isocitrate dehydrogenase, becomes inactive immediately following the
depletion of nitrogen. This leads to isocitrate not being metabolized, and conse-
quently, both it and citric acid, with which it is in equilibrium, rapidly accumulate.
This event occurs in the mitochondria of the cells, but the citrate is quickly trans-
ported out of this compartment into the cytoplasm of the oleaginous cell and is
immediately cleaved by an enzyme known as ATP:citrate lyase (1):
Citrate ATP coenzyme A ! acetyl-coenzyme A oxaloacetate ADP

inorganic phosphate . . . . . . . . . . . . . . . 1
This is a key reaction as it generates the C2 building unit (acetyl-CoA) for fatty acid
biosynthesis. Without this enzyme being present, there would be no abundant sup-
ply of the acetyl-CoA units and, indeed, many if not all of the nonoleaginous yeasts
do not possess this enzyme. The oxaloacetate generated in this cleavage reaction is
immediately converted to malate by malate dehydrogenase and then the malate, in
turn, is converted to pyruvate by the action of malic enzyme (2).
Malate NADP !pyruvate CO2 NADPH . . . . . . . . . . . . 2
This is the second key enzyme needed to produce high amounts of lipid as the reac-
tion catalysed simultaneously produces the necessary reducing equivalent, NADPH,
by which the growing long acyl chain, derived from acetyl-coenzyme A (see
above), is reduced to the final long-chain fatty acid. Fatty acid biosynthesis, and
consequently lipid accumulation, requires both a continous supply of acetyl-CoA
and reducing power (NADPH), and these are provided by the key reactions men-
tioned above.
How the oleaginous microorganism then produces variable amounts of lipid (see
Table 1) lies with the activity of malic enzyme rather than in ATP:citrate lyase.
Malic enzyme activity is controlled by the genetic makeup of the cell: in cells
that accumulate considerable amounts of lipid (up to and even above 70%), the
gene that controls the synthesis of the malic enzyme is kept switched on all the
time, whereas in the low-lipid cells, the gene is switched off shortly after nitrogen
exhaustion. When this happens, malic enzyme activity quickly disappears and, con-
comitantly, lipid accumulation ceases. Thus, it is now possible to explain not only
the reason why some microorganisms are able to accumulate lipid and other cannot,
but also why the amount of lipid accumulated can vary considerably within the
lipid-accumulating organisms (8). This biochemical information is now leading to
the identification of the key genetic elements that are involved in lipid biosynthesis.
From this will then come the opportunity to genetically modify microorganisms
to give them increased quantities of lipid and, simultaneously, to produce more of

the desired fatty acids.
2. COMMERCIAL MICROBIAL OILS
2.1. Initial Ventures Into Single Cell Oil Production

2.1.1. Fungal Oils Rich in Gamma-Linolenic Acid Although microbial lipids,
SCO, were considered as possible commercial sources of oils and fats for almost
all of the last century, no industrial production of a microbial oil took place
simply because no economically attractive target had been identified. However,
by the late 1970s, certain polyunsaturated fatty acids were being used medicinally as
over-the-counter treatments for a variety of disorders, chief among which was the
use of evening primrose oil (9) as a possible treatment for multiple sclerosis.
Evening primrose oil contains the relatively unusual fatty acid, gamma-linolenic
acid, 18:3n-6. (GLA) (see Table 2). Claims for the efficacy of this oil in the treat-
ment of many disorders, including atopic eczema (10, 11), which go back many
centuries in a historical context, are then attributed to the presence of this fatty
acid, which does not occur in most other plant seed oils. Other claims for the effi-
cacy of GLA-rich oils include the treatment of premenstrual tension, which is the
main selling point of this oil in the United Kingdom, as well as for various cancers
(9, 12). These have added impetus to identifying a cheaper and perhaps more reli-
able source of GLA as the cultivation of evening primrose is not easy, with it being
a biennial crop and which produces very tiny seeds requiring careful harvesting and
processing (13). As it was known that GLA is also present in a large number of
simple fungi, known as the lower fungi or Zygomycetes since the 1940s, a fungal
route to its production was recognized as a feasible alternative to evening primrose
TABLE 2. Fatty Acid Profiles of GLA-SCOsMicrobial Oils Rich in Gamma-Linolenic

Acid (18:3 n-6)in Comparison with Plant Sources of GLA [Adapted from (4) and 6)].
Relative% (w/w) of Major Fatty Acids

Oil
Content 16:0 16:1 18:0 18:1 18:2 18:3 18:3 20:1 22:1
(% w/w) (n-6) (n-6) (n-3)
Mucor circinelloides1 25 22 1 6 40 11 18
Mortierella isabellina2 ND 27 1 6 44 12 8 0.4
Mortierella ramanniana 40 24 5 51 10 10
Mucor hiemalis 30 25 1 10 32 12 15
Evening primrose 16 6 2 8 75 8 0.2 0.2
Borage 30 10 4 16 40 22 0.5 4.5 2.5
Blackcurrant 30 6 1 10 48 17 13
1
Production organism used by J & E Sturge Ltd., Selby, N. Yorks., UK.
2
Production organism used by Idemitzu Ltd., Japan.
NDNot disclosed but believed to be 40% to 50%.
COMMERCIAL MICROBIAL OILS 131
oil. Work was carried out in the authors laboratory at the University of Hull begin-
ning in 1976 by screening a large number of these fungi for their potential to pro-
duce an oil rich in this fatty acid. This led to Mucor circinelloides (formerly known
as Mucor javanicus) being identified as the most productive organism (3): although
this fungus was not the highest producer of GLA, there were three important and
interdependent variables that had to be satisfied:
The organism had to be able to grow quickly and to a high cell density. A
working target was for an organism to achieve more than 50-g dry cell per
liter of fermenter in a time not to exceed 4 days.
The organism had to have more than 20% oil content; otherwise extraction
would be difficult, and the costs of oil production would be increased. Equally,
the extracted oil should be at least 90% as a triacylglycerol so that subsequent
refinement and encapsulation would be relatively easy to accomplish.
The content of GLA in the oil had to be considerably higher than that in
evening primrose oil, which was only about 10%; a working target of 20%
was therefore chosen.
A fourth preference was for the chosen organism to be already recognized as
being safe to use for food purposes: the so-called Generally Recognised As
Safe (GRAS) category. However, this was not regarded as absolutely essential
as the product would be an extracted, purified oil. It would then be the quality
and safety of this that would be assessed rather than the safety of the whole
organism. Nevertheless, one would not want to use a microorganism that had
any association with any disease for very obvious reasons. Also, the chosen
organism should not be a plant pathogen because of possible environmental
risks when growing the organism on a large scale. (All organisms being used
in large-scale cultivations have to be intrinsically safe to handle and even
organisms that could cause allergic reactions in factory operatives are best
avoided.)
The finally chosen production organism, Mucor circinelloides, satisfied all of the
above criteria, including it being a GRAS-status organism as it has a historically
long association with tempe, a well-known oriental food material. The profile of
its fatty acids is given in Table 2 along with its lipid content.
In Japan, Idemitzu Co. Ltd. adopted a slightly different strategy to isolate GLA-
producing organisms. They opted to go primarily for organisms producing high oil
contents and, seemingly, hoped that the GLA also would be high in such organisms.
In the event, although high oil-producing species were found, none produced more
than 10% of the total fatty acids as GLA (see Table 2). For a reason not yet under-
stood, but possibly related to the limited capacity of the cells to generate NADPH
that is used both in fatty acid synthesis and in fatty acid desaturation (see Section
1.3), there is an inverse relationship between oil content and GLA formation. The
higher the oil content of the cells, the lower is the GLA content (see Table 2). The
Idemitzu oil from Mortierella isabellina, therefore, had only half the GLA content
of the U.K. organism, Mucor circinelloides, although the Japanese organism had
twice the oil content of the U.K. organism. Commercially, however, the higher
the GLA content of an oil, the greater its value, and an oil with 20% GLA is prob-
ably more than twice as valuable as an oil with only 10% GLA, so it is more cost-
effective to produce more GLA at the expense of producing less oil.
The GLA-SCO was produced in the United Kingdom by John & E. Sturge,
Selby, North Yorkshire, using Mucor circinelloides. It was first offered for sale in
1985 under the name of Oil of Javanicus, which took account of the oriental origins
of the organism and its original name, Mucor javanicus. The oil was sold com-
mercially throughout the United Kingdom between 1985 and 1990. Production
was at about 1015 tons of oil per year. The fermenters used were 220,000 L
and were normally used for the production of citric acid using Aspergillus
niger. The overall fermentation configuration was similar to that shown in
Figure 1. To grow M. circinelloides for oil production, rather than A. niger for citric
acid, all that had to be done was to reformulate the growth medium so that it
now contained an insufficient supply of nitrogen and so that cell proliferation
would cease after about 3640 hours and lipid accumulation would then begin
(see Figure 2).
The Oil of Javanicus was of high purity and passed all toxicity tests: It was
superior in safety evaluations to conventional plant oils, which always contain
very small residual amounts of pesticides, insecticides, and herbicides from the var-
ious sprays that are used on such crops. These levels, though, are always below the
recommended threshold values laid down by regulatory authorities. Being culti-
vated in fermenters, M. circinelloides does not, of course, need to be sprayed
with any chemical pesticide, herbicide, or fungicide.
The arrival of this novel oil on the market resulted in a sharp decline in the price
that evening primrose oil was being offered for sale. Competition between it and
Oil of Javanicus was much fiercer than had been anticipated. Although for all
intents and purposes, the fungal oil was superior to the plant oil, in that it contained
twice the GLA content of the evening primrose oil, nevertheless, and perhaps not
surprisingly, there was a certain reluctance on behalf of the general public, who
were buying these oils to switch to an oil of fungal origin. However, the marketing
of the oil carefully eschewed specific mention of the word microorganism or
fungus, but nevertheless there was a reluctance on the behalf of major purcha-
sers, the health-food stores, over-the-counter medicine shops, and so on, to pur-
chase the oil in spite of its technical superiority. Although interest in this GLA-
rich oil was high, it quickly became apparent that it was being outcompeted by eve-
ning primrose oil in terms of price. Unfortunately, within the European Union, agri-
cultural crops not designated as food crops could be financially subsidized from the
Common Agricultural Policy. This meant that growers of evening primrose oil
received cash benefits directly from the EU for growing this plant, which was desig-
nated as a non-food crop. At the same time, the fermentation process was being
financially penalized by the sugar used as the feedstock had to be bought within
the EU at EU prices and not at world prices, which were less than half this cost.
Sugar within the EU has a tariff placed on it so that farmers in the EU can receive
adequate remuneration for growing this crop. Thus, the Oil of Javanicus was doubly
disadvantaged by its commercial rival, evening subsidized primrose oil, which was
subsidized by the EU, and at the same time the cost of production was increased by
the artificially high price of the fermentation feedstock.
The final blow to the production of the fungal oil came with the introduction of
borage oil (Borago officinalis) as a superior source of GLA. This new oil had a
GLA content of 2022% (see Table 2) and, although it was technically just as dif-
ficult to grow and process as evening primrose oil, it was considered superior and
was, of course, cheaper, than both the evening primrose oil and the fungal oil.
Again, growing this borage crop still enjoyed the financial benefits that had accrued
from the EU Common Agriculture Policy for evening primrose cultivation.
In 1990, and against this background of increasing erosion of the profit for the
microbial SCO, production of Oil of Javanicus ceased. The 6 years in which the oil
had been in production, though, established a number of important points. First,
microbial oils could be produced on the very largest of scales, up to 220 m3 in
this case. The oil was intrinsically safe, posed no safety problems, and passed all
toxicity trials to which it was subjected. It was well accepted by all people who
consumed it, and no adverse reactions to it were ever recorded. There were no
particular difficulties in extracting the oil (a process using hexane was used), and
the oil could be easily purified using conventional procedures of the oils and fats
industry. The oil was also remarkably stable from oxidation, presumably because of
the presence of endogenous natural antioxidants. Thus, the way was now open for
other microbial oils to enter the market even though Oil of Javanicus was no longer
in commercial production.
In Japan, the GLA-SCO from the Idemitzu Co. Ltd process using Mort. isabel-
lina appears also to have ceased production, and GLA, when needed, is derived
either from evening primrose oil or borage oil.
2.1.2. Cocoa-Butter Equivalent Yeast Fat The initial interest in producing

high-value microbial oils in the late 1970s quickly led to serious consideration
being given to the prospects of producing a facsimile oil to cocoa butter, i.e., cocoa
butter equivalent fat or CBE (1416). The price of cocoa butter, which is used
extensively in chocolate manufacture, is very variable, but in the early 1980s, its
price was exceeding $8000 per ton. CBEs are traditionally made by palm oil frac-
tionation to give an oil containing a high content (approximately 3035%) of stearic
acid (see Table 3) with equal proportions of oleic acid and palmitic acid (17) and
command a price that is usually fixed at about 50% to 60% of the price of cocoa
butter. CBEs can be added into cocoa butter at up to 5% in various countries,
including the United Kingdom, without invalidating the claim for the product to
be called chocolate. In other countries, such mixtures can only be used in con-
fectionery chocolate.
The high price of cocoa butter at this time then made it an attractive target to
emulate. A sufficient margin of profit was considered possible if a microbial oil
could be produced that mimicked the fatty acid profile of cocoa butter and, most
importantly, had the same melting profile as cocoa butter. This manifests itself
TABLE 3. Fatty Acid Profiles of CBE-SCOsMicrobial Oils for use as a Cocoa Butter
Equivalentin Comparison with Cocoa Butter.
Relative % (w/w) of Major Fatty Acids
16:0 18:0 18:1 18:2 18:3 24:0 Ref

(n-3)
Cryptococcus curvatus WT a 30 15 45 5 0.5 2 27

C. curvatus NZ 18 24 48 3 1 2 28
C. curvatus R26-20 15 47 25 8 2 21, 23
C. curvatus R25-75 33 25 33 7 1 21, 23
C. curvatus F33.10 24 31 30 6 4 22, 23
Trichosporon cutaneum DRL-D221 30 13 47 7 24
T. cutaneum DRL-ole 26 38 16 14 24
Yeast isolate K7-2 26 25 38 6 1 1 29
Cocoa butter 23-30 32-37 30-37 2-4

a
WT wild type.
by the fat being solid at ambient temperature (up to about 25 C) but then melting
completely at 3032 C. As most microbial oils contain less than 10% of the total
fatty acids as stearic acid (18:0), the task was then to increase the content of stearate
and, simultaneously, to ensure that the resulting triacylglycerol had the correct fatty
acid distribution (i.e., was a sn-1 palmitoyl, sn-2 oleoyl, sn-3 stearoyl glycerol) so
that the ensuing oil would meet the very stringent requirement for its inclusion in
chocolate.
The most attractive production organisms appeared to be yeasts, which do not
usually contain high amounts of 18:2 or other polyunsaturated fatty acids and there-
fore were immediately attractive for this reason. They also could be grown extre-
mely rapidly and to high cell densities with high lipid contents. Attempts in the
early 1980s had been made by several groups to increase the amount of stearic
acid in a yeast by growing the yeast on stearic acid, or esters of stearic acid, as feed-
stocks (5). Not surprising, such yeasts then contained high amounts of this acid, but
as no cheap source of the stearic acid existed at this time, this did not represent an
economic route to an SCO-CBE product.
An alternative strategy was developed by Cadbury-Schweppes plc, the large
U.K.-based, multinational chocolate and food company, to inhibit the conversion
of stearic acid to oleic acid, which is mediated by fatty acid delta-9 desaturase,
using sterculic acid, cis-9,10-methyleneoctadecenoic acid (18, 19). This inhibitor,
which can be derived from sterculia and kapok oils (18), increased the stearic acid
content of several oleaginous yeasts up to nearly 50% of the total fatty acids when
they were grown on glucose. However, the selectivity of the inhibitor was such that
the yeast still contained too much linoleic acid, and consequently, another inhibitor
of the delta-12 desaturase (converting oleic acid to linoleic acid) was needed.
This was cis-12,13-methyleneoctadecenoic acid, which had to be synthesized
chemically. In the presence of both inhibitors at about 100 mg/L, one yeast,
Rhodosporidium toruloides, now produced a SCO-CBE product that was close
to the required fatty acid profile (5, 19). The costs of the fatty acid desaturase
inhibitors unfortunately proved to be too high for the process to be sustained,
and there was also clear unease at using metabolic inhibitors that might find their
way into the final product. Consequently, this interesting and novel approach was
abandoned.
Nevertheless, the clear conclusion was reached from the work with sterculic acid
that an SCO-CBE was possible if the activity of the delta-9 (and the delta-12) desa-
turase could be diminished. This was then the concept behind the subsequent muta-
tion program that was developed by several groups but most notably by a group at
the Free University of Amsterdam, The Netherlands. The yeast chosen for this
mutational work was Candida curvata (also known as Apiotrichum curvatum and
now renamed Cryptococcus curvatus). Using conventional mutational procedures,
various mutants were produced that had lost the ability to synthesize oleic acid and
now needed this fatty acid to be included in the growth medium (2023) (see also
Table 3). Clearly the gene coding for the delta-9 desaturase had been affected by the
mutations.
By judicious selection of the various mutants and adjustment of the mutational
makeup, it was possible to select mutants that had a low activity of the delta-9 desa-
turase and no longer needed oleic acid to be added as a supplement to the growth
medium. The composition of the fatty acids of two of these mutants, R26-20 and
R25-75, are shown in Table 3. The best results were, obtained though, using a
hydrid of two mutants, F33.10, which gave an almost ideal fatty acid profile for
an SCO-CBE (see Table 3). Similar mutational programs were carried out by other
research groups (2426), although with no greater success than the Dutch group
(23).
One factor that is a prerequisite for using mutated strains of micro-organisms is
their long-term genetic stability and, in particular, their stability when grown in
large-scale fermenters where they have to undergo many generations to reach the
required cell densities. Also, it is important that the mutants should grow as rapidly
as the original parent organism. During the growth of the mutants from small cul-
tures (say, 200 mL) up to the final level at perhaps 100,000 L or higher (see Figure 3),
there has to be complete genetic stability; otherwise the organism reverts to its ori-
ginal constitution and the desired product is no longer produced. For these reasons,
the mutants of the yeast used for SCO-CBE production were not entirely depend-
able when used in large-scale growth trials and yields of the required CBE were
below expectation. This is probably attributed to there being other mutational
changes to the DNA, besides the alteration in the oleic acid desaturase gene, which
then affect the long-term performance of the organism. Today, current molecular
techniques would allow one to identify the key genes that needed to be changed
and then these could be individually manipulated in such a way that the remainder
of the DNA would not be affected. Thus, in principle, specifically, genetically mod-
ified yeasts could now be produced that would yield a very high-quality SCO-CBE
even when grown in large-scale fermenters. However, in the 1980s, before this
current technology was available, further efforts to produce a yeast oil CBE focused
on controlling the production of oleic acid using a restricted supply of oxygen to
the cells.
As the conversion of stearic acid to oleic acid (and indeed all desaturase reac-
tions) requires the active participation of O2 in the reaction, it was attractive to con-
sider that the content of stearic acid in the yeast oil could be increased by restricting
the supply of air to the growing cultures. This, in fact, was achieved by Davies et al.
(28) using the yeast, C. curvatus. To achieve the necessary low levels of oxygen
within the fermenter, a 500-L system had to be employed. With smaller fermenters,
it was not feasible to restrict the oxygen supply sufficiently to cause the desired
effects. The fatty acid profile of the yeast when grown under such conditions is
shown in Table 3 and was considered to be a reasonable approximation to that
needed for the oil to be considered as a suitable CBE (27).
Extensive large-scale cultivation of various yeasts, but principally C. curvatus,
for the production of SCO-CBE was carried out by Davies et al. in New Zealand
from the early 1980s until the 1990s. The concept behind this work was the use of
cheese creamery waste, which in New Zealandthe location of Daviess labora-
torywas a major waste resource. As whey contains 4% to 5%(w/v) lactose,
this could be used as an appropriate substrate for yeast growth. The yeasts chosen
by Davies (see 29), therefore, were all lactose-users, including C. curvatus, which
had originally been isolated by Earl Hammond et al. from dairy waste-processing
areas (30). By virtue of the extensive work carried out on this process, Davies et al.
was able to calculate the likely costs of production of an SCO-CBE (27, 29).
Davies, writing in 1992 (27), calculated that the manufacturing cost of producing
1 ton of refined yeast oil would be $800$1000. This was based on a process
that would use 200,000 m3 of whey per year and that would be available at a single
location in New Zealand. These costs, though, did not include plant depreciation,
the interest payable on the capital investment needed, nor the manufacturing over-
heads. Collectively, these costs could then double the original estimate. At this time
(late 2003), cocoa butter had slumped in world prices from its all-time high of
$8000/ton in the early 1980s to $3500/ton. As a CBE could only command about
two-thirds of this price, this meant that the yeast CBE would only be valued at
$2000$2500/ton. The margin of profit, therefore, was not great enough to warrant
the large investment of capital that would be needed to establish the process.
Consequently, after huge efforts by many groups around the world, it was generally
agreed that this approach was not economically viable. As far as the present authors
are aware, no further commercial interest has been shown in developing a
yeast-based process for the production of a CBE. Even if the economics of cocoa
butter production were to change in the next decade, there are now sufficiently
good alternative systems, including the use of immobilized lipase technology, to
produce CBE material more cost-effectively than the microbial route of manufac-
ture (17, 31). It is worth pointing out, however, that the cost-calculations of Davies
for the bioengineering aspects of SCO-CBE production could form the basis for
future calculations of such processes aimed at the production of other microbial
oils.
SCOS IN CURRENT (2003) PRODUCTION 137
3. SCOS IN CURRENT (2003) PRODUCTION
As already discussed, the cost of SCO manufacture is high because of the high capi-
tal costs involved in the construction of large fermenters and associated machinery
as well as in the costs of operation. If a microbial oil is to be exploited commer-
cially, then the SCO must command a premium price (32, 33). In reality, this means
that microbial sources for oils can only be a commercial reality if the SCO pro-
duced is (1) destined for human consumption and (2) not readily available from
traditional sources, either plant or animal.
Such opportunities exist in terms of long-chain polyunsaturated fatty acid (LC-
PUFA) rich oils that have well-defined and publicly recognized benefits for human
health. Currently, >95% of the global SCO production are oils rich in two distinct
fatty acids:: arachidonic acid (ARA, 20:4 n-6) and docosahexanoic acid (DHA,
22:6 n-3). Both are being produced either directly by, or under contract, for Martek
Biosciences Corporation, MD. The production of SCOs is no longer a small-scale
operation. The blend of SCO produced by Martek as a neonate nutritional supple-
ment (for inclusion in infant formula) has established itself worldwide in the past
few years, to the extent that current production is measured in the hundreds of
tons annually. An estimate has been made that for 2003, about 560 tons of SCOs
will be produced, the majority of which will be SCOs rich in either ARA or DHA.
For 2004, this amount is predicted to double and possibly double again in 2005.
ARA and DHA are LC-PUFA, having a carbon chain of greater than 18 carbons
and are not available from plant sources. Agricultural plants usually produce
only 18 carbon (or shorter) fatty acids, and hence, the most unsaturated fatty acids
they produce are 18:3 of the n-3 or n-6 type. Although some plants, though, do pro-
duce C20 fatty acids, these are monounsaturated such as erucic acid (22:1, n-9) or
nervonic acid (24:1, n-9) and do not possess the beneficial physiological effects of
ARA and DHA. Animals (including fish) are potential sources of a multitude of
polyunsaturated fatty acids with carbon chains as long as 22 and as many as six
double bonds. However, these animal oils have potential problems of generally
low LC-PUFA content and complex fatty acid profiles, as well as acceptability pro-
blems to various sectors of the community based on religion or lifestyle (34).
Furthermore, there are increasing concerns about the presence of viruses and prions
in materials of animal origin, and with fish, there are worries concerning their long-
term availability as a cheap resource. More recently, reports of the possible accu-
mulation of toxic pollutants, including heavy metals, from the marine environment
into fish livers have added another dimension to the argument against the use of fish
oils. Fish livers, together with whole fish bodies, are, of course, the current major
source of LC-PUFAs.
SCOs have the advantage over these traditional sources for the provision of LC-
PUFA for the very reason that they are produced by fermentation. The quality and
supply of SCO can be closely controlled and guaranteed on a yearly basis (32, 33).
Such guarantees are hard to provide for plant-or animal-derived oils because of
environmental conditions and variations that are outside of the producers control.
Weather, diet (for animal oils), and environmental pollution (including the spraying
of almost all commercially grown crops with mixtures of herbicides, pesticides, and
fungicidal chemicals) have a considerable impact on the supply and quality of oils
from traditional sources. Also, the growth of micro-organisms is very rapid in com-
parison with agricultural crops and animals; usually fermentation times are 4 to 10
days. As a result, the supply of SCO can far more easily be matched to the market
requirement, avoiding oversupply or shortage problems that can be an issue with
traditional oils.
Initially, public reluctance to consume a microbe-derived oil in preference to
those from plant or animal sources was a considerable problem, contributing to
the decline of Oil of Javanicus (the first commercially produced SCO); see Sec-
tion 2.1.1. However, a mixture of more astute marketing, rebranding of SCO as
either fermented oils, designer oils, or even vegetarian oils when in com-
petition with animal oils, and the improved public acceptance of microbial
foods, as exemplified by the success of the mycoprotein (SCP) product, QuornTM,
in the United Kingdom and Europe, has helped establish PUFA-SCOs as acceptable
products.
3.1. Archidonic Acid

After the success (at least temporarily) of Oil of Javanicus, the next fatty acid to
be developed on a commercial basis was ARA. The early development occurred in
the Far East, with two Japanese companies, Lion Corp. and Suntory. Both compa-
nies developed processes for producing arachidonic acid using fungal sources. It is
interesting to note that the interest of Lion Corp. in an arachidonic acid-rich oil was
not for nutritional applications but as the basis of cosmetic creams, another area that
can withstand commodities with a premium price. This development resulted in
patents being awarded to these two companies in the late 1980s covering the pro-
duction of arachidonic acid from Mortierella alpina (35, 36).
ARA-rich SCO continues to be produced commercially in the Far East by Sun-
tory and is produced in Europe by Dutch State Mines (DSM), formerly Gist-
brocades. DSM produces ARASCOTM (an oil produced from the fungus Mortierella
alpina and containing >40% w/w ARA) under contract for Martek Biosciences
(USA) for inclusion in the neonate Formulaid1 nutritional supplement, which
is used in infant formulas and baby foods in 60 countries worldwide (including
the United States since February 2002). Currently, infant formula is, by far, the
most important market for SCO. Over 95% of global sales of SCO is made up
by a single product, Marteks Formulaid1 (Figure 4).
The astonishing success of this product is evidenced by the fact that although it
was introduced into infant formula in the United States in February 2002 (and the
formula containing this supplement commands a premium price), by the time of
writing (September 2003), the market share of infant formula containing Formu-
laid1 in the United States has reached 30%. The ARA-rich SCO produced by Sun-
tory, which is similarly used in infant formula, is mainly for export as the Japanese
market for such products is still being developed. However, the oil is available with-
in Japan as a health supplement. In China, the Wuhan Alking Bioengineering Co.
Figure 4. The production and uses of commercial SCO over the past 20 years and comparison
with current estimated production and use (Dr. David J. Kyle, personal communication). (This
figure is available in full color at http://www.mrw.interscience.wiley.com/biofp.)
Ltd., located in Wuhan City, has been producing arachidonic acid using Mort. alpi-
na since 2001 (see www.alking.com.cn). Production appears to be at the 50,000-L
level although larger fermenters may be under commission. The oil is thought to be
used in infant milk powder but also may be exported to unknown destinations.
As Mort alpina, unlike Mucor circinelloides (also known as Mucor javanicus),
did not, prior to the production of ARA-rich SCO, have any definite historical asso-
ciation to any food product, extensive toxicological studies were performed to con-
firm the safety of ARASCO (3740). This requirement for toxicological screening
is a hurdle that must be overcome for any SCO that cannot claim GRAS status by
association with a preexisting human food. Obtaining this toxicological data
ab initio can be a costly and time-consuming process.
Many micro-organisms have been examined for the ability to produce substantial
amounts of ARA, but the overwhelming consensus is that fungi of the genus
Mortierella (and in particular the subgenus of Mortierella Mortierella) are the organ-
isms of choice (41 43). Although a number of species of Mortierella have been
suggested, and even perhaps developed, as production organisms (44), the commer-
cial production of ARASCOTM is currently carried out using Mortierella alpina.
The major culture collections ATCC (14 strains), IMI (4 strains), CBS (19 strains),
and so on, contain multiple isolates of Mortierella alpina, which have been exhaus-
tively examined for their commercial viability. The most productive strain will
depend to a large degree on the process developed as strain selection, and process
development must occur in parallel. The interdependence of the strain/process com-
plicates the isolation/selection of new improved strains once a process is established.
The most productive strain in the open access culture collections appears to
be ATCC 32222 (45, 46). However, the group of Shimizu working in collaboration
with Suntory in Japan have isolated a very high ARA-producing environmental
strain that has been designated Mortierella alpina 1S-4 (32, 36, 41). This strain
has several unusual and advantageous characteristics (ease of sporulation amongst
them). Much experimental data pertaining to this strain exist in the literature (32,
47 49). However, because of the proprietary nature of this strain, little of this pub-
lished data has had the opportunity to be corroborated by other workers in the field.
Although relatively high-producing strains of Mortierella alpina are available
off the shelf from culture collections, it is inevitable that commercial production
will use proprietary improved strains. The processes used to obtain these strains
involve a certain amount of lore because of the difficulties obtaining spores of
the organism, which are a prerequisite for carrying out a successful mutation
program. Also, as increased ARA/oil production within a mutant organism confers
no easily recognizable trait in a mutant strain, considerable effort must be expended
to screen many thousands of potential mutants to isolate one new strain that over-
produces ARA.
3.1.1. Commercial Production The commercial production of ARA-rich SCO

starts with the thawing of a cryovial of a certified stock culture of the production
strain of Mort. alpina. The thawed cryovial is used to inoculate a series of shake-
flask cultures and seed fermenters of increasing volume to maintain a 510% inocu-
lum for each successive stage of the seed train. Final production scale is typically
50 to 200 m3 (see Figure 2) (50).
Media used for ARA-rich SCO production vary depending on the process/strain
used; however, they tend to be relatively simple complex media composed of a base
of yeast extract and glucose (51), although the ion composition is known to be cru-
cial for optimal productivity as well as the carbon to nitrogen ratio in the medium
(48, 52, 53).
Unlike classic fungal fermentations, such as for the production of penicillin,
citric acid, and gibberellic acid (1), the desired product in the case of ARA-rich
SCO is an intracellular product and this leads to the ARA-rich SCO process having
some distinctive features. Although in the other processes fungal biomass is, in
effect, an unwanted byproduct and can be minimized, in the ARA-rich SCO pro-
cess, a very high cell density must be achieved. Cell densities as high as 5060 g/lL
dry weight have been reported (48), and this creates problems in terms of culture
mixing and mass transfer as LC-PUFA biosynthesis is an arobic process caused by
the O2-requirement of the fatty acid desaturases.
In order to obtain maximal ARA-rich SCO productivity, the fungus must be cul-
tivated in the correct morphological form. Mort. alpina can grow as either dispersed
hyphal filaments or as pellets (of varying sizes; depending on conditions). Although
feather-like hyphal filaments yield the optimal ARA production at low cell den-
sities, because of the ready access of nutrients and ease of gas exchange between
the hyphae and the medium (54), this morphology is not suitable for cultivation at
high cell densities. Under intensive cultivation conditions, hyphal growth causes the
viscosity of the culture to increase and agitation of the culture to be difficult (48).
Under these conditions, gas exchange and nutrients become limited, thereby dele-
teriously affecting ARASCO production. For large-scale industrial ARA-rich SCO
production, the optimal morphology is pellets, which causes a decrease in the visc-
osity of the culture and promotes mixing. When pellets form, however, the culture
becomes heterogeneous as the biomass in the interior of the pellet can experience
nutrient limitation (49). Depending on the size of the pellets, the ARA-rich SCO
production when the fungus changes from hyphal to pellets can be decreased by
as much as 50% (45). Therefore, achieving the optimum morphology for ARA-
rich SCO production is not a trivial matter. It appears small pellets < 2 mm in dia-
meter are optimal to decrease culture viscosity, while maintaining adequate nutrient
and oxygen transfer to the pellet interior (48, 55).
The ARA-rich SCO fermentation operates under nitrogen-limiting conditions to
promote cell lipid accumulation. A glucose-fed batch system is employed to allow
high levels of carbon to be used without the toxic effects of very high initial glucose
concentrations. Once the biomass has accumulated a suitable amount of lipid and
ARA (approximately 200 mg ARA/g dry weight), the fungal biomass must be
removed from the culture medium, either by continuous centrifugation or filter
pressing, and dried before the oil can be extracted and purified using techniques
essentially identical to those used for vegetable oil production. The resulting oil is
a pale yellow brilliant oil with a relatively bland flavor that is remarkably resistant to
oxidation. The fatty acid profile of the commercial oil is given in Table 4 (55).
3.2. Docosahexaenoic Acid

The most unsaturated fatty acid found in significant quantities in nature (some algal
lipids contain trace amounts of fatty acids with as many as eight double bonds) is
DHA. This fatty acid is known to be especially important in the neural development
in animals; indeed, DHA alongside ARA are the predominant PUFA in neural
tissue (56). A deficiency in DHA in has been linked to impaired brain and visual
development in neonates (56).
Once the physiological role of DHA had been realized, this fatty acid became an
obvious candidate for commercial microbial production as no traditional lipid
source rich in this LC-PUFA alone was known. As mentioned above, plants gener-
ally do not synthesize fatty acids longer than 18 carbons and no plant makes any
DHA. DHA is found in animal-derived oils, particularly fish oils; however, the use
of fish oil as a DHA supplement is problematic. Fish, like all animals, possess fatty
acid profiles that, to a large extent, reflect their dietary intake rather than metabolic
ability to synthesize fatty acids. As a result, the fatty acid profiles of fish oil are
complex and contain a host of other PUFA as well as DHA. The presence of another
LC-PUFA, eicosapentaenoic acid (EPA, 20:5 n-3), is a particular problem asso-
ciated with fish oil if the intended use is neonate nutrition. EPA, a precursor of
important signal molecules (prostaglandins and eicosanoids), is thought to act
antagonistically to the beneficial effects of ARA-rich SCO supplementation of
infant formula and is associated with growth retardation in neonates (57).
TABLE 4. Fatty Acid Profiles of Microbial Oils Rich in Arachidonic Acid and Docosahexaenoic Acid that are Produced Commercially (from 55).
Fatty Acid Composition (Rel. % w/w)
12:0 14:0 16:0 16:1 18:0 18:1 18:2 18:3 18:3 20:3 20:4 22:5 22:6 24:0
Oil (n-6) (n-6) (n-3) (n-6) (n-6) (n-6) (n-3)
ARASCOa 0.4 8 0 11 14 7 4 4 49 0 1
DHASCOb 4 20 18 2 0.4 15 0.6 39
Schizo-SCOc 13 29 12 1 1 2 3 1 12 25
a
Production organism: Mortierella alpina (see 55).
b
Production organism: Crypthecodinium cohnii (see 55).
c
Production organism: Schizochytrium sp. (see 55).
NB: ARASCO and DHASCO are registered tradenames of Martek Corp. Inc.
Other problems associated with the administration of fish oil to the general popu-
lation, and in particular to neonates, is the potential for the presence of environmen-
tal toxins in fish oils. As the fish from which the oils are derived are free living in
the worlds oceans, they are prone to contamination with pesticide residues and
industrial wastes (including heavy metals) that are all too frequently released
into the marine environment. Although the levels of these chemicals are usually
low, they are of sufficient amounts to cause concern that the British Government
commissioned a report that has highlighted the extent of contamination of fish
stocks (58). In particular, the presence in fish oils of dioxins and polychlorinated
biphenyls (PCBs) was noted and the report recommended that oily fish should
not be eaten more than once a week. More recently, the specter of mercury in
fish oils has led to the recommendation by the Food and Drug Administration
(FDA) and the Environmental Protection Agency in America (59) that pregnant
women should restrict their oily fish intake. Furthermore, a recent study has deter-
mined that the mercury content of fish oil diminishes the cardiovascular benefits of
these oils (60). Any inclusion of the oils from such fish, mainly derived from liver,
an organ that often concentrates ingested toxins in infant formula, even if the levels
of contaminants is low and less than safety limits, must be seen as foolhardy. Alter-
native sources of DHA that would be outside the current supplies are clearly
needed, and such sources may then be expected to command a premium price
far in excess of that of fish oils.
3.2.1. Microbial Sources of DHA It has been recognized for over 30 years that
certain marine microorganisms have the ability to synthesize DHA de novo and to
accumulate significant amounts of this fatty acid in their cellular lipid in a relatively
pure form (61). Indeed, it is these microbes at the base of the food chain that
synthesize the DHA that eventually appears in fish oil (62). Two microalgal
sources, in particular one dinoflagellate (Crypthecodinium cohnii) and one chytrid
(Schizochytrium), have been found to synthesize sufficient quantities of DHA (in an
oil devoid of EPA) to become commercially viable. Although both organisms pro-
duce a DHA-rich SCO, important differences exist in the fatty acid profiles of the
resultant oils that have significant impact on their potential use. Although C. cohnii
oil contains DHA as >99% of the PUFA (and >40% of the total fatty acids), the oil
from Schizochytrium contains a significant amount of another PUFA docosapentae-
noic acid (DPA n-6; 22:5n-6). The presence of DPA in the chytrid oil was initially a
cause for sufficient concern that this oil was not considered suitable for neonate
nutritional applications. However, the DPA appears to be completely benign and
has, moreover, been recognized as a natural component of the phospholipids of
human blood platelets (63, 64). Studies of DPA metabolism in rat hepatocytes
(65) have further indicated that DPA is retroconverted to ARA if the ARA content
of the diet is low, but when ARA is administered along with a mixture of DPA and
DHA, then the DPA serves to maintain the DHA at a high circulating concentration.
Thus, there may be positive benefits of including DPA in a dietary supplement of
ARA DHA, and this would occur by using the Schizochytrium oil along with the
ARA-SCO.
Initially, the two organisms were developed as competing processes by two pio-
neering U.S. firms, Martek (using C. cohnii to produce DHASCOTM, an oil contain-
ing >40% DHA) and OmegaTech (using Schizochytrium to produce Sea GoldTM,
now known as DHAGoldTM, an DHA-rich SCO also containing DPA n-6). In April
2002, Martek acquired OmegaTech to combine the expertise of the two companies.
Currently, the DHA-rich SCO produced by Martek is almost exclusively DHAS-
COTM and is destined for infant formula as part of the SCO blend Formulaid1.
The production of the Schizochytrium oil continues at a low level and is aimed
more directly at food applications. Several other companies, Norferm in Norway
and Celanese in Germany, have developed competing processes to some level at
least, and indeed Celanese, through their subsidiary company Nutrinova, Frankfurt,
Germany, have recently expressed their intent to launch a chytrid-derived DHA-rich
oil, to be known as DHActiveTM. To date, however, no DHA-rich SCO other than
Marteks DHASCOTM has reached a large-scale market. In Japan, Nagase-Suntory
are thought to be producing small amounts of DHA-SCO to be used as a dietary
supplement, although it is uncertain to what extent production is taking place.
Both C. cohnii and Schizochytrium were selected for both their capacity to accu-
mulate large amounts of DHA containing oil (cells contain >30% oil containing
>40% DHA) and ease of cultivation in conventional stirred tank fermenters.
Although both C. cohnii and Schizochytrium are microalgae, they are heterotrophic
and therefore do not need light as a prerequisite for growth. As a result, both can be
grown in fermenters using similar technologies used for other microbial fermenta-
tions. One peculiarity of these organisms, however, is a consequence of their marine
heritage: Both organisms require the growth medium to contain a substantial con-
centration of salt (NaCl) in order to grow. Seawater contains 1819,000-ppm Cl,
which is not a problem in glass vessels, such as laboratory bench fermenters; how-
ever, it is a major problem at industrial scale where stainless steel vessels are the
norm. The standard grade of stainless steel, S 30400 (more normally referred to
as simply 304), which is relatively inexpensive and is used for a large percentage of
applications (approximately 50% of all stainless steel used is 304) from household
pans and cutlery to industrial plants. A total of 304 can withstand high Cl condi-
tions for short periods of time, assuming thorough washing in between, but is sus-
ceptible to crevice corrosion at above 200-ppm Cl if contact is prolonged. The
higher grade of stainless steel 316 is also used for industrial plants, although less
often because of increased expense and can withstand Cl concentration up to 1000
ppm, still insufficient for cultivation of microorganisms in media with salt concen-
trations similar to seawater. As a result, specific media have been developed (66),
and strains of both organisms that are capable of growth in low salt conditions
have been identified.
As with the ARASCO process, the organism is cultivated in sequentially larger
vessels to ensure sufficient inocula size, to final production vessels up to 200 m3
(see Figure 2). The development of the culture is followed to ensure optimal con-
ditions, and once sufficient biomass and oil have accumulated, the biomass is har-
vested by centrifugation, spray dried, and then hexane extracted using technology
essentially identical to that used for vegetable oils (50, 67). The resulting oil is a
PROSPECTS FOR PRODUCTION OF OTHER PUFAS BY MICROORGANISMS 145
transparent orange oil that is bland in taste and odor. Despite its high DHA content,
DHA-rich SCO and in particular DHASCOTM is very stable, far more stable than
fish oils (6769), to the extent that microbial oils have to be relatively badly abused
before they take on the taste/odor characteristics of even the most refined fish oils.
4. PROSPECTS FOR PRODUCTION OF OTHER PUFAS

BY MICROORGANISMS
4.1. Eicosapentaenoic Acid (20:5, n-3)

The markets for ARA and DHA are now well established in Europe and the United
Stated and are clearly being developed in Japan as well as in China. The next PUFA
that appears likely to be produced is EPA, which could be used as a nutraceutical
for over-the-counter sales or, more likely, as a possible pharmaceutical, as there
have been numerous reports of its benefits for the treatment of various diseases
and disorders. Such conditions as atherosclerosis, cancer, rheumatoid arthritis, psor-
iasis, bipolar disorder, schizophrenia, and Alzheimers disease have all been said to
have been improved or relieved by the oral administration of EPA, sometimes on its
own, although sometimes with DHA (34, 7072). EPA and DHA, of course, occur
together in many fish oils, and when treatment with both these PUFAs is advocated,
then the use of fish oils would seem to be the appropriate recommendation.
EPA has long been recognized as having anti-inflammatory properties and there-
fore has potential uses against autoimmune diseases such as arthritis. However,
inflammation is now being recognized as a major factor in the progression of heart
disease (73). Thus, the cardioprotective activity of EPA may be associated with its
anti-inflammatory properties. There is also some evidence that growth of some
mouse tumors can be decreased by oral administration of large doses of EPA
(7476). The EPA may work by affecting fatty acid uptake by the tumor cells
and, in particular, by inhibiting linoleic acid uptake, which is converted into the
13-hydroxy derivative that is a positive promoter of tumor growth (77).
Opportunities for producing EPA, besides purifying it from fish oils, which will
always leave some DHA behind, using micro-organisms are currently somewhat
limited. Mortierella alpina strain 1S-4, which is used commercially to produce ara-
chidonic acid (see above) can produce EPA if it is grown during the lipid accumu-
lation stages at 12 C rather than at the normal 28 C (78, 79). Also, EPA can be
produced in increased quantities if alpha-linolenic acid (18:3, n-3) is fed to this par-
ticular strain of M. alpina (80). Up to 42-mg EPA/g cells has been achieved. How-
ever, ARA is still present in all of these oils. Some species of the fungus, Pythium,
has also been shown to produce EPA at up to 34-mg/g cells (8183), but other
PUFAs, including arachidonic acid, are also synthesized simultaneously, making
production of an EPA-rich oil, devoid of ARA, somewhat difficult.
The best current sources of EPA would appear to be the photosynthetic micro-
algae, of which Porphyridium cruentum, Isochrysis galbana, Nannochloropsis ocu-
lata, and Phaeodactylum tricornutum appear to be the prime candidates (8487).
All produce oils with EPA between 25% and 38% of the total fatty acids (see
Table 1), but in each case, there is always some ARA or DHA, or even both, that is
present. The oil content of these algae is also not very encouraging as it is generally
less than 20% of the cell mass. The oil is not high in triacylglycerols, which is the
desirable form for it, but it is a variety of complex lipids involved in the photosyn-
thetic apparatus of the cells.
Thus, there is as yet no equivalent oil to the DHASCO from Crypthecodinium
cohnii, in which EPA is the sole PUFA that is present and moreover is present prin-
cipally as its triacylglycerol. Some researchers, though, have advocated the use of
genetically engineered microorganisms to produce EPA by taking appropriate genes
from bacteria (a few of which can also produce EPA) and cloning them into more
amenable species, but this is far from an easy task if one also needs to increase the
total oil content of the species. There is considerable effort currently being placed
into finding suitable micro-organisms for the production of EPA-rich oils, and it is
not inconceivable that within the next 2 to 3 years, such organisms will be found,
but at the present time, no microbial source of EPA is in production.
4.2. Other PUFAs

Prospects of producing a variety of other PUFAs using microorganisms exist,
although whether any of these will reach commercial fruitition will depend on
the demand for such materials. In many cases, these unusual fatty acids can be pro-
duced using Mortierella alpina, which normally produces arachidonic acid (see
above). However, Shimizu et al. have shown (32, 88) that it is possible to mutate
this fungus so that the normal sequence of desaturation and elongation of the C18
PUFAs can be disrupted at various key points. Thus, mutants of this organism have
been created, which can accumulate:
Stearidonic acid (18:4, n-3) [here the elongation of gamma-linolenic acid

(18:3, n-6) to 20:3, n-6, had been prevented by a mutation that knocked out
the gene coding for this elongase enzyme activity; the GLA was now
desaturated with an existing delta-15 desaturase]. The production of this
PUFA has, however, also been demonstrated in transgenic crop plants by
Monsanto, so the potential for this PUFA as the basis of a SCO is limited.
Dihomolinolenic acid (DHGLA; 20:3, n-6) (where the final desaturation of
this fatty acid to arachidonic acid by a delta-5 desaturase had been either
deleted by mutation or blocked by the presence of a specific inhibitor of the
desaturase).
Eicosatrienoic acid (ETA; 20:3, n-9: Mead acid) (here a double mutant of the
parent fungus was produced that lacked both delta-12 and delta-15 desaturase
activities so that oleic acid, 18:1,n-9, could only be desaturated between its
existing double bond and the carboxylate head group).
Eicosatetraenoic acid (EteA; 20:4, n-3) [where the mutant that had accu-
mulated DHGLA (see above) was now grown with linseed oil (as a source of
alpha-linolenic acid, 18:3, n-6) instead of the normal glucose; this fatty acid
was then channeled down the n-3 route of metabolism].
THE FUTURE OF MICROBIAL OILS 147
In all cases, the amounts of the various PUFA were relatively low, although for
DHGLA, the amount reached 23% of the total fatty acids. Some of the above
PUFAs also occur in small amounts in various marine microalgae, although no
routes to optimize their commercial productions have been indicated.
The unusual PUFA, docasopentaenoic acid (22:5, n-6), which occurs at about
12% of the total fatty acids of the Schizochytrium sp. used for DHA production (see
Table 4), is thought to be produced in Japan by Nagase-Suntory, possibly as a
byproduct from the purification of the DHA-rich oil that this organism produces,
although the exact means of production are not certain. The direct applications
of DPA are uncertain, although as discussed in Section 3.2.1, there may be some
benefits of including DPA along with both DHA and ARA as it may serve to main-
tain DHA at a higher circulating level (65).
5. THE FUTURE OF MICROBIAL OILS
The past two decades have seen the commercial productions of several single cell
oils: oils rich in gamma-linolenic acid, arachidonic acid, and docosahexaenoic acid.
Although gamma-linolenic acid rich-SCO was in production for only about 6 years
in the United Kingdom during the 1980s and is no longer considered an economic
reality, the production of DHA-and ARA-rich-SCOs has demonstrated the potential
of this technology, in dramatic style. These SCOs are enjoying a period of enor-
mous growth in demand, to the point where the market is supply- rather than
demand-limited. The main global supplier of SCO, Martek Bioscience Corporation
(>95% of the global market in SCO are Martek-oils) is doubling its production
capacity every year for at least the next two years (2004 and 2005). As a conse-
quence, Martek has become a major fermentation company with millions of Liters
of fermentation capacity solely dedicated to SCO production.
Against this background, it is more than likely that an SCO rich in eicosapen-
taenoic acid will also become commercially viable within the next few years as the
demands for improved supplies of this fatty acid are now beginning to accelerate.
Stearidonic acid could well be the next one after that, although in this particular
case, this PUFA can be produced using selected species of plants known as Echium,
and transgenic crop plants producing this fatty acid have been produced by Monsanto.
The microbial route to production will, however, always be expensive, and thus,
only the most expensive of the PUFAs will justify being produced by this means.
Fermentation technology is never going to be a cheap alternative to agriculture. The
logical progression, therefore, is to see the next phase of PUFA production moving
from micro-organisms toward the use of plants. This can, though, only take place
by genetic manipulation as no plants are known that produce ARA, EPA, or DHA
and it is these three PUFAs that are the major targets for production because of the
potential size (both in volume and in price) of the markets for each of them.
To design a plant that will produce these oils, it is necessary to clone into the
plant of choice (probably sunflower or rapeseedcanola) genes that will code
for one or two elongating enzymes (to convert C18 fatty acids into C20 fatty acids
and then to elongate the C20 fatty acids into C22 fatty acids) and a number of desa-
turases that will then convert a diunsaturated fatty acid (i.e., 18:2, which is the
major fatty acid of these plants) into, eventually, a hexa-unsaturated fatty acid. Thus,
to produce DHA (via a classic fatty acid route), a minimum of six different genes
will be needed and possibly more to ensure that these genes will function correctly.
The expression of the introduced genes must also be carefully controlled so that
the proteins (i.e., enzymes that they are coding for) will be specifically targeted
(both spatially and temporally) to ensure that the LC-PUFA are produced only in
seed tissue and only during the oil accumulation phase. Although this targeting is
not a major technical obstacle, the apparent innate inability of plants to metaboli-
cally process LC-PUFA could be a considerable challenge. Although the introduc-
tion of fatty acid desaturase genes (even those not usually fond in plants) leads to an
appreciable build-up in new unsaturated fatty acids, the introduction of an fatty acid
elongase has not yet engendered a significant accumulation of C20 or C22 PUFAs
(to the authors knowledge). It is possible that this failure to produce C20 PUFA is
not caused by a lack of elongase activity per se, but it is associated with the inabil-
ity of plant acyltransferases to efficiently transport C20 fatty acids. The acyltrans-
ferases are enzymes involved in the shuttling of fatty acids into cellular
membranes for their final desaturation and then incorporation into triacylglycerols
(TAG). If this is the case, then several more genes (for LC-PUFA acyltransferases)
may need to be introduced into plants to facilitate LC-PUFA accumulation, and this
would add enormously to the complexity of producing a satisfactory GM plant for
LC-PUFA production.
Some of the problems associated with the transgenic production of LC-PUFA in
plants (although not the problems associated with fatty acid transport) may be
solved by harnessing the newly discovered PKS-like route for LC-PUFA bio-
synthesis. This route, completely separate from classic fatty acid biosynthesis,
is catalysed by a single enzyme complex, and not using fatty acid desaturases or
discrete elongases appears to operate in certain marine prokaryotes and in Schizo-
chytrium (89). The PKS-like enzymatic machinery is encoded by three or four
open reading frames, decreasing the number of genes that would be required to be
transformed into a potential plant host. However, the very large size of these genes
could introduce problems of their own as it becomes increasingly more difficult to
clone genes as their size increases.
A further, and as yet unresolved, problem develops as to how the necessary redu-
cing power that is needed in each of the elongating and desaturating reactions will
be generated. The plant of choice for genetic manipulation is already in metabolic
balance: It produces what it needs, no more, no less. When an increased metabolic
burden is then placed on a plant to produce products that require additional
resources from the central metabolic pathways, it is not clear how these resources
will be achieved. An increased demand for reducing power in one part of the plant
means that metabolic economies will have to take place elsewhere. Plants do not
have the capacity to increase their metabolic capacity at the whim of genetic
engineers. They are governed by the availability of light and of CO2. Thus, how
the plants that are being designed to produce high contents of PUFAs, particularly
ACKNOWLEDGMENTS 149
DHA, will be able to achieve this is, at least to the present authors, far from clear. It
is not beyond the bounds of possibility that such genetically modified plants may
produce much less oil than the normal plant as the energy and reducing power
needed (by the desaturase and elongase reactions) to produce the PUFAs must
come from the same sources that are being used to synthesize the normal fatty
acids. The situation in plants may parallel what was found with the production of
GLA in Mucor spp. (see Section 2.2.1): You can either find strains that produce a lot
of oil but have little GLA, or you can find strains that produce a lot of GLA but have
little oil, but you cannot have both occurring simultaneously. It may then take addi-
tional genetic manipulations to correct this imbalance, but distortion of one meta-
bolic pathway can only be achieved at the expense of another so this is not likely to
be a trivial task.
Thus, although the GM plant route to PUFA formation is attractive, it is by no
means going to be as simple a task as the molecular biologists would appear to con-
sider. One can expect that it will take the labors of many people many years to
achieve these objectives, but, given the high rewards that are at stake, it may
seem to many people just a question of time before the genetic manipulators are
successful. The question then that microbial oil producers have to decide is just
how long they have before the GM plant people achieve their goals, for when
that happens, it will be the end of SCO productions. They will simply be too expen-
sive to compete. The pessimists would suggest no more than 10 years; the optimists
might suggest 20 or even 30 years.
But there is one final question that no one can yet resolve. Will GM plants be
accepted for the production of PUFAs? The public opinion of GM crops is, at
present, very much against their use in Europe. A survey conducted in the
United Kingdom during 2003 has indicated that over 80% of the population is
opposed to the introdution and use of GM crops, with only about 7% of the popula-
tion being willing to consume any such GM product (90). The view against the use
of GM plants, which has no scientific basis but is wholly based on irrational fears,
now appears to be spreading to the United States and the rest of the world. If there
should be a moratorium against the use of GM crops in general, then it may be
that the industrial companies who are currently funding much of this work will
pull out these endeavors. If this should happen, no matter how regrettable this might
be scientifically, then this would spell the end of GM plant PUFAs. If this should
happen, the only way in which the demand for these desirable oils is going to be
met will be by the microbial route. Single cell oils could then have a long and dis-
tinguished future ahead of them.
ACKNOWLEDGMENTS
We are extremely grateful to Professor Sakayu Shimizu, Kyoto University, Japan,

for providing information about the current commercialization of microbial oils in
Japan. We are also indebted to Dr. David Kyle, Advanced BioNutrition Corp., MD,
for his perceptive reading of this manuscript and for his shrewd comments.
REFERENCES
1. C. Ratledge and B. Kristiansen, Basic Biotechnology, 2nd ed., Cambridge University

Press, Cambridge, U.K., 2001.
2. M. Woodbine, Progr. Indust. Microbiol., 1, 179245 (1959).
3. C. Ratledge, in D. J. Kyle and C. Ratledge, eds., Industrial Applications of Single Cell
Oils, American Oil Chemists Society, Champaign, Illinois, 1992, pp. 115.
4. C. Ratledge, in B. S. Kamel and Y. Kakuda, eds., Technological Advances in Improved and
Alternative Sources of Lipids, Blackie, Glasgow, U.K., 1994, pp. 235291.
5. C. Ratledge, in H. J. Rehm, R. Reed, A. Puhler, P. Stadler, H. Kleinkauf, and H. von
Dohren, eds., Biotechnology, Vol. 7: Products of Secondary Metabolism, 2nd ed., VCH,
Weinheim, Germany, 1997, pp. 133197.
6. C. Ratledge, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker,
New York, 2001, pp. 351399.
7. C. Ratledge and S. J. Wilkinson, eds., Microbial Lipids, Vol. 1, Academic Press, London,
U.K., 1989.
8. C. Ratledge and J. P. Wynn, Adv. Appl. Microbiol. 51, 151 (2002).
9. Y. S. Huang and D. E. Mills, eds., g-Linolenic Acid: Metabolism and its Roles in Nutrition
and Medicine, American Oil Chemists Society, Champaign, Illinois, 1996.
10. D. F. Horrobin and P. F. Morse, Lancet, 345, 260261 (1995).
11. D. F. Horrobin, Amer. J. Clin. Nutr., 71, 367S372S (2000).
12. Y. S. Huang, and A. Ziboh, eds., g-Linolenic Acid: Recent Advances in Biotechnology &
Clinical Applications, American Oil Chemists Society, Champaign, Illinois, 2001.
13. P. M. Clough, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker,
New York, 2001, pp. 75117.
14. R. J. Davies, Report of the Department of Scientific and Industrial Research, No. IPD/T50/
2011 DSIR, New Zealand, 1983.
15. R. J. Davies, Food Technol. New Zealand, June 3337 (1984).
16. R. J. Davies, in R. S. Moreton, ed., Single Cell Oil, Longman, U.K./John Wiley, New York,
1988, pp. 99145.
17. K. W. Smith, in F. D. Gunstone, ed., Structured and Modified Lipids, Marcel Dekker,
New York, 2001, pp. 401 422.
18. R. S. Moreton, Appl. Microbiol. Biotechnol., 33, 4145 (1985).
19. R. S. Moreton, ed., Single Cell Oil, Longman, U.K./John Wiley, New York, 1988,
pp. 132.
20. A. Ykema, E. C. Verbee, H. J. J. Nijkamp, and H. Smit, Appl. Microbiol. Biotechnol., 32,
7684 (1989).
21. A. Ykema, E. C. Verbee, I.I. G. S. Verwoert, K. H. van der Linden, K. H. Nijkamp, H. J. J.
Nijkamp, and H. Smit, Appl. Microbiol. Biotechnol., 33, 176182 (1990).
22. I. I. G. S. Verwoert, A. Ykema, J. A. C. Valkenburg, E. C. Verbree, H. J. J. Nijkamp, and
H. Smit, Appl. Microbiol. Biotechnol., 32, 327333 (1989).
23. H. Smit, A. Ykema, E. C. Verbree, I. I. G. S. Verwoert, and M. M. Kater, in D. J. Kyle and
C. Ratledge, eds., Industrial Applications of Single Cell Oils, American Oil Chemists
Society, Champaign, Illinois, 1992, pp. 185195.
REFERENCES 151
24. M. Beavan, A. Kligerman, R. Droniuk, C. Drouin, B. Goldenberg, A. Effio, P. Yu,

B. Guiliany, and J. Fein, in D. J. Kyle and C. Ratledge, eds., Industrial Applications of
Single Cell Oils, American Oil Chemists Society, Champaign, Illinois, 1992, pp. 156184.
25. M. Hassan, P. J. Blanc, A. Pareilleux, and G. Goma, Biotechnol. Lett., 16, 819824 (1994).
26. M. Hassan, P. J. Blanc, A. Pareilleux, and G. Goma, World J. Microbiol. Biotechnol., 10,
534537 (1994).
27. R. J. Davies, in D. J. Kyle and C. Ratledge, eds., Industrial Applications of Single Cell
Oils, American Oil Chemists Society, Champaign, Illinois, 1992, pp. 196218.
28. R. J. Davies, J. E. Holdsworth, and S. L. Reader, Appl. Microbiol. Biotechnol., 33, 569
573 (1990) Erratum: 34, 832833 (1991).
29. R. J. Davies, and J. E. Holdsworth, Adv. Appl. Lipid Res. 1, 119159 (1992).
30. N. J. Moon, and E. G. Hammond, J. Amer. Oil Chem. Soc., 55, 683688 (1978).
New York, 2001, pp. 1135.
32. M. Certik and S. Shimizu, Appl. Microbiol. Biotechnol., 54, 224230 (2000).
33. C. Ratledge, Trends Biotechnol., 11, 278284 (1993).
35. S. Kazuhiko et al., Japanese Patent 1038007, 1989.
36. Y. Shinmen, S. Shimizu, and Y. Hideaki, European Patent 0276541, 1988.
37. H. Streekstra, J. Biotechnol., 56, 153165 (1997).
38. K. Boswell, E-K. Koskelo, L. Carl, S. Glaza, D. J. Hensen, K. D. Williams, and D. J. Kyle,
39. R. A. Hempenius, J. M. H. van Delft, M. Prinzen, and B. A. R. Lina, Food Chem. Toxicol.,
35, 573581 (1997).
40. R. A. Hempenius, B. A. R. Lina, and R. C. Haggit, Food Chem. Toxicol., 38, 127139
(2000).
41. Y. Shinmen, S. Shimizu, K. Akimoto, H. Kawashima, and H. Yamada, Appl. Microbiol.
Biotechnol., 31, 1116 (1989).
42. S. Stredanska and J. Sajbidor, Folia Microbiol., 37, 357358 (1997).
43. V. K. Eroshin, E. G. Dedyukhina, T. I. Chistyakova, V. P. Zhelifonova, C. P. Kurtzman, and
R. J. Bothast, World J. Microbiol. Biotechnol., 12, 9196 (1996).
44. T. Aki, Y. Nagahata, K. Ishihara, Y. Tanaka, T. Morinaga, K. Higashiyama, K. Akimoto,
S. Fujikawa, S. Kawamoto, S. Shigeta, K. Ono, and O. Suzuki, J. Amer. Oil Chem. Soc., 78,
599604 (2001).
45. A. Singh, and O. P. Ward, Appl. Microbiol. Biotechnol., 48, 15 (1997).
46. P. K. Bajpai, P. Bajpai, and O. Pl Ward, J. Indust. Microbiol., 8, 179186 (1991).
47. S. Shimizu, Yukagaku, 44, 804813 (1995).
48. K. Higashiyama, T. Yaguchi, K. Akimoto, S. Fujikawa, and S. Shimizu, J. Amer. Oil Chem.
Soc., 75, 18151819 (1998).
49. K. Higashiyama, K. Murakami, H. Tsujimura, N. Matsumoto, and S. Fujikawa, Biotech-
nol. Bioeng., 63, 442 448 (1999).
50. D. J. Kyle, Lipid Technol., 8, 107110 (1996).
51. H. Yamada, S. Shimizu, Y. Shinmen, H. Kawashima, and K. Akimoto, J. Despersion Sci.
Technol., 10, 561579 (1989).
52. J. Sajbidor, D. Kozelouhova, and M. Certik, Folia Microbiol., 37, 404406 (1992).
53. Y. Koike, H. J. Cai, K. Higashiyama, S. Fujikawa, and E. Y. Park, J. Biosci. Bioeng., 91,
382389, (2001).
54. E. Y. Park, Y. Koike, K. Higashiyama, S. Fujikawa, and M. Okabe, J. Biosci. Bioeng., 88,
6167, (1999).
55. L. M. Arterburn, K. D. Boswell, T. Lawlor, M. A. Cifone, H. Murle, and D. J. Kyle, Food.
Chem. Toxicol., 38, 971976 (2000).
56. G. Hornstra, Am. J. Clin. Nutr., 71, 1262S1269S (2000).
57. K. Boswell, E.-K. Koskelo, L. Carl, S. Glaza, D. J. Hensen, K. D. Williams, and D. J. Kyle,
58. BBC Online Network. (Sept. 20, 1999). Toxin warning on oily fish. BBC News. Available:
http://news.bbc.co.uk/1/hi/health/452548.sym.
59. United States Environmental Protection Agency. Press Release. (Jan. 12, 2001). Available:
http://www.epa.gov/mercury/fishadv.pdf.
60. J. Gomez-Aracena, J. D. Kark, R. A. Riemersma, J. M. Martn-Moreno, and F. J. Kok, New
Eng. J. Med., 347, 17471754 (2002).
61. G. W. Harrington, D. H. Beach, J. E. Dunham, and G. G. Holtz, J. Protozool., 17, 213219
(1970).
62. R. J. Henderson, Lipid Technol., 11, 510 (1999).
63. S. M. Innis and J. W. Hansen, Am. J. Clin. Nutr., 64, 159167 (1996).
64. J. A. Conquer and B. J. Holub, Lipids, 32, 341345 (1997).
65. P. S. Y. Tam, R. Umeda-Sawada, T. Yaguchi, K. Akimoto, Y. Kiso, and O. Igarashi, Lipids,
35, 7175 (2000).
66. W. R. Barclay, U. S. Patent 5,340,742, 1994.
67. P. W. Behrens and D. J. Kyle, J. Food Lipids, 3, 259272 (1996).
68. W. Barclay, R. Abril, P. Abril, C. Weaver, and A. Ashford, World Rev. Nutr. Diet, 83, 6176
(1998).
69. C. C. Becker and D. J. Kyle, Food Technol., 52, 6871 (1998).
70. C. A. Drevon, I. Baksaas, and H. L. Krokan, eds., Omega-3 Fatty Acids: Metabolism and
Biological Effects., Birkhauser Verlag, Basel, 1993.
71. A. P. Simopoulos, R. R. Kifer, R. E. Martin, and S. M. Barlaw, eds., Health Effects of n-3
Polyunsaturated Fatty Acids., S. Karger AG, Basel, 1991.
72. D. J. Garcia, Food Technol., 52, 4449 (1998).
73. G. J. Blake and P. M. Ridker, J. Intern. Med., 252, 283294 (2002).
74. A. S. Whitehouse and M. J. Tisdale, Biochem. Biophys. Res. Commun., 285, 598294
(2002).
75. M. J. Tisdale, J. Nutr., 129, 243S246S (1999).
76. M. J. Tisdale, Support Care Cancer, 11, 7378 (2003).
77. L. A. Sauer, R. T. Dauchy, and D. E. Blask, Cancer Res., 60, 52895295 (2000).
78. S. Shimizu, Y. Shinmen, H. Kawashima, K. Akimoto, and H. Yamada, Biochem. Biocphys.
Res. Commun., 150, 335341 (1988).
79. S. Shimizu, H. Kawashima, Y. Shinmen, K. Akimoto, and H. Yamada, J. Amer. Oil Chem.
Soc., 65, 14551459 (1988).
REFERENCES 153
80. S. Shimizu, H. Kawashima, K. Akimoto, Y. Shinmen, and H. Yamada, J. Amer. Oil Chem.
Soc., 66, 341347 (1989).
81. S. R. Gandhi, and J. D. Weete, J. Gen. Microbiol., 137, 18251830 (1991).
82. D. J. OBrien, M. J. Kurantiz, and R. Kwoczak, Appl. Microbiol. Biotechnol., 40, 211214
(1993).
83. E. E. Stinson, R. Kvoczak, and M. J. Kurantz, J. Ind. Microbiol., 8, 171178 (1991).
84. E. Molina Grima, J. A. Sanchez Perez, F. Garcia Camacho, J. M. Fernandez Sevilla, and
F. G. Acien Fernandez, Appl. Microbiol. Biotechnol., 41, 2327 (1994).
85. A. Seto, K. Kumasaka, M. Hosaka, E. Kojima, M. Kashiwakura, and T. Kato, in D. J. Kyle
and C. Ratledge, eds., Industrial Applications of Single Cell Oils, American Oil Chemists
Society, Champaign, Illinois, 1992, pp. 219234.
86. Z. Cohen, J. Amer. Oil Chem. Soc., 67, 916920 (1990).
87. W. Yongmanitchai and O. P. Ward, J. Amer. Oil Chem. Soc., 69, 584590 (1992).
88. M. Certik, E. Sakuradani, and S. Shimizu, Trends Biotechnol., 16, 500505 (1998).
89. J. G. Metz et al., Science, 293, 290293 (2001).
90. (2003). Available: www.gmpublicdebate.org/ut_09/ut_9_6.htm.
6
Transgenic Oils
Thomas A. McKeon
USDA-ARS Western Regional Research Center
Albany, California
1. INTRODUCTION
Most commodity seed oils consist of triacylglycerols containing varying per-

centages of palmitic, stearic, oleic, linoleic, and linolenic acids esterified to the
glycerol backbone. These oils and the fatty acids derived from them are used pri-
marily for food and feed; yet they have important industrial applications as well.
They provide precursors for soaps and other surfactants, derivatives used in produc-
tion of certain plastics and polyamides, applications in low volatile organic carbon
(VOC) coatings, cosmetics, as well as lubricant and grease compounds. The useful-
ness of an oil for food use lies in caloric value and the presence of essential fatty
acids, specifically, those fatty acids that are not produced by humans, linoleic acids,
and a-linolenic acids. In many cases, the usefulness of a seed oil to industry derives
from a high proportion of a specific fatty acid in the oil; for example, the high
linolenate content of linseed oil results in its good properties as a drying oil.
This article will briefly cover progress in developing new oils, examine the state
of the art, and describe industrial oilseed crops projected to be developed for
commercial use.
Oils containing high levels of oleic acid are considered to be beneficial for
human health consequences. Such oils are stable to oxidation compared with oils
155
156 TRANSGENIC OILS
containing polyunsaturated fatty acids (PUFAs) and contain low levels of saturated
fatty acids. Olive oil (70% oleate) is generally considered to have a beneficial com-
position for oleate content. It should be noted that continued research is adding
information for fatty acid requirements and benefits. Linoleic and a-linolenic acids
have been known for decades to be essential fatty acids. More recently, the need for
eicosapentaeneoic and docosahexaenoic acids as part of the diet, especially for the
developing fetus, and even g-linolenic, underlie the need for intake of oils and fats
from multiple sources that can supply what seems to be an expanding list of nutri-
titionally important fatty acids.
Although oils containing PUFAs can be converted to high monounsaturate con-
tent by partial hydrogenation, the process results in the production of trans-
fatty acids. There is a negative perception of trans-fatty acids, which are thought
to behave physiologically as saturated fatty acids. These acids are considered to
increase arterial plaque formation and may contribute to the development of type II
diabetes. Thus, a considerable research-and-development effort has been put into
designing food oils with a high content of oleic acid. However, for commercial
use, the market for food-grade oils is often driven by price, with quality traits pro-
viding premium value.
Most vegetable oil that is produced is consumed as food. These food oils also
have important industrial uses. For example, approximately 15% of soybean oil
is used for industrial products, including inks, plasticizers, coatings, and composite
materials. Other commodity oils are useful industrially because they contain un-
common fatty acids. Castor oil is 90% ricinoleate (12-hydroxy-octadec-9-enoate),
and the hydroxy group imparts dramatically different physical and chemical proper-
ties that make castor oil an important industrial feedstock. Rapeseed oil contains
up to 60% erucate (docosa-13-enoate), which is used in several lubricant appli-
cations. Tung oil contains up to 80% eleostearate (octadeca-9c,11t,13t-trienoate),
a conjugated fatty acid that makes tung oil a prized drying oil because it does
not yellow during the drying process. Palm-kernel oil and coconut oil both contain
high levels of the medium-chain saturated fatty acids laurate (C12) and myristate
(C14), which have excellent foaming properties for production of soaps and other
surfactants. Thus, several features of a vegetable oil can impart industrial chemical
value. Chemical functionality can alter physical properties or provide reactive sites
that allow useful derivatives to be made. Another industrially useful feature is the
presence of a highly enriched single component. Some oils also have unique uses as
a result of their composition; e.g., cocoa butter is unique in its melting character-
istics, which makes it an excellent component of cosmetics in addition to its food
uses. Consistent composition is also important for industry, and this is usually
closely tied to a high content of a desired component. The goal of developing oil-
seeds for industrial use is to introduce one or more of these desirable characteristics
into the oil of an agronomically suitable crop.
Seed oils also contain potentially useful fatty acids that have not been introduced
into commerce because the plant has not yet been adapted to large-scale planting.
Examples of such plants include Vernonia anthelmintica and Euphorbia lagascae,
INTRODUCTION 157
which produce oils high in vernolate (octadeca-12,13-epoxy, 9-enoate); Cuphea sp.,

many of which produce oils containing medium-chain fatty acids from caprylate
(C8) to myristate at levels up to 95% of a single fatty acid; Lesquerella sp., which
contain up to 55% lesquerolate (eicosa-14-hydroxy-9-enoate) that can replace
ricinoleate in some applications. Each crop has been the target of New Crops
research, which is aimed at breeding out undesirable agronomic characteristics
and introducing desirable traits, such as higher yield or indehiscence. Although
considerable progress has been made in each crop, the problem encountered is
a Catch 22: It is hard to get farmers to grow the crop because there is not yet
a significant market for the product, and it is hard to develop a market for the crop,
because no reliable source exists.
Breeding programs have expanded the potential uses of vegetable oils over the
years. Canola, high oleic safflower, high-oleic sunflower, and low-saturate soybean
oil are all the result of extensive traditional breeding programs, based on crossing
available crop germplasm. The introduction of mutagenesis provided a new tool
to breeders, and it is most often useful in eliminating an undesirable trait. The intro-
duction of genetic engineering greatly expanded the ability of the breeder to
introduce desired traits, with genes from incompatible crops, microorganisms,
insects, animals, or any other organism being added to the breeders toolbox.
Moreover, genomic sequencing efforts have revealed the extent of synteny among
plants. Synteny refers to a correspondence in genomic arrangement, and this has
allowed identification of specific genes associated with agronomically useful traits,
e.g., dwarfing. By comparing genomes, useful traits across species and genera can
be identified and selected for directly rather than through extensive breeding
programs.
Crop genetic engineering holds great promise as a means for developing oilseed
crops with unique characteristics that add both commercial and nutritive value,
increase utilization, and benefit the environment. Currently, the four genetically
engineered (transgenic) crops that have been adopted are all oilseed crops: soy, corn,
cotton, and canola. They are a commercial success and account for 99% of trans-
genic crops planted worldwide. Over 70% of the soy, 50% of the corn, and 70% of
the cotton grown in the United States are genetically engineered. Canola is a rela-
tively small crop in the United States, but approximately 60% of the canola planted
in the United States is transgenic. Most canola grown in Canada, a leading pro-
ducer, is transgenic. An increasing number of countries have adopted the techno-
logy. The United States, Argentina, Canada, Brazil, China, and South Africa account
for 99% of the transgenic crops produced, with an additional 12 countries adopting
the technology (1). The growth in planting of transgenic crops is remarkable in that
it has all occurred in the last 8 years, from the time the first transgenic crops were
introduced in 1996. At this time, each crop has been modified for input traits,
reducing or eliminating the need for chemical applications by the introduction of
genes encoding herbicide tolerance (soy, canola), insect resistance (corn, cotton),
or both (cotton). Currently, cotton is the only crop with a significant share of the
crop carrying both (stacked) traits. The introduction of genes for input traits is
158 TRANSGENIC OILS
TABLE 1. Composition of Transgenic Oilseeds (%).
Crop 16:0 16:1 18:0 18:1 18:2 18:3 20:0 20:1 22:0 22:1 Xa
Cotton-ctrl 27.7 0.6 2.7 15.3 43.2 0.2 0.2 2.3
Cotton-GM 26.8 0.7 2.7 15.5 45.9 0.3 0.2 1.7
Corn-ctrl 9.9 1.9 27.4 58.7 1.1 0.4 0.3 0.2
Corn-GM 9.9 1.9 27.5 58.6 1.1 0.4 0.3 0.2
Soybean-ctrl 11.2 4.1 19.7 52.5 8.0 0.4 0.2 0.5
Soybean GM 11.2 4.1 19.7 52.3 8.2 0.4 0.2 0.5
Canola-ctrl 4.6 0.3 1.6 57.5 19.4 13.8 0.6 1.4 0.3
Canola-12:0 3.3 0.2 1.1 35.1 14.6 8.8 0.3 0.7 0.2 35.5
Xa: uncommon fatty acid content; for cotton, malvalic, sterculic, and dihydrosterculic; Canola-12:0, 31.3% 12:0
and 4.2% 14:0.
Cotton control is Coker 312; Cotton GM is glyphosate tolerant, selection 1445 (Monsanto), from (2); Corn
control is GA21 segregant lacking the gene for glyphosate tolerance; GM variety is GA21 segregant carrying
glyphosate tolerance gene, from (3); Soybean control is A5403, and Soybean GM is GTS 40-3-2, derived from
particle-bombardment of A5403 with genetic material containing the CP4 ESPS gene for glyphosate tolerance
(4); Canola, control (Westar) compared to canola seed with gene for acyl-ACP thioesterase from California
Bay laurel (5).
required for maintenance of substantial equivalence in the crop. Table 1 displays

data for soybean oil, cottonseed oil, and corn oil from plants lacking or including a
gene for glyphosate tolerance, and it indicates that the introduced gene has had
essentially no effect on composition, especially given the variability that is observed
in varietal and climate-related differences in fatty acid composition.
Developments affecting output traits, the product of interest in the case of this
article being oil, have not yet achieved commercial success. In addition to oilseed
crops derived from breeding and mutagenesis, this article will describe the two
transgenic oilseeds that have been commercially introduced, briefly describing the
biochemical basis of their development and the problems faced in their commercia-
lization. That discussion will be followed by a description of other oils that have
been proposed for development through transgenic technology. Finally, the article
will discuss issues related to acceptance of transgenic crops.
2. TECHNOLOGY FOR ALTERING FATTY ACID COMPOSITION
Several approaches lead to oilseed crops with altered fatty acid composition. The
most ancient is evolution, which is a long-term, seemingly random process.
Although it is not a practical means for purposefully altering fatty acid composi-
tion, especially in a brief time span, evolution has, in fact, yielded a broad range
of oilseeds with differing characteristic fatty acid compositions. In the same species
and genera, these differences usually consist of varying percentages of the same
fatty acids. In some plant families and genera, considerable variation exists as well
in the types of fatty acid made in seed oil. These differences provide the variants
needed for successfully breeding varieties with altered fatty acid composition.
Breeding programs have successfully used available germplasm to develop major
CANOLA FROM TRADITIONAL BREEDING OF OILSEED CROPS 159
crops soy, corn, rapeseed (Canola), and sunflower that have a more desirable oil
content or fatty acid composition. Where evolution may not have provided suitable
germplasm, approaches also can be taken to alter fatty acid composition. Random
mutagenesis followed by screening and breeding has produced varieties with altered
fatty acid composition in oil (6). As the mutagenic approach is geared to eliminat-
ing genes, usually this approach has reduced levels of undesirable fatty acid compo-
nents or increased levels of a desired fatty acid. A recent innovation in this approach
is targeting induced local lesions in genomes (TILLING), which uses a
mutagenic approach but introduces high throughput screening of the M2 generation
(second-generation, mutated lines that have been self-pollinated) to identify specific
genes that have been altered or inactivated by mutagenic events (7). Plant selections
carrying these mutated genes can then be screened directly for desired characteris-
tics. The TILLING process thus moves most of the screening effort into the labo-
ratory, which considerably reduces the population that would otherwise have to be
grown in the field for later screening.
The ability to manipulate fatty acid composition in oilseeds by genetic engineering
has resulted from a combination of three approaches. Biochemical characterization
has identified most steps in fatty acid biosynthesis (8, 9). Genetic identification
and chemical characterization of fatty acid biosynthetic mutants in mutagenized
Arabidopsis thaliana has provided an extensive genetic map of fatty acid and lipid
biosynthetic steps during plant growth and development (10). Identification, char-
acterization, and cloning of enzyme activities in plants that produce nutritionally
useful fatty acids, such as g-linolenic acid, or uncommon, industrially useful fatty
acids, such as vernolic acid (12,13-epoxy oleate), have provided the additional in-
formation needed to broaden the spectrum of fatty acids available from commodity
oilseeds (11). Hundreds of other fatty acids with unusual chemical functionalities
are produced by one or more oilseed plants. A considerable amount of research has
gone into elucidating the biosynthetic process by which such fatty acids are made,
and much enzymology underlying the introduction of unsaturation, conjugated
unsaturation, hydroxyl, acetylenic, and epoxy functionality is now understood. The
enzymes that carry out each reaction are interrelated, can be interconverted by
engineering appropriate amino acid residues, and to a limited extent, can have their
specificity for chain length and positional-selectivity altered in a predictable
manner (12). The specificity of the chemistry carried out on what is essentially a
straight hydrocarbon chain is unprecedented for traditional bench chemistry, and
in the future, it may represent the development of green chemistry carried out in
plants to produce desired chemical precursors.
3. CANOLA FROM TRADITIONAL BREEDING OF OILSEED CROPS
Rapeseed has long been a source of cooking oil and has important industrial uses
such as lubricants for high-temperature applications, especially those leading to
environmental release of the lubricant; antislip agents in plastics manufacturing;
fabric softeners; and additional oleochemical applications. However, the erucic acid
160 TRANSGENIC OILS
component has been considered a potential health problem, and as a result, a

low-erucic acid rapeseed (LEAR) was desired to meet food, feed, and export needs.
An intensive breeding program was initiated in the 1950s by R.K. Downey of Agri-
culture Canada in Saskatoon (13) and B.R. Stefansson (14) of the University of
Manitoba to reduce the content of erucic acid and eliminate glucosinolates from
the seed, as these were feeding deterrents and impeded use of rapeseed meal for
animal feed. Each researcher identified lines of rapeseed with very low erucic acid,
based on crossing out Brassica napus with B. juncea. The low-erucic varieties
developed had less than 2% erucic, compared with 55% in rapeseed. The low-erucic,
low-glucosinolate double-low varieties derived from this research were renamed
Canola in 1979 by the Western Canadian Oilseed Crushers but are also known as
LEAR. Although Canola is a minor crop in the United States, it is the major oilseed
grown in Canada and Northern Europe. Although the oil is used primarily for food,
its high-oleate content (60%) makes it useful for industrial processes requiring an
oxidatively stable oil. Its high yield of oil (per hectare) has resulted in its use as a
major source of biodiesel in Europe. Other changes have been bred in Canola
selections; these include low linolenate and lower saturate content. Other high-
oleate oils have been obtained through traditional breeding in sunflower, safflower,
and corn, whereas peanut and olive oil, among others, have a naturally high oleate
content.
4. HIGH-OLEIC SUNFLOWER FROM MUTAGENESIS

OF OILSEED CROPS
When available germplasm with desired characteristics is limited, mutagenesis can

help to provide additional germplasm. Chemical and radiation mutagenesis have
been used in breeding programs to obtain suitable germplasm for generating novel
traits. Sunflower has been mutagenized with and screened for fatty acid composi-
tion (15). The normal composition of sunflower oil is high in linoleate (6075%),
but some mutagenized lines were found that had higher oleate (>80%). In back-
crosses, the high oleate trait remained stable, which indicates a nonrevertible muta-
tion had resulted in the high-oleate phenotype. Later research demonstrated that
the mutation is in the oleate desaturase gene (16), a single copy gene in sunflower.
Biochemically, this process would block conversion of oleate to linoleate and allow
oleate to accumulate in the seed (Figure 1). Mutagenesis has altered fatty acid
composition of oils derived from other crops, including soybean, cotton, flax, and
canola (15).
5. APPLICATIONS OF HIGH-OLEATE OILS
High-oleate oils are highly desirable for food use. They are stable to oxidation and
therefore good for frying and can be stored without spoilage for a longer time than
oils with high polyunsaturate content. Oleate is the prevalent fatty acid in the
APPLICATIONS OF HIGH-OLEATE OILS 161
Acetyl-ACP
CO2 2 ACP
1
Malonyl-CoA Acetyl-CoA
1. AcetylCoA
Carboxylase Malonyl-ACP 4
3
ACP
2. AcetylCoA ACP
Acetyl-ACP 5
Transacylase + Butyryl-ACP
3. MalonylCoA ACP Malonyl-ACP
Transacylase Malonyl-ACP 5
4. Condensing Enzymes,
KAS III Caproyl-ACP
Malonyl-ACP 5
KAS I
KAS II Capryloyl-ACP
7. StearoylACP
Malonyl-ACP 5
Desaturase
8. AcylACP Thioesterase
Capryl-ACP
9. Fatty AcylCoA
Synthetase Malonyl-ACP 5
10. Lysophosphatidic
Acid AcylCoA
Lauroyl-ACP
Transacylase
11. Oleoyl Desaturase Malonyl-ACP 5
Myristoyl-ACP
Malonyl-ACP 5
Linoleoyl-PC B
Palmltate Palmitoyl-ACP
Malonyl-ACP 6
11
B
Oleoyl-PC Stearate Stearoyl-ACP
10 7
9 8
Oleoyl-CoA Oleate Oleoyl-ACP
Figure 1. Fatty acid biosynthetic pathway.
Mediterranean diet based on olive oil and popularly thought to be the best fat
to consume for long-range health benefits. The oxidative stability of high-oleate oils
also meets industrial needs (17). Such oils are useful in cosmetic applications
as they are established to be safe for consumption. They are useful as sources of
162 TRANSGENIC OILS
oleate, as they can reduce or eliminate the need for purification from other fatty acid
components, which adds significant expense to the cost of obtaining pure oleate. As
they remain liquid at room temperature and below, and have high-oxidative stabilty,
they are useful in applications such as hydraulic fluids and oil-based insulators.
Although problems are associated with using seed oil in these latter applications,
the oil has the benefit of being biodegradeable and nontoxic in case of a spill (17).
6. ALTERED POLYUNSATURATE CONTENT

THROUGH MUTAGENESIS
Linseed oil from the flax plant (Linum usitatissimum L.) has a high content of
a-linolenic acid, an essential fatty acid for the human diet. Although present in
many seed oils at levels from near 0% to 15%, it makes up 55% of the fatty acid
content of linseed oil. It is this high content of the oxygen-sensitive linolenate that
imparts excellent drying oil qualities to linseed oil, which makes it useful for
production of coatings and compound materials such as linoleum. As a source of
o-3 fatty acid, it provides an essential fatty acid for the human diet and thus can
play an important role in many physiological processes. However, linseed oil is
highly susceptible to air oxidation, turning rancid and developing objectionable
odors on exposure to air. Because the meal from the flaxseed has been found to con-
tain valuable nutritive components, such as the lignans, it is desirable to have both
linseed oil and flax meal that are more stable to air oxidation. To this end, a research
group led by Allan Green developed low linolenate strains of flax by mutagenesis of
flaxseed followed by screening for low linolenate and high oil content (18). The
resulting isolates are high in linoleic acid and less susceptible to rancidity than
linseed oil. The normal composition of linseed oil is approximately 13% linoleic
and 49% linolenic, whereas the oil derived from these plants, designated Linola
or solin, is up to 70% linoleic and 2% or less linolenic (19).
7. IMPROVED OIL COMPOSITION OF A TRANSGENIC SOYBEAN
Soybean oil linolenic acid content normally ranges from 5% to 12%, with most in
the 8% range, which makes the oil susceptible to oxidation and spoilage. As a
result, soy oil is often partially hydrogenated to stabilize the oil by reducing the
linolenate content. However, the hydrogenation process introduces trans-fatty
acids, which are considered undesirable dietary components. It has been shown
that reducing the content of linolenic acid in soy oil would significantly stabilize the
oil, which makes it useful for frying and other high-temperature cooking operations
without the need for hydrogenation (20). A recent introduction, the Vistive soybean,
has been bred to contain less than 2% a-linolenic acid, thereby producing an oil that
does not need to be hydrogenated for food use. Interestingly, the low linolenate trait
was introduced by traditional breeding into a soybean line genetically engineered
IMPROVED INDUSTRIAL USE FROM GENETICALLY ENGINEERING OILSEED CROPS 163
to carry the gene for glyphosate resistance. It is thus a hybrid of traditional and
transgenic technology. Vistive will be commercialized in 2005 (21).
8. IMPROVED INDUSTRIAL USE FROM GENETICALLY

ENGINEERING OILSEED CROPS
8.1. Laurate Canola

The first commercial oilseed genetically modified for industrial use is laurate
canola, altered to produce lauric acid (22). Laurate oils produce soaps and other
surfactants because of the excellent foaming properties of the medium-chain fatty
acid, and no temperate climate crop produces laurate. Because the price of laurate
oils, derived from coconut and palm kernel, is subject to considerable fluctuation,
it was thought that development of a stable, temperate climate crop that could
produce laurate would provide a valuable renewable resource to meet a significant
domestic need. Moreover, the higher temperature melting properties of laurate oils
make them useful in baking and confections where a melting temperature similar to
butter or cocoa is desired.
In the usual fatty acid biosynthetic pathway (Figure 1), the major product is
oleate with varying amounts of palmitate and stearate produced, depending in
part on the relative activities of the acyl-ACP (acyl carrier protein) thioesterases,
stearoyl-ACP desaturase, and keto-acyl-ACP synthase II (KAS II). Certain plants
that produce uncommon fatty acids have different enzyme(s) present that result
in alterations of this pathway. Researchers at Calgene (Davis, CA) identified an
enzyme in California Bay laurel seed, which produces oil containing approximately
60% laurate. The enzyme, a lauroyl-ACP thioesterase (FATB gene), specifically
releases laurate during the course of fatty acid biosynthesis, which prevents the fatty
acid chain from being further elongated and makes the laurate available for incorpo-
ration into the triacylglycerol fraction (22). By introducing the gene for this enzyme
into Canola, and driving the expression of the gene with a promoter for napin, a
seed storage protein that is highly expressed in Brassica sp., laurate-producing
cultivars averaging 40%, and some up to 60% (on a mole basis), laurate were
obtained (Table 1, reference 23). The triacylglycerols in the oil were acylated with
laurate in the sn-1 and sn-3 positions. Because coconut oil contains laurate in all
3 sn- positions, the coconut lysophosphatidic acid acyltransferase (LPAAT) was
purified and was shown to specifically incorporate laurate into the sn-2 position.
The gene for the enzyme was cloned, expressed in laurate canola, and resulted in
cultivars with a laurate content averaging over 50% (molar basis) (24). Despite the
overwhelming scientific success of this approach, and the development of what is
logically a valuable industrial crop, the commercialization of laurate canola has not
yet been successful. The crop has reduced yields compared with normal canola
(25), requires special handling to keep it separate from other canola, has added
cost to recoup the research and development of the crop, and includes a premium
paid to contracted growers. Given these cost items and the coinciding low price of
164 TRANSGENIC OILS
laurate oils from tropical sources, laurate canola could not achieve commercial
success as a replacement for palm-kernel and coconut oils.
8.2. High-Oleate Soybean

High-oleate soybean oil, which contains over 80% oleic acid, was developed and
commercialized by the DuPont Corporation (Delaware) (26). The presence of high
levels of linoleate in a food oil is undesirable, as the presence of two methylene-
interrupted double bonds in a fatty acid makes it more sensitive to oxidation than
those high in oleate, which reduces its applicability in certain long-term uses. As
linoleate is further desaturated to a-linolenate in soy, this make the oil even more
sensitive to oxidation. Moreover, oleate is generally considered a more desirable
fatty acid for dietary intake. Linoleate is derived from the action of the oleoyl-
desaturase; thus, if the oleoyl desaturase activity could be suppressed in soybean,
the oil composition should be primarily oleate, which is ideal for food use and some
industrial uses mentioned above. Although antisense technology (in which a gene
is introduced to be transcribed in the reverse, or antisense, direction) often blocks
gene expression, the DuPont group used gene suppression, which results in
stable reduction of gene expression when the gene is inserted in the sense (same)
direction. The cultivars obtained produced oil containing up to 80% oleate, with
concomitant reduction in the amount of linoleate, some reduction in the amount
of linolenate, and little difference in levels of palmitate and stearate. However,
the expense of the seed, the availability of other oilseed crops that can also produce
high oleate, and the expense of identity preservation (IP) to keep the seeds separate
from normal soybean have inhibited commercial success for the high-oleic trans-
genic soybean as well (27).
9. FUTURE DIRECTIONS FOR TRANSGENIC OILSEEDS
Research efforts are geared to developing oils that meet changing food, feed, and
chemical feedstock needs. Hundreds of fatty acids are produced in plant sources,
and hundreds more are produced in organisms from microbes to mammals. Some
of these would be of great value if they were available in suitable amounts from a
crop source. The two general transgenic approaches used to develop such sources
are as follows:
Transgenic crops genetically modified to produce the desired fatty acid

Transgenic modification of a source plant to make it agronomically suitable
Research groups are pursuing one or both courses to enhance the value and uses of
vegetable oil for food and to expand industrial crop production and develop renew-
able resources that can replace products derived from petroleum. Although none of
these have been commercialized yet, the following examples present anticipated
new oils.
POTENTIAL NEW OILS FOR FOOD, FEED, AND INDUSTRUIAL USE 165
10. POTENTIAL NEW OILS FOR FOOD, FEED,

AND INDUSTRUIAL USE
10.1. New Polyunsaturated Fatty Acid Components

In addition to the essential fatty acids linoleate and a-linolenate, it is becoming
clear that dietary intake of other polyunsaturated fatty acids has important benefits
for proper development and health. Eicosapentaenoic acid (20:55,8,11,14,17) (EPA)
and docosahexaenoic acid (22:64,7,10,13,16,19) (DHA) are known to play an impor-
tant role in fetal neurological development (28), and they are also associated with
reduction of chronic inflammatory diseases and improved psychological mood (29).
These o-3 fatty acids are derived from fish oils in the human diet, with algae and
phytoplankton providing the original source of the fatty acids for fish. Both EPA
and DHA can be produced by humans, via successive elongation and desaturation
of a-linolenate. Linoleate is also subject to the same set of elongation reactions,
which leads to production of arachidonic acid (20:45,8,11,14) (AA). EPA and
DHA lead to formation of the o-3 eicosanoids, which tend to be anti-inflammatory,
whereas AA leads to the formation of o-6 eicosanoids, which tend to promote
inflammation (29). The two types, thus, counterbalance each other. However, the
modern diet tends to be richer in linoleate, so there is considerable interest in
expanding the availability of EPA and DHA in the diet.
Numerous biosynthetic routes to EPA and DHA exist across the spectrum of
organisms that synthesize them (29). One research group (30) has combined genes
encoding enzymes from a marine microalgae (Isochrysis galbana), from Euglena
gracilis, and from the oleaginous fungus Mortierella alpina to introduce the bio-
synthetic steps for EPA biosynthesis into Arabidopsis thaliana. The resulting triple-
transformed plant produced 3% EPA in its leaf tissue and 6.6% arachidonic acid.
This successful engineering feat can be followed up to provide seed oil containing
AA, EPA, and DHA as a more concentrated product of these fatty acids (31).
An alternative route to EPA and DHA can come from elongation and further
desaturation of stearidonic acid (18:46,9,12,15). Certain plants, including black-
berry, borage, and evening primrose, contain up to 25% of g-linolenic acid
(18:36,9,12) in their seed oil, with considerably smaller amounts of stearidonic
acid. The g-linolenate arises from the action of a 6-desaturase on linoleate. Small
amounts of o-3-desaturase present in these seeds account for the stearidonate
produced. When 6-desaturase from borage was introduced into soy, plants produ-
cing up to 29% g-linolenate (precursor to arachidonate), with up to 4% stearidonate
in oil, resulted (32). Further desaturation to stearidonate could be promoted with
high expression of an o-3-desaturase.
10.2. Oils Containing Hydroxy Fatty Acids for Industrial Use

Castor (Ricinus communis) is a model industrial crop. The seed is up to 60% oil,
which is composed of 90% ricinoleic acid (12-hydroxy oleate), a fatty acid that
produces literally hundreds of products, which include lithium grease, low VOC
166 TRANSGENIC OILS
coatings, plasticizers, Nylon 11, and cosmetics, among others. The laxative effect
of the oil proscribes use of castor as a food crop, and it seems to be a monotypic
genus. Thus, many concerns expressed for genetic modification of food crops do not
apply to castor. However, the presence of a potent allergen and the toxic protein
ricin in the seed complicate utilization of the seed meal remaining after oil extrac-
tion and prevent widespread cultivation of castor as a crop (33). Initial research
efforts were aimed at producing a castor oil substitute in an alternative crop.
The gene for the enzyme that made ricinoleate, the oleoyl-12-hydroxylase, was
cloned (34) and expressed in plants including Arabidopsis and canola (35). How-
ever, these transgenic plants never made oil containing more than 20% hydroxy
fatty acid. It became apparent that in addition to the oleoyl hydroxylase, other
enzymes involved in the biosynthetic pathway for high ricinoleate oil may also
have evolved with the pathway, and developed substrate specificities not present
in alternative crop plants. This result seem to be the case for the diacylglycerol
acyltransferase (DGAT), the terminal step in triacylglycerol biosynthesis. The
enzyme from castor displays preference for substrates containing ricinoleate
when compared with a DGAT from Arabidopsis (36). The biochemical elucidation
of castor oil biosynthesis should eventually provide the molecular tools necessary
to engineer synthesis of a high-ricinoleate oil in an agronomically suitable crop
(37, 38).
As the toxin and allergen are both proteins and have previously been identified
and cloned, it is possible to use transgenic technology to block their expression.
This approach is being pursued, and has resulted in the development of a transfor-
mation system for castor, a plant that had proven to be recalcitrant to transformation
and regeneration of intact plants (39).
10.3. Oils Containing Novel Monounsaturated Fatty Acids

The fatty acid petroselenic acid (octadeca-cis-6-enoate) is produced in Umbellifereae
plants such as coriander, with levels approaching 90% in the oil. Unlike oils with
high oleate, oils high in petroselenate are solid at room temperature and are a
precursor of adipic acid for Nylon 6,6. Although first postulated as arising from
a simple variant of the stearoyl-ACP desaturase that produces oleate, its production
in plants is more complicated. The biosynthesis of the fatty acid occurs by desatura-
tion of palmitoyl-ACP at the C-4 position, followed by elongation to petroselenate
(40). Although biochemically analogous to stearoyl desaturation, the protein
factors, such as ACP and ferredoxin, involved in the reaction appear specific for
the petroselenate pathway. It is now thought that, in general, when a plant produces
an unusual fatty acid, it may require an entire complement of additional genes to
effectively produce the fatty acid (41, 42).
Numerous fatty acids have considerable commercial potential if they can be
produced in suitable crops at a high level (42). Some of these are included in
Table 2. With perhaps the exception of medium-chain fatty acids, there is not yet
a major success in producing commercially useful levels of any uncommon fatty
acid in a transgenic crop plant.
ISSUES RELATED TO TRANSGENIC OILSEEDS 167
TABLE 2. Industrially Useful Fatty Acids for Transgenic Plant Production.
Fatty Acid (type) Source Use
Eleostearic (conjugated) Tung, Bitter Melon Drying oil

Octadeca-9c,11t,13t-trienoic
Erucic (very long chain) Rapeseed, Crambe Lubricants,
Docosa -13c-enoic Anti-slip agent
g-Linolenic Borage, Blackberry Nutraceutical
Octadeca-6c,9c,12c-trienoic
Medium chain (saturated) Cuphea, Coconut, Detergents
6 to 14 carbons Bay Laurel
Oleic Many Hydraulic oil,
Octadeca-9c-enoic Oleochemicals
Petroselenic (isomer) Coriander Nylon 6,6
Octadeca-6c-enoic
Ricinoleic (hydroxylated) Castor Lubricants,
Octadeca-9c,12-OH-enoic Polymers
Vernolic (epoxy) Vernonia, Euphorbia lagascae Coatings, plasticizer
Octadeca-9c,12,13-O-enoic
Very long chain Algae nutraceutical
polyunsaturated VLCPUFA
11. ISSUES RELATED TO TRANSGENIC OILSEEDS
One inhibiting factor in commercial development of transgenic oilseeds with novel

traits is public acceptance. The primary principle upon which approval has been
based is known as substantial equivalence, which means that aside from any
introduced changes, the composition of the plant or seed remains essentially
unchanged. However, the concept of unintended consequences expands the
scope of substantial equivalence, which establishes criteria that must be examined
and met. Satisfying the concern for unintended consequences broadened the con-
cept of substantial equivalence to include transcripts, the proteome, metabolome,
and even genome sameness (43). In the approval process for a transgenic plant,
these issues become a key part of the risk assessment both for food crops (44)
and for industrial crops (45).
Most transgenic oilseeds with altered fatty acid composition remain research
subjects, with commercial introduction limited to two crops, neither of which have
yet achieved success in the marketplace. The expected benefits from transgenic
crops with altered fatty acid composition include improved stability properties;
enhanced nutritive value; expanded use of renewable resources to replace petroleum
derived materials; replacement of chemical processes, such as epoxidation of fatty
acid double bonds; and gradual expansion of agriculture as a chemical industry, a
concept long ago known as chemurgy. It is possible to predict some issues that
168 TRANSGENIC OILS
will arise from commercialization and widespread planting of these crops. It is

noteworthy that the major commercially successful transgenic crops are all oil-
seeds, and some understanding of issues and effects of transgenic oilseeds can be
drawn from these crops.
The major transgenic crops grown in the United States and elsewhere are
soybean, cotton, maize, and canola. Considerable effort has gone into and continues
in optimizing these crops for food use. Industrial applications serve as a supplemen-
tal market for vegetable oils. The key to converting oilseeds to enhance food value
or expand their use as an industrial feedstock lies in predictable alterations of bio-
synthetic pathways that will lead to production of the desired product. In the case
of these four commercial crops, they have been engineered for the input traits,
herbicide tolerance and insect resistance. Studies presenting long-range predictions
on profitability of transgenic crops done in the 1980s, before any crops were near
commercialization, indicated that sales of seeds for transgenic crops would be the
major source of profit. Thus, several manufacturers of agricultural chemicals
acquired seed companies and developed research programs that addressed financial
and environmental concerns. Some seeds developed under these programs relied on
application of a product from the company, thus providing a secure market for the
seed and the agrochemical. At the same time, the transgenic crops required consid-
erably less pesticide or herbicide applied, which provided benefit to the farmer via
lower capital outlay, an increase in yield, and considerably lower amounts of agro-
chemicals applied to crops and released into the environment. The secure market for
herbicides also provided an economic incentive for carrying out the mandated regis-
tration of agrochemicals for each crop.
Although the benefits of transgenic crops to consumers are somewhat abstract, as
the level of agrochemical residues on crops is already very low, the benefits to farm-
ers include higher profitability as a result of reduced chemical input and reduced
toxic exposure. For example, farmers in some nations experienced a 75% reduction
in exposure to the toxic effects of agrochemicals when growing transgenic cotton
(46). Reductions in chemical exposure are clearly beneficial to the farmer, farm
workers, and wildlife.
Although oilseed crops may be engineered for industrial use, areas of concern
relate to the food supply. The introduction of allergens in the form of new proteins
is a concern, highlighted by the Starlink episode (47). As yet, no human case of an
allergic reaction related to Starlink has been identified. However, any food crop
modified to produce a toxic, noxious, or bioactive compound can present a potential
health hazard. These hazards would include oilseeds expressing ricinoleate, which
is a laxative; vernolate, which might be an irritant; and other fatty acids with
unwanted physiological effects. Such crops and components of the crop must be
isolated from the food supply by using a sound IP system.
Any process that causes comingling or cross contamination of food and non-food
crops is a concern. Cross-pollination with food crops is a particular concern, and
several strategies for preventing it have been described (48). For example, crops
producing oils for use in industry and containing a non-food fatty acid should
not comingle or cross with related food crops. Such crops would have to be grown
in limited areas and surrounded by a buffer crop to block cross-pollination (or

low-probability revertants if male-sterile) (48, 49), a particular concern for canola,
which must be buffered from HEAR when grown. An additional concern for
nonsterile crops is seed drop during harvesting, which can result in germination
and growth of the crop among the crop planted in the next growing season.
Recently, some corn seed grown to produce a pharmaceutical protein in one grow-
ing season was left in a field and germinated among a crop of soybeans. As a result,
the soybeans had to be recovered and destroyed. The end result of this incident was
increased regulatory oversight of non-food and industrial transgenic crops to pre-
vent such incidents in the future.
Industrial oilseed crops are beneficial because they are renewable resources,
yield biodegradeable products, and are environmentally benign. To the extent they
supplant products derived from petroleum, they are clearly green alternatives
to synthetic chemicals and inherently benefit human health by reducing exposure to
atmospheric and aqueous emissions from petroleum processing plants. As much
controversy about genetic modification is concerned with the safety of food derived
from genetically modified plants, genetic modification of industrial crops should
be relatively free of controversy. This situation is not the case, for several reasons.
Many byproducts of industrial crop processing enter the food supply. For example,
after oilseeds, such as industrial rapeseed, have been processed to remove the oil,
the remaining meal is protein rich and processed for use as animal feed. In some
cases, the meal may produce foods for human consumption; for example, flax meal
from linseed oil processing provides nutritional benefit in the form of lignans and
omega-3 oil residue (50). Additionally, many crops have dual uses. Soybean is pri-
marily a food crop, but soybean oil and soybean protein are also used for non-food
applications, such as inks, coatings, and adhesives. The Starlink maize incident has
made it clear that approving a food crop for strictly non-human use (animal feed) is
not sufficient to prevent it from entering the human food supply. No genuinely
harmful consequences of the Starlink corn to human health were found, only to
positive perception of the transgenic crop industry (47). The case described above
involving comingling of transgenic corn carrying a therapeutic protein underlines
the need for a sound (IP) system and appropriate quarantine (both space and time),
if food crops are to produce potentially noxious products, and if food crops are to be
planted at a later date in the same field. In the case of Starlink, with animal feed
being an inherently cheaper end-use than human food, no economic motivation
existed to maintain it separately from other maize. In the case of the transgenic
corn, the product would be much higher value than any food use but extremely dif-
ficult to prevent some seed from remaining in the harvested field.
For transgenic crop approval in the United States, the action and approval
of three Federal agencies is required. The Food and Drug Administration evalu-
ates the crop for direct and indirect food use, the Environmental Protection Agency
registers the crop for potential environmental effects, and the U.S. Department of
Agriculture, Animal and Plant Health Inspection Service (USDA-APHIS) evaluates
genetically modified crops for their potential to disrupt the growth habit of domestic
plants. Canada and Australia have similar oversight, and it is likely that this
170 TRANSGENIC OILS
multipartite regulatory process has engendered consumer confidence in the safety

of transgenic crops (44).
IP is the means by which a crop is maintained separately during storage, ship-
ping, and processing, and it has been in practice for several specialty crops that
have added value, for example, low-saturated soybean. Insofar as comingling a
transgenic industrial crop with other crops could result in health problems, it is
essential to segregate transgenic crops from other crops, in the field and postharvest.
The concept of IP and crop isolation is also important for market perception.
Currently, those persons opposed to eating food from transgenic crops require IP
of the nontransgenic crops. The estimated cost of IP is in the range of 510% of
the crops value (45). If a given transgenic crop delivers higher value to the end
user, at least an economic incentive exists to keep it from contaminating a lower
value crop.
11.1. Risk Assessment

The genetic engineering of oilseed crops to make them better suited for food and
industrial purposes necessitates an evaluation of their potential for hazard. The
primary consideration is currently whether the modification to the crop alters it
unpredictably; i.e., does the introduced gene maintain the crop as substantially
equivalent, or are unintended consequences associated with the introduced gene?
In every case, the designed crop, by definition, will produce the desired product.
Based on existing toxicological data, it is likely that either the chemical product
or related compounds will provide the means to develop a toxicological profile
and determine any potential harm arising from production. A procedure based on
close comparison of a nontransgenic control with the transgenic has been described
(51). A novel product in the plant will require an independent toxicological assess-
ment. The crop residue remaining after extraction of the product will also need to
be evaluated for the remaining product, as well as changes in the crop residue that
result from the alterations required to make the product. Changes in metabolism
brought about to enhance production of a single product can be predicted based
on knowledge of biosynthetic pathways affected by the alteration, but a broad-based
approach is required to identify unintended changes (51). Regulatory agencies
require a demonstration of substantial equivalence depending on the intended
uses of the product or the crop residue remaining (47). Finally, migration of the
transgene(s) into other crops must be evaluated from the standpoint of potential
for harm and the likelihood that it will actually occur. Knowledge of agronomic
habit allows assessment of the latter, and toxicological analysis provides the needed
information for the former. Any potentially toxic or noxious product can be
restricted to sterile strains. Although pollen release from transgenic oilseeds,
such as the Brassica sp., is a common scenario for concern, a more significant
problem may arise from comingling as a result of seed drop during harvest. It
is inevitable that crops that are not controlled to prevent gene release will
ultimately not be permitted (52). Transgenic technology holds out great promise
for expanding the use of renewable resources in production of industrial products.
Accordingly, it is essential that such a benefit be implemented in a way to prevent

any harmful effects.
11.2. Allergenicity
It is expected that beyond the different product in an engineered transgenic oil-
seed, the crop will retain substantial equivalence to the nontransgenic crop. The
genes for given characteristics have been cloned and sequenced; perhaps quantities
of the protein have been isolated after being expressed in bacteria or yeast. In many
instances, the activity of the native protein has only been demonstrated by the
change brought about in planta. Therefore, its potential for becoming a problem
allergen remains unknown. In the case of oils, where the primary product is free
of the transgenic crop protein, the allergenicity of the protein is not a food health
issue. However, because most oilseed meals are used as food or animal feed, then
protein allergenicity clearly becomes a consideration. In plants that have undergone
metabolic engineering, the introduced gene(s) is (are) often overexpressed to
redirect the flow of metabolites to the desired product, which leads to a relatively
high level in the plant (23). Altered timing for expression may also be implemented.
Promoter technology is still in a relatively primitive state. In oilseeds, it is common
to use promoters that drive the synthesis of storage proteins and to restrict expres-
sion of the introduced gene to the seed (23). As storage protein promoters are
geared to produce a high level of protein, high levels of the expressed protein
can accumulate, and as storage proteins are expressed late into seed development,
proteins produced to alter metabolism of the immature seed may persist to a high
degree in the harvested seed, whereas they would not normally be present. Methods
for demonstrating potential allergenicity exist, for example, model pepsin digestion
reactions (47, 51) and a decision-tree for assessing allergenicity have been
described, with linear epitope analysis and partial sequence identity to allergens
as indicators (51). Animal models for allergenicity have also been proposed to sup-
plement the decision-tree. In cases where the protein is not available in isolation,
theoretical models to predict allergenicity from the protein sequence are essential to
ensure the safety of associated byproducts, such as seed meal.
11.3. Pollen Transfer

Ecological concern exists about transmission of pollen from some types of plant,
such as the Brassicas and tree crops, either into weedy relatives or into crops grown
at some distance. This problem is not limited to transgenic plants. Canola, a rape-
seed cultivar bred to produce low glucosinolates and low erucic acid, must be
planted in isolation from industrial rapeseed, as each crop will result in seed with
altered composition from the ideal: that is, excessive erucic acid in the canola
and less erucic acid in the industrial rapeseed. Because the products of industrial
crops are not intended for consumption, and may even be noxious, risk manage-
ment and containment, including the prevention of intercrop cross-pollination, is
172 TRANSGENIC OILS
essential. The approaches described (53) can address the problem of out-crossing
from transgenic crops.
11.4. Economics
A major argument for promoting transgenic technology has been the need to pro-
vide more food. Some industrial applications use surplus products, for example,
soybean oil, and provide a buffer against surplus crops, to prevent a decline in
farm profitability. However, if industrial transgenic farming expands, it is not clear
to what extent agriculture may be diverted away from food production, which will
result in increased food costs, if industrial crops are grown in higher volume and
have a higher value than food crops.
If the success of transgenic industrial oilseeds is to be measured on the basis of
their commercial success (see (54) for an economic analysis of genetically modified
industrial crop profitability), then the success of such crops can affect the prosperity
of the industries they replace, such as chemical manufacturing. Although the over-
all benefit will be great, as renewable resources replace potentially limited resour-
ces, industries and workers may be displaced.
Several developing nations produce specialty crops that meet current industrial
needs, such as castor oil, rubber, and lauric acid, all of which are products from
Southeast Asia. If these products were to be replaced by transgenic crops grown
in temperate regions, economic displacement of the less wealthy countries could
occur. However, because transgenic products may also be produced more cheaply
in these countries, they should ultimately benefit from the same new technologies.
Where new uses, such as biodiesel fuel and fuel additives produced from castor oil
and laurate (45), will greatly expand demand, economic disruptions may be offset.
Many countries already have research programs in transgenic crop technology. The
impact of transgenic technologies in industrial agriculture on the world economy
remains to be seen.
Transgenic technology remains a powerful tool for developing a broad range of
useful food and industrial oils. To date, attempts to use this technology have been
limited to crop input traits, but in the long term, novel crops with altered output
traits will fill important niches in the food supply and will help to shift the petro-
leum economy to renewable resources.
REFERENCES
1. C. James. Global Status of Commercialized Transgenic Crops: 2003 Executive Summary

International Service for the Acquisition of Agri-Biotech Applications. (2004). Available:
http://www.isaaa.org.
2. D. L. Nida, S. Patzer, P. Harvey, R. Stipanovic, R. Wood and R. L. Fuchs, J. Agric. Food
Chem., 44, 19671974 (1996).
3. R. S. Sidhu, B. G. Hammond, R. L. Fuchs, J. N. Mutz, L. R. Holden, B. George, and
T. Olson, J. Agric. Food Chem., 48, 23052312 (2000).
REFERENCES 173
4. S. R. Padgette, N. B. Taylor, D. L. Nida, M. R. Bailey, J. MacDonald, L. R. Holden, and

R. L. Fuchs, J. Nutr., 126, 702716 (1995).
5. W. E. Neff, T. L. Mounts, and W. M. Rinsch, Lebensrm. Wiss. u.-Technol., 30, 793799
(1997).
6. P. F. Knowles, in G. Robbelen, R. K. Downey and A. Ashri, eds., Oil Crops of the World,
McGraw-Hill, New York, 1985, pp. 260282.
7. S. Henikoff and L. Comai, Annu Rev. Plant Biol., 54, 375401 (2003).
8. J. L. Harwood, Biochim. Biophys. Acta, 1301, 756 (1996).
9. T. Voelker and A. J. Kinney, Annu. Rev. Plant Physiol. Mol. Biol., 52, 3353361 (2001).
10. J. Browse and C. Somerville, Annu. Plant Physiol. Plant Mol. Biol., 42, 467506
(1991).
11. J. J. Thelen and J. B. Ohlrogge, Metabol. Eng., 4, 1221 (2002).
12. J. Jaworski and E. B. Cahoon, Curr. Opin. Plant Biol., 6, 178184 (2003).
13. R. K. Downey, Can. J. Plant Sci., 44, 499504 (1964).
14. B. R. Stefansson, F. W. Hougen, and R. K. Downey, Can. J. Plant Sci., 41, 218219 (1961).
15. L. Velasco and J. M. Fernandez-Martinez, J. Crop Prod., 5, 309344 (2002).
16. S. Lacombe and A. Berville, Mol. Breeding, 8, 129137 (2001).
17. S. Erhan and S. Asadauskas, Ind. Crops Prod., 11, 277282 (2000).
18. A. G. Green, Can. J. Plant Sci., 66, 499503 (1986).
19. A. Green, Aus. New Crops Newsl. 3(1). (1995). Available: http://www.newcrops.uq.edu.
au/newslett/ncnl3-92.htm.
20. T. L. Mounts, K. Warner, G. R. List, W. E. Neff and R. F. Wilson, JAOCS, 71, 495499
(1994).
21. Monsanto Press Release. (2004). Available: http://www.monsanto.com/monsanto/layout/
media/04/0901-04.asp.
22. T. A. Voelker, A. C. Worrell, L. Anderson, J. Bleibaum, C. Fan, D. J. Hawkins, S. E. Radke,
and H. M. Davies, Science, 257, 7274 (1992).
23. T. A. Voelker, T. R. Hayes, A. M. Cranmer, J. C. Turner, and H. M. Davies, Plant J., 9,
229241 (1996).
24. D. S. Knutzon, T. R. Hayes, A. Wyrick, H. Xiong, H. M. Davies, and T. A. Voelker, Plant
Physiol., 120, 739746 (1999).
25. A. E. McCoy, M. Pippin, and R. D. MacNeill IV. Laurate Canola Variety Trial. (1998).
Available: http://www.griffin.peachnet.edu/swvt/1998/rp981rat.htm.
26. A. J. Kinney, in J. P. Williams, M. U. Khan and N. W. Lem, eds., Physiology, Biochemistry
and Molecular Biology of Plant Lipids, Kluwer Academic, Dordrecht, the Netherlands,
1997, pp. 298300.
27. R. Frerichs. (2003). Available: http://web.aces.uiuc.edu/value/factsheets/soy/fact-oleic-
soy.htm.
28. A. Rodriguez, D. Raederstorff, P. Sarda, C. Lauret, F. Mendy, and B. Descomps, Eur. J.
Clin. Nutr., 57, 727734 (2003).
29. O. V. Sayanova and J. A. Napier, Phytochemistry, 65, 147158 (2004).
30. B. Qi, T. Fraser, S. Mugford, G. Dobson, O. Sayanova, J. Butler, J. A. Napier, A. K.
Stobart, and C. M. Lazarus, Nature Biotech., 22, 739745 (2004).
31. A. G. Green, Nature Biotech., 22, 680682 (2004).
174 TRANSGENIC OILS
32. S. Sato, A. Xing, X. Ye, B. Schweiger, A. Kinney, G. Graef, and T. Clemente, Crop Sci.,
44, 646652 (2004).
33. D. L. Auld, R. D. Rolfe, and T. A. McKeon,. J. New Seeds, 3, 6169 (2001).
34. F. J. van de Loo, P. Broun, S. Turner, and C. Somerville, Proc. Nat. Acad. Sci. (USA), 92,
67436747 (1995).
35. P. Broun and C. Somerville, Plant Physiol., 113, 933942 (1997).
36. X. He, C. Turner, G. Q. Chen, J. T. Lin, and McKeon, Lipids, 39, 311318 (2004).
37. T. A. McKeon and J. T. Lin, in T. M. Kuo and H. W. Gardner, eds., Lipid Biotechnology,
Marcel Dekker, New York, 2001, pp. 129139.
38. M. Smith, H. Moon, G. Chowrira, and L. Kunst, Planta, 217, 507516 (2003).
39. T. A. McKeon and G. Q. Chen, U. S. Patent 6,620,986 B1, 2003.
40. E. B. Cahoon and J. B. Ohlrogge, Plant Physiol., 104, 827837 (1994).
41. M. C. Suh, D. J. Schultz, and J. B. Ohlrogge, Planta, 215, 584595 (2002).
42. H. Drexler, P. Spiekermann, A. Meyer, F. Domergue, T. Zank, P. Sperling, A. Abbadi, and
E. Heinz, J. Plant Physiol., 160, 779802 (2003).
43. F. Cellini, A. Chesson, I. Colquhoun, A. Constable, H. V. Davies, K. H. Engel, A. M. R.
Gatehouse, S. Karenlampi, E. J. Kok, J.-J. Leguay, S. Lehesranta, H. P. J. M. Noteborn,
J. Pedersen, and M. Smith, Food Chem. Toxicol., 42, 10891125 (2004).
44. J. Thomson, Trends Food Sci. Technol., 14, 210228 (2003).
45. T. A. McKeon, Trends Food Sci. Technol., 14, 229241 (2003).
46. J. Huang, S. Rozelle, C. Pray, and Q. Wang, Science, 295, 674677 (2002).
47. H. A. Kuiper, G. A. Kleter, H. P. J. M. Noteborn, and E. J. Kok, Plant J., 27, 503528
(2001).
48. H. Daniell, S. J. Streatfield, and K. Wycoff, Trends Plant Sci., 6, 219226 (2001).
49. P. J. Dale, B. Clarke, and E. M. G. Fonte, Nature Biotech., 20, 567574 (2002).
50. B. D. Oomah, J. Sci. Food Agric., 81, 883888 (2001).
51. A. Konig, A. Cockburn, R. W. R. Crevel, E. Debruyne, R. Grafstroem, U. Hammerling,
I. Kimber, I. Knudsen, H. A. Kuiper, A. A. C. M. Peijnenburg, A. H. Penninks, M. Poulsen,
M. Schauzu, and J. M. Wal, Food Chem. Toxicol., 42, 10471088 (2004).
52. S. Smyth, G. G. Khachatourians, and P. W. Phillips, Nature Biotech., 20, 537541 (2002).
53. G. van den Eede, H. Aarts, H.-J. Buhk, G. Corthier, H. J. Flint, W. Hammes, B. Jacobsen,
T. Midtvedt, J. van der Vossen, A. von Wright, W. Wackernagel, and A. Wilcks, Food
Chem. Toxicol., 42, 11271156 (2004).
54. B. Hitz, Curr. Opin. Plant Biol., 2, 135138 (1999).
7
Tree Nut Oils
Fereidoon Shahidi and Homan Miraliakbari
1. INTRODUCTION
Tree nut oils are primarily composed of triacylglycerols, but also contain diacylgly-
cerols, monoacylglycerols, free fatty acids, and other minor components, including
natural antioxidants and fat-soluble vitamins. The chemical composition of edible
fats and oils largely determines their stability, quality, nutritional value, sensory
properties, and potential health effects. Tree nuts, in many cases, provide rich
sources of food lipids; up to 75% lipid on a weight basis (1). With a few exceptions,
tree nut lipids exist as oils at room temperature. Generally, tree nuts are rich in
monounsaturated fatty acids, predominantly oleic acid, but contain much lower
amounts of polyunsaturated fatty acids, predominantly linoleic acid and small
amounts of saturated lipids (1). In many parts of the world, such as the Middle
East and Asia, tree nuts are cultivated for use as oil crops and are important sources
of energy and essential dietary nutrients as well as phytochemicals (2). Tree nut oils
are also used as components of some skin moisturizers and cosmetic products (3).
Tree nuts, tree nut oils, and tree nut byproducts (defatted meals and hulls) are
known to contain several bioactive and health-promoting components. Epidemio-
logical evidence indicates that the consumption of tree nuts may exert several car-
dioprotective effects, which are speculated to derive from their lipid component that
includes unsaturated fatty acids, phytosterols, and tocols (4). Recent investigations
175
176 TREE NUT OILS
have also shown that dietary consumption of tree nut oils may exert even more ben-
eficial effects than consumption of whole tree nuts, possibly due to the replacement
of dietary carbohydrate with unsaturated lipids or other components present in the
oil extracts (5). Tree nut byproducts are used as sources of dietary protein and as
health-promoting phytochemicals such as natural antioxidants. This chapter sum-
marizes the characteristics and potential health effects of several tree nut oils and
their byproducts, including almond oil, hazelnut oil, pecan oil, walnut oil, pistachio
oil, Brazil nut oil, pine nut oil, and macadamia nut oil, among others. Protein
compositions of tree nut byproducts are also discussed collectively at the end of
this chapter, with emphasis on the completeness of these proteins based on their
amino acid compositions.
2. ALMOND
The almond tree (Prunus delcis and Prunus amara) and its fruit (containing the
almond kernel or almond) have long been recognized as commercially valuable
and nutritionally important. California and Italy are the major almond-producing
regions of the world, however, other parts of Europe, Asia, and Australia also con-
tribute to a lower level of production (6). The only other economically important
product of almond trees is the almond hull, which is traditionally used in animal
feed preparations. Several studies have reported that almond consumption may
improve blood lipid profiles by lowering low-density lipoprotein (LDL) cholesterol
and raising plasma high-density lipoprotein (HDL) cholesterol levels. Thus, there is
much current interest in almond oil as a health-promoting edible oil (7).
The proximate composition of almond includes 50.6% lipid, 21.3% protein,
19.7% carbohydrate, 5.3% water, and 3.1% ash (w/w) (1). The most common meth-
od for producing almond oil is hexane extraction that affords high oil yields, how-
ever, cold pressing is another commercially used procedure for almond oil
production (8). Shi et al. (8) assessed the fatty acid composition of almond oil;
oleic acid was major fatty acid present (68%), followed by linoleic acid (25%),
palmitic acid (4.7%), and small amounts (<2.3%) of palmitoleic, stearic, and ara-
chidic acids (Table 1). Almond oil is also a rich source of a-tocopherol (around
390 mg/kg) and contains trace amounts of other tocopherol isomers as well as phyl-
loquinone (70 mg/kg) (1). Almond oil contains 2.6 g/kg phytosterols, mainly b-
sitosterol, with trace amounts of stigmasterol and campesterol (1).
Sattar et al. (9) examined peroxide formation during light-induced oxidation of
several tree nut oils, including almond oil, pine nut oil, and walnut oil. The oils
were oxidized for a 7-week period under four different conditions: (1) by direct
exposure to light (540 lux), (2) exposure to light in clear glass containers, (3) expo-
sure to light in amber-colored glass containers, and (4) unexposed oils, which were
used as controls. The initial peroxide value (PV) of the almond oil was 2.8 meq
oxygen/kg oil, which was second only to pine nut oil (9). Results for almond oil
peroxidation rate under each condition were expressed as increase in PV/day
(PV/day); the oxidation rate was highest in almond oil directly exposed to light
ALMOND 177
TABLE 1. Fatty Acid Composition

of Almond Oil.a
Fatty Acid (%)
16:0 4.7
16:1 <1
18:0 <1
18:1 (n-9) 68.0
18:2 (n-6) 25.0
18:3 (n-3) <1
24:0 <1
a
Adapted from Shi et al. (8).
(0.82 PV/day), followed by almond oil stored in glass containers (0.43 PV/day),
then almond oil stored in amber-colored containers (0.15 PV/day), and lowest
in unexposed almond oil (0.11 PV/day). Under all four oxidative conditions,
almond oils showed greater oxidative stability than pine nut oil and walnut oil, pos-
sibly due to a higher content of tocopherols in almond oil (9). Salvo et al. (10) stu-
died the peroxidation rate and compositional changes of almond oil over a 3-year
period at 4 C and ambient temperature. The almond oils used were extracted from
sweet (P. delcis) and bitter almonds (P. amara). The initial composition of the two
almond oils were similar, having identical fatty acid compositions and total toco-
pherol contents; however, sweet almond oil contained only a-tocopherol (458 mg/kg),
whereas bitter almond oil contained 345 mg/kg a-tocopherol and 113 mg/kg
g-tocopherol. No changes in fatty acid composition were observed during the 3-
year storage period; however, the total tocopherol content fell to 245 mg/kg in
sweet almond oil and 121 mg/kg in bitter almond oil after one year, and became
totally depleted after three years when stored at 4 C (10). Both sweet and bitter
almond oils showed similar peroxide formation rates; the initial PV was 9.6 meq
oxygen/kg oil and rose to 21.3 after 1 year, 29.6 after 2 years, and 129.5 after 3
years of storage at 4 C. A similar, but faster, trend was observed in almond oils
stored at ambient temperatures. Thus, oxidative stability of almond oil (9, 10)
depended on the presence of tocopherols and possibly other substances contributing
to the stabilization of the oil.
The compositional characteristics of almond oil show that it is rich in several
health-promoting nutrients, many of which may be responsible for the observed
beneficial effects of dietary almond consumption in cardiovascular diseases (11)
and in weight management (12), however, few investigations have explored this
topic. Hyson et al. (13) conducted a dietary intervention study to determine whether
the consumption of whole almonds or almond oil for 6 weeks would result in simi-
lar or different effects on plasma lipids and ex vivo LDL oxidation. Both groups
consumed diets with identical almond oil and total fat levels. This study showed
that both whole almond and almond oil consumption caused similar reductions
in plasma cholesterol and LDL (4% and 6%, respectively) as well as a 14%
178 TREE NUT OILS
TABLE 2. Phenolic Acid Constituents (mg/g) of Selected Tree Nut Meals.a
Phenolic Acid Almond Hazelnut Chestnutb Pine Nut Walnut
p-Hydroxybenzoic 0.30 <0.01 0 0.44 0.16 0.06

Phenyl acetic 0 0 <0.01 <0.01 00.02
Vanillic 0.07 <0.01 0 0.14 0.09
Proto-catechuric 0.70 0.36 0.3 0.46 0.28 0.02
Syringic 0 0 0 0.06 0.23 0.02
Gallic <0.01 <0.01 0.7 4.21 0.08 0.02
Caffeic 0 <0.01 0.16 0.30 0.48 0.10
Ferulic 0 0 0 0.01 <0.01 <0.01
Total 1.08 0.36 1.65 4.97 1.37 0.51
a
Adapted from Senter et al. (14).
b
Range of values accounts for three chestnut varieties (American, Chinese, and Hybrid chestnut).
decrease in fasting plasma triacylglycerols. These findings indicate that the lipid
component of almond is responsible for its cardioprotective effects and warrants
further investigation (13).
The defatted meals and hulls of almonds contain several antioxidative com-
pounds as well as other health-promoting substances. Senter et al. (14) performed
a comparative analysis of phenolic acids in selected tree nut meals including
almond. The results of this study showed that gallic acid was the predominant phe-
nolic compound in all tree nut meals except pine nut (caffeic acid), almond, and
hazelnut (protocatechuric acid). Other phenolic compounds identified included
p-hydroxybenzoic, p-hydroxyphenylacetic, vanillic, syringic, and ferulic acids
(Table 2) (14). The antiradical activity of ethanolic extracts of almond and almond
byproducts, including brown skins and hulls, have been reported (15). The Trolox
equivalent antioxidant activity of brown skins and hulls were 13 and 10 times great-
er than that of the whole almond extracts. At a concentration of 200 ppm, ethanolic
extracts of almond skins and hulls had strong scavenging activities against super-
oxide radical (95% and 99%, respectively), hydrogen peroxide (91%), hydroxyl
radical (100% and 56%, respectively), and 2,2-diphenyl-1-picrylhydrazyl (DPPH)
(100%) (15). Sang et al. (16) isolated nine phenolic compounds from almond skins
and assessed DPPH scavenging activity of each compound; catechin and protoca-
techuic acid had the greatest antioxidant activity, followed by 30 -O-methylquercetin
3-O-b-D-galactopyranoside, then 30 -O-methylquercetin 3-O-b-D-glucopyranoside,
and 30 -O-methylquercetin 3-O-a-L-rhamnopyranosyl-(1 !6)-b-D-glucopyranoside
as well as vanillic and p-hydroxybenzoic acids, naringenin 7-O-b-D-glucopyrano-
side, and, finally, kaempferol 3-O-a-L-rhamnopyranosyl-(1 ! 6)-b-D-glucopyr-
anoside (16). Frison-Norrie and Sporns (17) quantitatively assessed the flavonol
glycoside composition of blanched almond skins using matrix-assisted laser deso-
rption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), showing the
presence of isorhamnetin rutinoside (51 mg/g), isorhamnetin glucoside (18 mg/g),
kaempferol rutinoside (18 mg/g), and kaempferol glucoside (6 mg/g). More recently,
Pinelo et al. (18) examined the total phenolics content and DPPH scavenging
HAZELNUT 179
HO
O O
OH
NH
OH
OH
O
OH
Figure 1. 1-O-b-D-Glucopyranosyl-(2S,3R,4E,8Z)-2-[(2R)-2-hydroxyhexadecanoylamino]-4,8-
octadecadiene-1,3-diol.
activity of almond hull ethanolic extracts and reported values of 3.74 mg/g and
58%, respectively. Sang et al. (19) have also isolated potential health-promoting
sterols, nucleotides, and one sphingolipid, namely 1-O-b-D-glucopyranosyl-
(2S,3R,4E,8Z)-2-[(2R)-2-hydroxyhexadecanoylamino]-4,8-octadecadiene-1,3-diol
(Figure 1), from defatted almond meals. In light of this data showing that tree nuts,
tree nut oils, and tree nut byproducts contain heath-promoting phytochemicals,
Davis and Iwashi (20) examined the effects of dietary consumption of whole
almonds, almond oil, and almond meal on aberrant crypt foci development in a
rat model of colon carcinogenesis. This landmark study showed that both almond
oil and almond meal reduced aberrant crypt foci development, but whole almonds
showed a significantly stronger anticancer effect in this model, implying a synergis-
tic anticancer activity between the lipidic and nonlipidic constituents of almonds
(20).
3. HAZELNUT
Hazelnuts or filberts (Corylus sp.) are a rich source of energy with a 6163% lipid
content (w/w) (1, 21). Other components of hazelnuts are carbohydrate (15.3%),
protein (13.0%), water (5.4%), and ash (3.6%) (1). Turkey is the worlds largest
producer of hazelnuts, accounting for approximately 75% of total hazelnut produc-
tion, followed by Italy, which accounts for 10% of total global production. In the
United States, the state of Oregon is the largest producer and in Canada, southwes-
tern British Columbia produces a small amount of hazelnuts. North America
contributes less than 5% to the total world hazelnut production, which is about
850,000 metric tons (unshelled basis) (22).
The fatty acid composition of hazelnut oil is as follows: 7883% oleic acid,
910% linoleic acid, 45% palmitic acid, and 23% stearic acid as well as other
minor fatty acids (Table 3) (1, 22). Parcerisa et al. (23) examined lipid class com-
position of hazelnut oil, showing that triacylglycerols constituted 98.4% of total
lipids, glycolipids comprised 1.4% of total lipids, and trace amounts (<0.2%) of
phosphatidylcholine and phosphatidylinositol were also present. Hazelnut oil
contains 1.21.14 g/kg of phytosterols primarily in the form of b-sitosterol and is
a very good source of a-tocopherol (382472 mg/kg) (1, 22). The main odorant in
180 TREE NUT OILS
TABLE 3. Fatty Acid Composition of Hazel-

nut Oil.a
Fatty Acid (%)
14:0 0.03
15:0 0.02
16:0 4.85
16:1 0.16
17:0 0.04
17:1 0.07
18:0 2.73
18:1 (n-3) 82.7
18:2 (n-6) 8.89
18:3 (n-3) 0.10
20:0 0.14
20:1 (n-9) 0.16
24:0 0.01
24:1 (n-9) 0.02
a
Adapted from Alasavar et al. (22).
hazelnut oil responsible for its characteristic flavor is 5-methyl- (E)-2-hepen-4-one

or filbertone, which can produce intense hazelnut oil-like aroma at the very low
odor threshold of 5-ng/kg oil (24). The oil from unroasted hazelnuts typically con-
tains about 6-mg filbertone/kg oil, whereas the oil from roasted hazelnuts contains
over 315-mg filbertone/kg oil (24). The level of filbertone in hazelnut oil serves as a
useful index for assessing possible adulteration of other oils with hazelnut oil and
the extent of adulteration (25). Bernardo-Gil et al. (26) studied the composition and
oxidative characteristics of hazelnut oils obtained by hexane and supercritical
carbon dioxide fluid extraction. Under optimal supercritical fluid extraction condi-
tions (CO2 density: 880 kg/m3, superficial velocity: 0.000642 m/s, time: 240 min,
temperature range: 308321 K, pressure range: 1823.4 MPa), the total yield and
fatty acid compositions of the resultant hazelnut oils did not differ significantly
for the two extraction methods; however, the supercritical fluid extract was clearer
than the hexane extract, implying some degree of purification by the supercritical
fluid (SCF) extraction process (26). Interestingly, the SCF extract contained 17%
more tocopherols (458.7 mg/kg) than the hexane extract (382.8 mg/kg). The oxida-
tive stability of both hazelnut oils were assessed using the Rancimat apparatus at
110 C; the SCF extract was more resistant to oxidation (induction periods: 6.7 h
and 8.7 h, respectively) (26). Ozdemir et al. (27) studied the oxidative stability of
several commercial and experimental hazelnut oils using the stability index calcu-
lation ([mg/100-g a-tocopherol] x [% saturated fatty acids/% unsaturated fatty
acid]), showing stability index values between 5.4 and 6.3. Similar findings have
been reported by other research groups (22, 28, 29). The oxidative stabilities of
both stripped and nonstripped hazelnut oil in oil-in-water emulsion and bulk-oil
systems have been reported (28). In these systems, oxygen uptake, peroxide value,
2-thiobarituric acid reactive substances (TBARS), and depletion of a-tocopherol
HAZELNUT 181
levels during a 21-day oxidation cycle at 60 C were assessed. Nonstripped oils
were more stable than stripped oils, and both stripped and nonstripped oils were
more stable as bulk oils than as oil-in-water emulsions (28). More recently, Romero
et al. (29) studied the oxidative stability of stripped and nonstripped hazelnut oils
using the Rancimat method at 100 C; the effects of antioxidants in these systems
were also evaluated. These researchers studied lipid oxidation in five different
hazelnut oil systems: (1) nonstripped cold-pressed hazelnut oil, (2) stripped cold-
pressed hazelnut oil, (3) stripped cold-pressed hazelnut oil with 150-mg/kg
a-tocopherol added, (4) stripped cold-pressed hazelnut oil with 140-mg/kg a-
tocotrienol added, and (5) stripped cold-pressed hazelnut oil with 70 mg/kg of a-
tocopherol and 70-mg/kg a-tocotrienol added (29). The Rancimat studies showed
that a-tocotrienol prolonged the induction period to the greatest extent in the
stripped oil system (37.6 h), followed by the mixture of a-tocopherol and a-
tocotrienol (35.3 h), and finally a-tocopherol (32.6 h). The induction period in the
nonstripped hazelnut oil system (30.8 h) was less than all stripped oils with added
antioxidants, and better than the plain stripped oil system (3.5 h). These results
collectively show that a-tocotrienol exhibits a greater antioxidant activity than
a-tocopherol in this system; both a-tocotrienol and a-tocopherol exhibited greater
antioxidant activities than the minor constituents present naturally in the non-
stripped hazelnut oil (29). This research group also assessed the stability of the
hazelnut oil systems at 180 C by measuring the formation of polar components
and decomposition of tocols in all four antioxidant-containing oil systems over
an 18 h period; under these extreme conditions a-tocopherol was the most effective
antioxidant (29).
Several reports have shown that hazelnut is a health-promoting food and a con-
tributing factor for the beneficial health effects of the Mediterranean style diet (30);
however, few studies have investigated the health effects of hazelnut oil. Balkan
et al. (31) examined the effects of hazelnut oil administration on plasma peroxide
levels, plasma lipid profiles, plasma LDL and VLDL levels, and atherosclerotic pla-
que development in male New Zealand white rabbits. In this study, animals were
fed either control diets, control diets rich in cholesterol (0.5% w/w), control diets
rich in cholesterol (0.5% w/w) with hazelnut oil supplementation (5% w/w), or a
control diet with hazelnut oil supplementation (5% w/w) for 14 weeks. The results
showed that when supplemented in control diets, hazelnut oil reduced plasma cho-
lesterol and apoB-100 containing lipoprotein levels by an insignificant level. No
differences were observed in the high-cholesterol-diet group supplemented with
hazelnut oil, which implies that hazelnut oil may be an effective health-promoting
agent in diets with normal lipid intake, but cannot reverse the effects of high
cholesterol intake (31).
Some researchers have investigated the potential of hazelnuts as a source of
natural antioxidants. Yurttas et al. (32) assessed the phenolic composition of metha-
nolic extracts of hazelnuts, showing that gallic acid, p-hydroxl benzoic acid, caffeic
acid, and sinapic acid were the predominant phenolic acids reported. In addition,
quercetin and epicatechins were present. The composition of phenolic acid consti-
tuents in hazelnut meal has also been assessed by Senter et al. (14) (Table 2).
182 TREE NUT OILS
4. PECAN
Pecan tree (Carya illinoinensis) is native to the United States but has also been
naturalized for commercial pecan production throughout the world, including
Australia, South Africa, and several middle eastern and South America countries
(33). Fat is the predominant constituent in all pecan varieties, ranging from 65%
to 75% (w/w) (1, 33, 34). Other constituents include 13.9% carbohydrate, 9.1%
protein, 3.5% water, and 1.5% ash (w/w) (1). The predominant fatty acids present
in pecan oil are oleic (55%), linoleic (33%), linolenic (2%), palmitic (7%), and
stearic (2%) acids (Table 4) (34). The most predominant tocol in pecan oil was
g-tocopherol (176 mg/kg), followed by a-tocopherol (10 mg/kg), and then d- and
b-tocopherols (6.2 mg/kg) (1). Pecan oil also contains 0.73 g/kg phytosterols that
exist primarily as b-sitosterol (around 90%) (1).
Early studies have shown that pecan oil is a very stable food oil despite its high
content of unsaturated fatty acids, thus making it an excellent dietary oil (35).
Demir and Cetin (36) examined the total yields, compositions, and oxidative stabi-
lities of expeller-pressed and hexane-extracted pecan oils. Total yields were higher
for solvent-extracted batches (6779%, w/w) than pressed batches (36). The expel-
ler-pressed pecan oil had a significantly higher total tocopherol content when com-
pared with hexane-extracted oil (260 mg/kg and 23 mg/kg, respectively); however,
the solvent-extracted oil exhibited a greater oxidative stability with an induction
period of 6 h at 100 C, as compared with 5.5 h for pressed oil. These findings
may imply that antioxidative constituents, aside from tocopherol, may be contribut-
ing to the enhanced oxidative stability of the hexane-extracted oils, however, pre-
vious publications (33, 34), using similar solvent-extraction methods, have shown
much higher concentrations of tocopherols in pecan oils and thus contradict the
findings of Demir and Cetin (36). Toro-Vasquez and Perez-Briceno (37) studied
the oxidative stabilities of solvent-extracted pecan oils from 22 Mexican pecan
varieties; all varieties tested had high oxidative stability index values (8.510.8 h
at 110 C).
TABLE 4. Fatty Acid Composition of Pecan

Oil.a
Fatty Acid (%)
12:0 0.01
14:0 0.05 0.06
14:1 0.02 0.03
16:0 6.49 6.71
16:1 0.20 0.21
18:0 2.23 2.80
18:1 (n-9) 51.1 62.1
18:2 (n-6) 27.2 36.9
18:3 (n-3) 1.52 1.94
20:0 0.12
20:4 0.03
a
Adapted from Wakeling et al. (34).
WALNUT 183
Epidemiological findings show that pecan-enriched diets can favorably alter

serum lipid profiles in humans and thus reduce cardiovascular disease risk (38);
however, the effects of pecan oil intake on human blood lipid profiles have not
been reported.
5. WALNUT
Walnut (nux juglandes) is harvested from walnut tree (Juglans regia) and is the
most popular nut ingredient in North American cooking. Over 30 varieties of wal-
nut trees are currently harvested that have been developed for various characteris-
tics including pest tolerance, early/late harvest, and shell thickness. The major
walnut-producing nations are the United States (California), China, Turkey, India,
France, Italy, and Chile (39).
Walnuts contain about 65% lipids, however, considerable differences exist
among varieties (range: 5270%, w/w) (1, 40). Walnuts also contain 15.8% protein,
13.7% carbohydrate, 4.1% water, and 1.8% ash (w/w) (1). The fatty acid composi-
tion of walnut oil is unique compared with other tree nut oils for two reasons;
walnut oil contains predominantly linoleic acid (4963%) and a considerable
amount of a-linolenic acid (815.5%). Other fatty acids present include oleic
acid (13.826.1%), palmitic acid (6.78.7%), and stearic acid (1.42.5%)
(Table 5) (40). The tocopherol content of walnut oil varies among different cultivars
and extraction procedures and ranges between 268 mg/kg and 436 mg/kg. The pre-
dominant tocol isomer is g-tocopherol (>90%), followed by a-tocopherol (6%),
and then b- and d-tocopherols (41). Nonpolar lipids have been shown to constitute
96.9% of total lipids in walnut oil, whereas polar lipids account for 3.1%. The polar
lipid fraction consisted of 73.4% sphingolipids (ceramides and galactosylcera-
mides) and 26.6% phospholipids (predominantly phosphatidylethanolamine) (42).
Walnut oil contains approximately 1.8 g/kg phytosterols (1), primarily b-sitosterol
(85%), followed by -5-avenasterol (7.3%), campesterol (4.6%), and, finally,
cholesterol (1.1%) (42).
Several research groups have investigated the oxidative stability of walnut oil
and have shown that it is readily oxidizable. Demir and Cetin (36) investigated
the oxidative stability of expeller-pressed and hexane-extracted walnut oil at
TABLE 5. Fatty Acid Composition of Walnut

Oil.a
Fatty Acid (%)
16:0 6.78.7
18:0 1.42.5
18:1 (n-9) 1426
18:2 (n-6) 4963
18:3 (n-3) 816
a
Adapted from Zwarts et al. (40).
184 TREE NUT OILS
100 C by monitoring changes in peroxide value, showing that the induction period
of expeller-pressed walnut oil was 3.5 h compared with 4.5 h for hexane-extracted
oil. The induction period of different walnut oils was 1.52.0 h shorter than values
for pecan oil, as expected considering the existing difference in unsaturation
between the two oils (36). Savage et al. (41) examined the oxidative stability of
walnut oils from 13 different cultivars using the Rancimat method at 100 C; the
induction periods were 3.97.8 h. One notable trend in this study was that the var-
iation of induction period for various walnut oils was inversely correlated with the
levels of linoleic acid and the ratio of total unsaturation to total tocopherol contents
(41). Similar findings were reported by Oliveira et al. (43) using walnut oil obtained
by supercritical fluid extraction with compressed carbon dioxide. More recently,
Amaral et al. (44) examined the oxidative stability of oils from six walnut cultivars
obtained by petroleum ether extraction. In one set of experiments, these researchers
evaluated oil stabilities using the Rancimat method and showed induction periods
between 2.7 h and 3.3 h for walnut oils used. In another set of experiments, the
change in peroxide value after one year of storage was assessed; no obvious trends
were observed among the various oils (44). Findings from stability studies collec-
tively show that walnut oil has low oxidative stability when compared with other
common nut oils, which can be ascribed to its high content of polyunsaturated fatty
acids, mostly a-linolenic acid (44).
Evidence from epidemiologic and intervention studies as well as clinical trials
shows that walnut consumption has favorable effects on serum lipid levels in
humans, such as lowering LDL, raising HDL, and reducing total serum triacylgly-
cerol levels, all of which reduce the likelihood of suffering from a cardiovascular
event (4547). Many of the beneficial findings associated with walnut consumption
have previously been attributed to the polyunsaturated fatty acid intake and have
prompted health researchers to investigate which of these effects, if any, can be
attributed to the lipid component of walnuts. Lavedrine et al. (48) conducted a
cross-sectional study to assess the association between whole walnut and walnut
oil consumption and blood lipid levels. This study included 933 men and women
aged 1865 years living in Dauphine, France (a major walnut-producing area). Fac-
tors used to assess cardiovascular disease risk included a one-year dietary recall
questionnaire and serum levels of HDL, LDL, total cholesterol, and levels of the
apolipoproteins apoA1 and apoB. Results from this study showed that higher levels
of HDL cholesterol and apoA1 were associated with higher amounts of walnut oil
and kernel consumption, with a positive trend existing between the various degrees
of walnut oil/kernel consumption in this cohort. Other blood lipids did not show any
significant association with walnut consumption; the nature of the cohort group
made it impossible to separate the effects of whole walnut and walnut oil consump-
tion (48). More recently, Zibaeenezhad et al. (49) examined the effects of walnut oil
consumption on plasma triacylglycerol levels in hyperlipidemic men and women.
In this trial, 29 patients were given 3g/day walnut oil (six 500-mg capsules per day)
for 45 days, 31 patients were given placebo and were used as controls. Supplemen-
tation of walnut oil reduced serum levels of LDL, triacylglycerol, and total choles-
terol while increasing serum HDL levels, however, only the decrease in serum
PISTACHIO 185
triacylglycerol reached significance (49). The fatty acid composition of walnut oil
has been suggested as being responsible for its cardioprotective feature, but results
from studies, such as that of Espin et al. (50), show that the antioxidative compo-
nents of walnut oil have significant antiradical properties that may exert a protective
effect against the oxidation of biomacromolecules such as LDL, a known risk factor
for atheroma development and, thus, heart disease. Therefore, more studies are
needed to clarify the putative cardioprotective effects of walnut oil consumption
before a casual relationship between the two can be established. The defatted meals
of walnuts provide for an excellent source of natural antioxidants at a level of
0.51 mg/g phenolic acids (14) (Table 2).
6. PISTACHIO
The pistachio tree (Pistacia vera) is native to the Middle Eastern region and has
been naturalized in many parts of the world. The worlds largest producer of pista-
chio nuts is Iran (Kerman Province), with an annual output of 300,000 tons. Other
major producers are Turkey, the United States (California), and Syria (51). Pista-
chio contains 44% lipid, 28% carbohydrate, 21% protein, 4% water, and 3% ash
(w/w) (1). Other research groups have reported that pistachio nuts contain between
45% and 72% oil, depending on the variety and stage of harvest (52, 53). The main
uses of pistachio oil are in the cosmetics and condiment industries. The predomi-
nant fatty acid of pistachio oil is oleic acid (56 64%), followed by linoleic acid
(2331%), palmitic acid (913%), and small amounts of other fatty acids
(Table 6) (54). Pistachio oil contains large amounts of phytosterols (5 g/kg, 85%
b-sitosterol) (55), 270 mg/kg of tocopherols (1, 56), and has an acid value higher
(2.32-mg KOH/g oil) than other tree nut oils (52, 57). Evidence from several
epidemiologic studies suggests that pistachio consumption can reverse several
adverse blood lipid parameters such as hypercholesterolemia (58), however, inves-
tigations on the health effects of pistachio oil consumption are not readily available
or have not been conducted.

of Pistachio Oila
Fatty Acid (%)
16:0 913
16:1 <1
18:0 6.0
18:1 (n-9) 5664
18:2 (n-6) 2331
20:0 <1
24:0 <1
a
Adapted from Kamangar et al. (54).
186 TREE NUT OILS
TABLE 7. Fatty Acid Composition of Brazil

Nut Oil.a
Fatty Acid (%)
16:0 14
16:1 0.3
18:0 8.6
18:1 (n-9) 29
18:2 (n-6) 47
20:0 <1
24:0 <1
a
Adapted from Beuchat and Worthington (60).
7. BRAZIL NUT
Brazil nuts (Bertholletia excelsa) are widely consumed but are produced mainly in
South America, with total world production estimated to be about 20,000 metric
tons. Bolivia, Brazil, and Peru are the main Brazil-nut-producing nations (59).
Brazil nuts are traded mainly in the form of kernels (i.e., shelled) and are used
in confectionery, bakery, and health foods. Brazil nuts contain 6669% lipid,
14.3% protein, 12.2% carbohydrate, 3.5% ash, and 3.5% water (w/w) (1, 60). Brazil
nut oil is used in the areas it is produced as cooking oil and is being promoted on
the export market (59). As the export value of shelled Brazil nuts is so high, usually
only defective Brazil nuts (cracked and partially oxidized) are extracted for their
oils that can result in oils with acid values and peroxide values as high as
5.9-mg KOH/g oil and 7.6-meq oxygen/kg oil, respectively (61). The fatty acid
composition of Brazil nut oil includes 2948% oleic acid, 3061% linoleic acid,
1415% palmitic acid, 68% stearic acid, and 0.5% myristic acid (60, 62) (Table 7).
8. PINE NUT
Pine nuts (pinon or pignolia) are the edible seeds within the pine cone of several
varieties of pine trees (Pinus sp.) but most commonly Pinus pinea or stone pine.
Pine nuts are harvested all over the world, most notably in Russia, China,
North Korea, Spain, Italy, and Turkey, among others. Pine nuts contain 4861%
lipids by weight (1, 60). Other constituents of pine nut include carbohydrate
(19.3%), protein (11.6%), water (5.9%), and ash (2.2%) (1). Pine nut oil contains
predominantly linoleic acid (46.4%) and oleic acid (38.1%). Maritime pine nut
(Pinus pinaster) oil also contains two fatty acids that are unique among tree nut
oils; pinoleic acid and sciadonic acid (Figure 2), which exist at 7% each in pine
nut oil and may have antiatherogenic effects (Table 8) (63). The phenolic acid com-
position of defatted pine nut meal is given in Table 2 and shows that caffeic acid is
the predominant phenolic compound (14).
MACADAMIA NUT 187
H3C COOH
5,11,14-20:3)
Sciadonic Acid (
H3C COOH
5,9,12-18:3)
Pinoleic Acid (
Figure 2. Chemical structures of sciadonic and pinoleic acids.
TABLE 8. Fatty Acid Composition of Pine Nut Oils.a
Fatty Acid Pine Nut Oil (%) Maritime Pine Nut Oil (%)
16:0 5.8 3.6

16:1 0.3 0.2
18:0 3.8 2.4
18:1 (n-9) 38.1 18.1
18:2 (n-6) 46.4 55.9
18:3 (n-3) 0.8 1.3
20:0 0.7 NDb
20:1 0.9 0.8
5,9-18:2 ND 0.7
5,9,12-18:3 ND 7.1
5,11-20:2 ND 0.8
5,11,14-20:3 ND 7.1
a
Adapted from Beuchat and Worthington (60) and Asset et al. (63).
b
None Detected.
9. MACADAMIA NUT
Macadamia trees (Macadamia sp.) were originally cultivated in Australia, but the
United States (Hawaii) is currently the worlds largest producer of macadamia nuts.
Edible macadamia nuts are from two species; Macadamia integrifolia (smooth-shell
type) and Macadamia tetraphylla (rough-shell type). The macadamia nut industry
in Hawaii, Australia, and many other producing areas, is based primarily on the
smooth-shell type (64). Oil yields from macadamia nuts range from 59% to 78%
(w/w) (1, 65, 66). Macadamia nuts also contain 13.8% carbohydrate, 7.9% protein,
1.4% water, and 1.1% ash (w/w) (1). Compositional studies of macadamia nut oil
shows that it is rich in oleic and palmitoleic acids (Table 9) (67), has 1854 mg/kg
tocol isomers (predominatly a-tocotrienol), and up to 1.5 g/kg phytosterols (predo-
minatly campesterol) (66). Macadamia nut oil has been shown to have a relatively
high smoke point of 198 C. The Rancimat method has been used to assess the
oxidative stability of several varieties of macadamia nut oil, resulting in induction
periods of between 3.6 h and 19.8 h (66).
188 TREE NUT OILS
TABLE 9. Fatty Acid Composition of Maca-

damia Nut oil.a
Fatty Acid (%)
16:0 7.9
16:1 17
18:0 3.3
18:1 (n-9) 57.7
18:2 (n-6) 1.7
20:0 <1
24:0 <1
a
Adapted from Macfarlane et al. (67).
10. CASHEW NUT
The cashew (Anacardium occidentale L.) is an evergreen species native to tropical

America and contains 47% oil (w/w) (1, 68). Other components of cashew nuts
include carbohydrate (27.1%), protein (18.2%), water (5.2%), and ash (2.5%) The
predominant fatty acid in cashew nut oil is oleic acid (57.365.1%), followed
by linoleic (15.618.6%), and palmitic (9.014.2%) acids (Table 10) (68). Cashew
nut oil contains 1.4% unsaponifiable matter (w/w), of which 76.282.7% is
b-sitosterol. Other sterols present in cashew nut oil include 5-avenasterol,
campesterol, fucosterol, cholesterol, and stigmasterol (68). Cashew nut oil con-
tains 45.383.5 mg/100 g g-tocopherol; other tocopherols present are a-tocopherol
(2.88.2 mg/100g) and d-tocopherol (2.05.9 mg/100 g) (68).
11. USE OF DEFATTED TREE NUT MEALS AND OTHER

BYPRODUCTS AS PROTEIN SOURCES
Defatted tree nut meals and hulls are traditionally used as animal feeds due to their
low cost and the high nutritional value of their proteins and other constituents (69).
Tree nut byproducts have many food (70) and biochemical applications (71). Tree

of Cashew Nut Oil.a
Fatty Acid (%)
16:0 9 14.2
16:1 0.3 0.4
18:0 6.3 11.6
18:1 (n-9) 57.3 65.1
18:2 (n-6) 15.6 18.1
20:0 0.3 0.8
a
Adapted from Toschi et al. (68).
TABLE 11. Amino Acid Profiles (%) of Tree Nut Proteins.a
Amino Acid Almond Hazelnut Pecan Walnut Pistachio Brazil Nut Pine Nut Macadamia Nut Cashew
Alanine 4.54 4.67 4.46 4.41 4.59 3.71 7.24 3.73 4.18
Arginine 11.2 14.2 13.2 14.4 10.1 13.8 26.9 13.5 10.6
Aspartic Acid 12.4 10.5 10.4 11.6 9.06 8.67 12.6 10.5 8.96
Cystine 1.28 1.51 1.70 1.31 1.78 2.36 2.51 0.05 1.96
Glutamic Acid 23.5 23.3 20.5 17.8 19.1 20.3 23.5 21.8 22.5
Glycine 6.67 4.65 5.09 5.17 4.75 4.62 7.05 4.37 4.68
Histidine 2.69 2.16 2.94 2.48 2.52 2.48 3.31 1.87 2.27
Isoleucine 3.14 3.75 3.77 3.96 4.49 3.32 5.38 3.02 3.94
Leucine 6.68 7.26 6.72 7.42 7.75 7.44 9.98 5.79 7.35
Lysine 2.73 2.63 3.22 2.68 5.74 3.17 5.19 0.17 4.63
Methionine 0.85 1.07 2.05 1.49 1.68 6.49 2.47 0.22 1.80
Phenylalanine 5.22 4.53 4.78 4.50 5.29 4.06 5.30 6.40 4.75
Proline 4.40 3.36 4.08 4.47 4.05 4.23 7.44 4.50 4.05
Serine 4.57 4.41 5.32 5.92 6.11 4.40 5.87 4.03 5.38
Threonine 3.08 2.95 3.44 3.78 3.35 2.33 4.39 3.56 3.44
Tryptophan 0.87 1.42 1.04 1.07 1.36 0.90 1.74 0.64 1.43
Tyrosine 2.41 2.99 2.41 2.57 2.07 2.70 5.07 4.92 2.53
Valine 3.63 4.37 4.61 4.77 6.18 4.87 7.15 3.49 5.46
a
Values adapted from USDA Nutrient Database for Standard Reference (1).
190 TREE NUT OILS
nut meals are rich in several antioxidative compounds and other health-promoting
substances. This has led researchers to investigate their potential as functional food
ingredients and as possible sources of nutraceutical extracts (70). The predominant
nutritional component of tree nut meals is protein, constituting around 40% of total
weight (70). The protein component is of high quality with a reasonably balanced
composition of essential amino acids (1). As an example, cashew nut meal contains
42% crude protein and, compared with soybean meal, it has been shown to enhance
livestock weight-gain curves and a higher protein score (97 vs. 93, respectively)
(72). Similar findings have also been reported for walnut meal (73). The amino
acid compositions of proteins from tree nut meals are summarized in Table 11
and show that, in most cases, glutamic acid, arginine, and aspartic acid account
for about 40% of the amino acids in these proteins, whereas tryptophan is a limiting
amino acid in all tree nut proteins examined in this chapter, except macadamia nut
protein, which contains only trace amounts of cystine. Thus, defatted meals of tree
nuts serve as excellent sources of high-quality proteins.
12. CONCLUDING REMARKS
Several tree nut varieties serve as valuable oil crops due to their high oil yield,
unique flavors, and healthful lipid composition. Byproducts of tree nuts also have
several uses including functional food ingredients and as sources of nutraceutical
extracts and dietary protein. Compared with most other vegetable oils, tree nut
oils show high oxidative stability, which is due to high levels of monounsaturated
fatty acids rather than polyunsaturated fatty acids and high concentrations of minor
components with antioxidant activity. The use of tree nut oils and byproducts in
every day cooking is very common in some parts of the world and is becoming
more widespread due to increased consumer demand for alternative and health-
promoting foods. The consumption of high-fat tree nuts and their oils has been
shown to have antiatherogenic effects, which may be related to the known positive
cardiovascular health effects of unsaturated fatty acids, phytosterols, and tocol
isomers. Other minor phytochemicals present in tree nut oils may also contribute
to their observed health effects. Less information is available regarding the health
effects of tree nut byproducts.
REFERENCES
1. United States Department of Agriculture (USDA) Nutrient Database. (2004). Available:

www.nal.usda.gov/fnic/foodcomp/search.
2. J. S. Bonvehi and F. V. Coll, Food Chem., 48, 237241 (1993).
3. N. Madhaven, Inter. J. Toxicol., 20, 1520 (2001).
4. F. B. Hu, M. J. Stampfer, J. E. Manson, E. B. Rimm, G. A. Colditz, B. A. Rosner,
F. E. Speizer, C. H. Hennekens, and W. C. Willett, BMJ, 317, 13411345 (1998).
REFERENCES 191
5. S. M. Grundy, L. Florentin, D. Nix, and M. F. Whelan, Amer. J. Clin. Nutr., 47, 965969
(1988).
6. M. Martins, R. Tenreiro, and M. M. Oliveira, Plant Cell Rep., 22, 7178 (2003).
7. G. A. Spiller, D. A. J. Jenkins, O. Bosello, J. E. Gates, L. N. Cragen, and B. Bruce, J. Amer.
Coll. Nutr., 17, 285290 (1998).
8. Z. Shi, Q. Fu, B. Chen, and S. Xu, Chinese J. Chromatogr., 17, 506507 (1999).
9. A. Sattar, M. Jan, A. Ahmad, A. Hussain, and A. Khan, Nahrung, 33, 213215
(1989).
10. F. Salvo, M. Alfa, and G. Dugo, Rivista Italiana delle Sostanze Grasse, 63, 3740
(1986).
11. J. Sabate and G. E. Fraser, Curr. Opin. Lipidol., 5, 1116 (1994).
12. G. E. Fraser, H. W. Bennett, K. B. Jaceldo, and J. Sabate, J. Amer. Coll. Nutr., 21, 275283
(2002).
13. D. A. Hyson, B. O. Schneeman, and P. A. Davis, J. Nutr., 132, 703707 (2002).
14. S. D. Senter, R. J. Horvat, and W. R. Forbus, J. Food Sci., 48, 798799 (1983).
15. S. Siriwardhana and F. Shahidi, J. Amer. Oil Chem. Soc., 79, 903906 (2002).
16. S. Sang, K. Lapsley, W. S. Jeong, P. A. Lachance, C. T. Ho, and R. T. Rosen, J. Agric. Food
Chem., 50, 24592463 (2002).
17. S. Frison-Norrie and P. Sporns, J. Agric. Food Chem., 50, 27822787 (2002).
18. M. Pinelo, M. Rubilar, J. Sinero, and M. J. Nunez, Food Chem., 85, 267273 (2004).
19. S. Sang, H. Kikuzaki, K. Lapsley, R. T. Rosen, N. Nakatani, and C. T. Ho, J. Agric. Food
Chem., 50, 47094712 (2002).
20. P. A. Davis and C. K. Iwashi, Cancer Lett., 165, 2733 (2001).
21. C. Alasalvar, F. Shahidi, C. M. Liyanapathirana, and T. Ohshima, J. Agric. Food Chem.,
51, 37903796 (2003).
22. C. Alasalvar, F. Shahidi, T. Ohshima, U. Wanasundara, H. C. Yurttas, C. M. Liyana-
pathirana, and F. B. Rodrigues, J. Agric. Food Chem., 51, 37973805 (2003).
23. J. Parcerisa, D. G. Rishardson, M. Rafecas, R. Codony, and J. Boatella, J. Agric. Food
Chem., 45, 38873890 (1997).
24. P. Pfnuer, T. Matsui, W. Grosch, H. Guth, T. Hofmann, and P. Schieberle, J. Agric. Food
Chem., 47, 20442047 (1999).
25. M. L. Ruiz del Castillo, M. Mar Caja, M. Herrair, and G. B. Blanch, J. Agric. Food Chem.,
46, 51235132 (1998).
26. M. G. Bernardo-Gil, J. Grenha, J. Santos, and P. Cardoso, Eur. J. Lipid Sci. Technol., 104,
402409 (2002).
27. M. Ozdemir, F. Ackurt, M. Kaplan, M. Yildiz, M. Loker, T. Gurcan, G. Biringen, A. Okay,
and F. Seyhan, Food Chem., 73, 411417 (2001).
28. S. F. OKeefe, V. A. Wiley, and D. A. Knauft, J. Oil Chem. Soc., 70, 489 492 (1993).
29. N. Romero, P. Robert, L. Masson, J. Ortiz, J. Pavez, C. Garrido, M. Foster, and
C. Dobarganes, J. Sci. Food Agric., 84, 943948 (2004).
30. P. M. Kris-Eterton, J. Nutr., 129, 22802284 (1999).
31. J. Balkan, A. Hatipoglu, A. Gulcin, and M. Uysal, J. Agric. Food Sci., 51, 39053909
(2003).
32. H. C. Yurttas, H. W. Shafer, and J. J. Warthesen, J. Food Sci., 65, 276280 (2000).
192 TREE NUT OILS
33. L. T. Wakeling, R. M. Mason, B. R. DArcy, and N. A. Caffin, Food Aust., 52, 574578
(2000).
34. L. T. Wakeling, R. M. Mason, B. R. DArcy, and N. A. Caffin, J. Agric. Food Sci., 49,
12771281 (2001).
35. T. M. Pyriadi and M. E. Mason, J. Amer. Oil Chem. Soc., 437 440 (1968).
36. C. Demir and M. Cetin, Deutsche Lebensmittel Rundschau, 95, 278282 (1999).
37. J. F. Toro-Vazquez and F. Perez-Briceno, J. Food Lipids, 5, 211231 (1998).
38. S. Rajaram, K. Burke, B. Connel, T. Myint, and J. Sabate, J. Nutr., 131, 22752279
(2001).
39. Walnut Industry Fact Sheet. (2004). Available: www.walnut.org/pdfs/walnuts_fact-
sheet.pdf.
40. L. Zwarts, G. P. Savage, and D. L. McNeil, Int. J. Food Sci. Nutr., 50, 189194
(1999).
41. G. P. Savage, P. C. Dutta, and D. L. McNeil, J. Amer. Oil Chem. Soc., 1999, 76, 10591065.
42. G. Tsamouris, S. Hatzinantoniou, and C. Demetzos, J. Biosciences, 57, 5156 (2002).
43. R. Oliveira, M. Fatima Rodrigues, and M. G. Bernardo-Gil, J. Amer. Oil Chem. Soc., 79,
225231 (2002).
44. J. S. Amaral, S. Casal, J. A. Pereira, R. M. Seabra, and B. P. P. Oliveira, J. Agric. Food
Chem., 51, 76987702 (2003).
45. M. Abbey, M. Noakes, G. B. Belling, and P. J. Nestel, Amer. J. Clin. Nutr., 59, 995999
(1994).
46. D. Zambon, J. Sabate, S. Munoz, B. Campero, E. Casals, M. Merlos, J. C. Laguna, and
E. Ros, Ann. Intern. Med., 132, 538546 (2000).
47. J. Sabate, G. E. Fraser, K. Burke, S. F. Knutsen, H. Bennet, and K. D. Lindsted, N. Engl.
J. Med., 328, 603607 (1993).
48. F. Lavedrine, D. Zmirou, A. Ravel, F. Balducci, and J. Alary, Preventive Med., 28, 333
341 (1999).
49. M. J. Zibaeenezhad, M. Rezaiezadeh, A. Mawla, S. M. Ayatollahi, and M. R. Panjahshahin,
Angiology, 54, 411 414 (2003).
50. J. C. Espin, C. Soler-Rivas, and H. J. Wichers, J. Agric. Food Chem., 48, 648657 (2000).
51. Global Pistachio Production and Marketing Challenges. (2004). Available: http://eco-
nomics.ca/2003/papers/0460.pdf.
52. M. Yousfi, B. Nedjmi, R. Bellal, D. B. Bertal, and G. Palla, J. Amer. Oil Chem. Soc., 79,
10491050 (2002).
53. M. Aslan, I. Orhan, and B. Sener, Int. J. Food Sci. Nutr., 37, 333335 (2002).

54. T. Kamangar, F. Farrahi, and M. Mehran, J. Amer. Oil Chem. Soc., 52, 512513 (1975).
55. M. Yildiz, S. Turcan Gurcan, and M. Ozdemir, Fett/Lipid., 3, 8486 (1998).
56. A. Danechrad, Oleagineux., 29, 153154 (1974).
57. A. M. H. Al-Shaibani, L. D. K. Al-Doori, and O. M. Nori, Zanco., 4, 93101 (1986).
58. K. Edwards, I. Kwaw, J. Matud, and I. Kurtz, J. Amer. Coll. Nutr., 18, 229232 (1999).
59. Economic Viability of Brazil Nut Trading in Peru. (2004) Available: www.nri.org/NRET/
brazilnuts.pdf.
60. L. R. Beuchat and R. E. Worthington, J. Fd. Technol., 13, 355358 (1978).
REFERENCES 193
61. E. M. R. Gutierrez, M. A. B. Regitano-DArce, and A. M. O. Rauen-Miguel, Ciencia e

Tecnologia de Alimentos., 17, 2227 (1997).
62. E. W. Ecky, Vegetable Fats and Oils, van Nostrand Reinhold Publishing Co, New York,
1954.
63. G. Asset, E. Bauge, R. L. Wolff, J. C. Fruchart, and J. Dallongeville, Eur. J. Nutr., 40,
268274 (2001).
64. H. Ako, D. Okuda, and D. Gray, Nutrition, 11, 286288 (1995).
65. C. Croy, Inform, 5, 970971 (1994).
66. A. Kaijser, P. Dutta, and G. Savage, Food Chem., 71, 6770 (2000).
67. N. Macfarlane, and R. V. Harris, AOCS Monograph, AOCS Press, Champaign, Illinois,
1981, pp. 9, 103108.
68. T. G. Toschi, M. F. Caboni, G. Penazzi, G. Lercker, and P. Capella, J. Amer. Oil Chem.
Soc., 70, 10171020 (1993).
69. D. Harvey, Commonwealth Bureau of Animal Nutrition, 19, 105113 (1970).
70. M. L. Souza and H. C. Menezes, Ciencia e Tecnologia de Alimentos, 24, 120128
(2004).
71. W. J. Wolf and S. K. Sathe, J. Sci. Food Agric., 78, 511521 (1998).
72. G. Piva, E. Santi, and T. Ekpenyong, J. Sci. Food Agric., 22, 2223 (1971).
73. G. P. Savage, Plant Foods Hum. Nutr., 56, 7582 (2001).
8
Germ Oils from
Different Sources
Nurhan Turgut Dunford
Oklahoma State University
Stillwater, Oklahoma
1. INTRODUCTION
Commercial plant germ oils are mainly obtained from cereal grains, such as corn,
wheat, and rice. In most cases, the endosperm of the grain is the part of interest for
the industrial use of cereals. Starch and proteins are the major components of the
endosperm. Endosperm represents 7585% of the grain and protects the embryo,
which is also referred to as germ.
The germ of a cereal constitutes about 23% of the grain and can be separated in
a fairly pure form during the milling operation. Although lipids are present in rela-
tively small quantities in grains, they play an important role in cereal processing
and nature of the products by affecting the properties of protein and starch. Most
of the oil present in the grain is found in the germ fraction of the cereal. For some
cereals such as rice, corn, and wheat, oil has economic significance. Some cereals,
such as oats, contain a considerable amount of lipid in the endosperm. Lipid content
of the germ varies with the grain type, and it can be as high as 60%. Even though a
small amount of germ oil is commercially extracted from other grains, such as oats
and barley, the most important commercial germ oils are rice, corn, and wheat germ
oil (WGO).
195
196 GERM OILS FROM DIFFERENT SOURCES
TABLE 1. World Wheat Production.1
Production
(x106 Metric Tons)
Year 2001 2002 2003

Canada 21 16 24
China, Mainland 94 90 86
India 70 73 65
USA 53 44 64
World 591 574 556
1
FAOSTAT Database (http://apps.fao.org) (11).
Several reviews have been published on wheat (1, 2) corn (37), and rice bran
oils (810) over the years. Therefore, this chapter, with sufficient background
information, will emphasize the latest literature on composition, nutritional charac-
teristics, and processing methods of plant germ oils.
2. WHEAT GERM
2.1. Production and Use

Wheat is one of the leading grain crops produced, consumed, and traded worldwide.
About 590 million metric tons of wheat is produced globally each year. China, the
United States, and Canada are among the largest wheat growers (Table 1) (11).
Statistical data on wheat germ production is not readily available. However it
can be estimated that about 10 million tons of wheat germ could be obtained
from wheat milling operations worldwide based on the fact that germ constitutes
about 2% of the whole wheat grain. Although wheat varieties differ in oil content,
in general, the whole kernel contains about 24% lipids. Lipid content of endo-
sperm is usually less than 2%. Wheat germ contains about 814%, (w/w) oil (12).
Typical proximate composition of commercial wheat germ is shown in Table 2
(2). Wheat germ has great potential as a highly nutritious food supplement. Wheat
germ provides three times as much protein, seven times as much fat, 15 times
as much sugars, and six times as much mineral content as wheat flour (13). The
TABLE 2. Wheat Germ Proximate Composition.1
Compound % (w/w)
Protein 26
Crude Fiber 3
Starch 20
Sugars 16
Oil 10
Moisture 6
Ash 4
1
Adapted from Barnes (2).
WHEAT GERM 197
protein content of wheat germ is about 30% (w/w). Although of plant origin, germ
protein has similar nutritive value as animal protein (14). All these properties make
wheat germ very attractive for enrichment or supplementation of various processed
food products.
Wheat germ oil has a number of nutritional and health benefits, such as reducing
plasma and liver cholesterol levels, improving physical endurance/fitness, and
delaying aging (15). These effects are attributed to the high concentration of bioac-
tive compounds present in the germ. Wheat germ is one of the richest natural
sources of a-tocopherol, which possesses Vitamin E activity (16). The wheat
germ market is mainly based on its high Vitamin E content, and WGO is marketed
in bottles or in capsules as a dietary supplement. Wheat germ oil is also added to
lecithin and cod liver oil. Wheat germ oil has been reported to improve human phy-
sical fitness, and this effect is attributed to its high polycosanol (PC), specifically to
its high octacosanol (OC) content (15). There is a growing interest in wheat germ
octacosanol as a potential nutraceutical and functional food ingredient.
Wheat germ oil is used in products such as foods, biological insect control
agents, pharmaceuticals, and cosmetic formulations (15). Wheat oil betaine, ami-
doaminelactate, sulfosuccinate, and amidopropylamine oxide are some of the
WGO-based surfactants used in cosmetics (17, 18). Wheat germ oil and its volatile
components, such as C13C16 saturated and unsaturated hydrocarbons, branched
hexylbenzene, octanoic acid, g-nanolactone, substituted naphthalenes, and cyclic
branched ketones, have been reported to have activities as biological insect control
agents (19, 20). Wheat germ products are also marketed as dietary supplements for
farm animals, racehorses, pets, and mink.
2.2. Grain Structure

A sketch of the longitudinal and histological section of a wheat grain is shown in
Figure 1 (22). The wheat caryopsis (the fruit of the cereal) contains a single seed.
The outermost layer of a seed is the pericarp, which forms a tough protective layer
at maturity. The tissue directly surrounding the seed is referred to as testa or seed
coat. The whole grain consists of two parts, namely the embryo and the endosperm.
The starchy endosperm is the major portion of the seed at maturity and serves as a
food reserve to be used by the embryo at germination. The aleurone layer is the
outermost layer of the endosperm. The embryo or germ is formed of thin-walled
cells and consists of embryo axis and scutellum. The embryo axis develops into
the first roots and shoots of the new plant. The scutellum is a tissue layer located
between the embryo and the endosperm, and it plays a role in the translocation of
hydrolyzed sugars from endosperm to the embryo during germination. Depending
on the variety of cereal, either the germ is surrounded by the endosperm or placed
laterally and produces a slight protuberance from the grain.
2.3. Germ Recovery

Wheat germ is separated from the endosperm during the milling operation. The
basic principles of modern wheat milling technology are used worldwide, whereas
Figure 1. Structure of wheat grain (21).
slight variations are observed regionally. In general, wheat is transferred from ele-
vator to screen room to separate out chaff, dust, stones, mud balls, glass, nonferrous
metal, and grains other than wheat (oats, barley, etc.). Magnets placed at various
positions throughout the mill remove ferrous metals. Most mills use dry rather
than wet systems to clean wheat before tempering. Tempering is the controlled
addition of moisture to wheat to achieve the following objectives: (1) to toughen
the skin to resist powdering during milling; (2) to facilitate the physical separation
of endosperm and bran; and (3) condition the endosperm for easy size reduction and
sieving for flour production. Tempered wheat is then milled. The milling process
involves grinding of the grain and separation of the fractions. Milling breaks
open the grain and scrapes off as much endosperm from the bran skin as possible.
A basic milling operation consists of the following systems: (1) a series of break
rolls; (2) grading; (3) purification; (4) sizing (scratch); (5) size reduction; (6) flour
dressing; and (7) millfeed (byproducts of wheat flour milling).
WHEAT GERM 199
In a conventional mill, wheat germ stock occurs primarily in the chop (ground
material leaving a break roll from the first two breaks). Germ recovery is completed
in a germ separator. The separation and purification of the germ is achieved by tak-
ing advantage of two physical characteristics of germ. Germ tends to fragment into
particles coarser than bran and endosperm particles because of its plasticity, which
is due to its high oil content. Germ also tends to flatten, rather than crush like endo-
sperm. These properties facilitate germ separation by size. The density of germ is
greater than that of the other wheat fractions. Thus, wheat germ can also be isolated
based on specific gravity differences. In some mills, which are designed to maxi-
mize germ recovery, wheat grain is first passed through an impact machine that
releases practically the entire germ along with a small quantity of fines. Broken
wheat is then sent to the first break rolls after sifting out the fines. Then germ is
separated and purified in a germ separator based on the above described principles.
After separation, germ is passed through smooth reduction rolls to produce flakes.
Flat, thin, and large flakes characterize good-quality wheat germ.
Scutellum is relatively friable and difficult to separate from the other milling
fractions. Separation of the embryo axis is easier. After size reduction, the embryo
axis goes on to the reduction rolls with coarse middlings where it is flaked and can
be separated by sieving. Thus, commercial wheat germ consists predominantly of
the embryo axis. The bran and endosperm content of commercial germ may vary
with the extent and sophistication of the milling operations. During the flaking pro-
cess, some oil is pressed out of the germ and transferred to the flour. Hence, the
lipid content of the commercial WGO is different than that of the original embryo
axis. In mills with no special germ-separation equipment, germ recovery yields are
usually very low, about 0.5% for food-grade germ. Yields can be as high as 1.52%
in mills with advanced germ-recovery systems.
2.4. Oil Content

The oil content of wheat germ varies with variety, purity of germ, and extraction
method. Pure germ fractions prepared in the laboratory (dissected by hand) contain
a much higher amount of oil (15%) than commercially flaked wheat germ
(711%), because some oil is lost by expression and transferred to flour during
the commercial operations (23). In general, dissected germ contains both embryo
and scutellum. It has been reported that scutellum contains about two times more
oil than does the embryo (24). The higher oil content in scutellum explains higher
oil content of laboratory-dissected germ. Higher oil yields (2530%) were obtained
when total acyl lipid content of wheat germ was determined by acid hydrolysis,
which releases oil from the germ matrix (25).
2.5. Oil Properties and Lipid Composition

Physicochemical properties of WGO are summarized in Table 3 (26). The specific
gravity of WGO varies from 0.925 to 0.938. Typical refractive index of WGO is in
the range of 1.469 to 1.483. Iodine and saponification values of the oil are 115128
TABLE 3. Physicochemical Properties of Wheat Germ Oil.1
Property
Specific Gravity 0.9280.938

(15.5/15.5 C)
0.9250.933
(25/25 C)
Refractive Index 1.4741.483 (25 C)
1.4691.478 (40 C)
Iodine Value 115128
Saponification Value 179190
Unsaponifiable (%) 25
1
Adapted from Firestone (26).
and 179190, respectively. Free fatty acid content of WGO is usually less than 6%;
however, it can be as high as 25% if the germ separation, storage, and oil-extraction
conditions are not controlled properly. Solvent-extracted crude oil usually has a
lower FFA content than that of the mechanically expelled oil. Free fatty acids
are not desirable in the oil as a result of their contribution to a bitter and soapy fla-
vor in food products; hence, they have to be removed during the edible oil-refining
process. Unsaponifiable matter content of WGO is higher, 1.58%, than that of the
most other edible oils. Composition of the unsaponifiable content of WGO is
discussed later in this chapter.
2.5.1. Fatty Acid Composition The fatty acid composition of commercial and
laboratory-extracted wheat germ oil has been reviewed by Barnes (2). Significant
variations were observed in the fatty acid composition of commercial WGO. These
variations were attributed to differences in varieties of wheat, growth conditions,
storage conditions of the germ, method of lipid extraction and analysis, adulteration
with other vegetable oils, and post extraction treatments such as removal of fatty
acids. However, fatty acid compositions of laboratory-extracted oils (hexane
extracts) obtained from wheat germ processed at different milling operations
were similar (2).
Hexane-extracted wheat germ consisted of about 56% linoleic acid (18:2 n6),
which is an essential fatty acid (Table 4) (27). Total unsaturated and polyunsatu-
TABLE 4. Comparison of Fatty Acid Composition1of Wheat Germ Oil Extracted

with SC-CO22 and Soxhlet3 Methods.
Fatty Acid 16:0 18:0 18:1n-9 18:2n-6 18:3n-3 20:0 20:1n-9 20:2n-6 22:1n-9 24:1n-9
SC-CO22 16.4 0.6 14.0 56.2 6.1 0.2 1.6 0.2 0.5 0.1
Soxhlet3 16.7 0.7 14.6 56.5 6.2 0.2 1.5 0.2 0.4 0.2
1
GC area percentage.
2
Extracted at 550 Bar and 40 C.
3
Extracted with hexane.
WHEAT GERM 201
rated fatty acid (PUFA) content of wheat germ oil was about 81% and 64%, respec-
tively. It has been well documented that unsaturated fatty acid, especially PUFA,
intake reduces coronary hearth disease (CHD) (28). Several scientific studies
have shown that n-3 fatty acids have health benefits, such as lowering CHD risk
(29). It has been also suggested that n-6/n-3 ratio of ten or less results in reduction
in fatal CHD risk (29). The n-6/n-3 recommendations of the World Health Organi-
zation, Sweden, and Japan are 510/1, 5/1, and 2/1, respectively. Wheat germ oil
has very high unsaturated and polyunsaturated fatty acid content and an excellent
n-6/n-3 fatty acid ratio (9/1). A high concentration of PUFA is a positive attribute in
the functional foods and nutraceutical market. However, a high content of C18:3
fatty acid makes the oil susceptible to oxidative rancidity.
Research findings of Dunford and Zhang (27) also demonstrated that there was
no significant change in fatty acid composition of the oils extracted with various
organic solvents (hexane, ethanol, isopropanol, and acetone) even though the
extract yield was affected considerably by the solvent type.
Wang and Johnson (30) have reported that neutralized WGO had higher palmi-
tic, stearic, and oleic acids content and lower linoleic and linolenic acids content as
compared with crude and degummed oils. This phenomenon was attributed to two
reasons: (1) selective hydrolysis of triacylglycerols (TAG) during germ separation
and oil extraction, and (2) removal of phospholipids, which usually contain more
PUFA as compared with the neutral oil during the deacidification process.
2.5.2. Acyl Lipids The composition of nonpolar acyl lipids in WGO was
reported by Barnes (2). Triacylglycerols are the major lipid class in WGO
(Table 5). According to Nelson et al. (31), about 30% of the TAG consists of
1-palmito-2,3-dilinolein. Trilinolein (16%) and 1-palmito-2-linoleo-3-olein
(12%) are the two other major TAGs present in WGO. Details on distribution
of specific fatty acids in the TAG of WGO are discussed by Barnes (2) and
Nelson et al. (31).
Diacylglycerols (DAGs) content of commercial oils varies from 2% to 11% (2).
Monoacylglycerols (MAGs) content of WGO (0.11.0%) is usually lower than the
DAG content. It had been also reported that fatty acid composition of MAG was
significantly different than that of the other acyl lipids (32). Linoleic acid content
of MAG was significantly lower than that of the other acyl lipids.
TABLE 5. Nonpolar Acyl Lipid Composition

of Laboratory Extracted Wheat Germ Oil.1
Lipid Class Composition (%)
Triacylglicerols 63.988.5
Diacylglicerols 1.86.9
Monoacylglycerols 0.31.1
Free Fatty Acids 0.622.0
Phytosterol Fatty Acid Esters 5.15.8
1
TABLE 6. Effect of Processing on the Phosphorous

Content of Wheat Germ Oil.1
Oil Type Phosphorous (ppm)
Crude 1428
Degummed 1082
Neutralized 99
Bleached 22
Cold-processed 786
Cold pressed 74
1
Adapted from Wang and Johnson (30).
Wheat germ oil also contains polar lipids. Reported values for WGO polar lipids
vary considerably with the extraction method and solvent used for lipid recovery.
Hargin and Morrison (33) reported over 20% polar lipids in dissected wheat germ
chlorform-methanol extracts. Commercial oils produced by pressure expelling
contain 0.21.8% (expressed as of total acyl lipids) polar lipids. Commercially
solvent-extracted crude WGO consists 0.310% polar lipids as of the total acyl
lipids (32). Analytical data on WGO phospholipids (PL) is scarce. In dissected
wheat germ, phosphatidylcholine (PC) represents about 4060% of total PL and
phosphatidylethanolamine (PE) (915%), and phosphatidylinositol (PI) (1320%)
are also present in large amounts (33). The effect of processing on the phosphorous
content of the oil was examined by Wang and Johnson (Table 6) (30). Wheat germ
oil degumming was very difficult as a result of the presence of a large amount of non-
hydratable PL caused by the phospholipase D activity during wheat milling (30).
2.5.3. Unsaponifiable Components Wheat germ oil is quite rich in unsaponi-

fiable compounds, in particular phytosterols and tocopherols (Table 7). Small quan-
tities of triterpenols, n-alkanols, carotenoids, and hydrocarbons are also present
in the unsaponifiable fraction of the oil. Tocopherols and n-alkanols are commer-
cially the most important unsaponifiable components in WGO.
2.5.3.1. Tocols Tocopherols constitute about 18% of the unsaponifiable fraction

of WGO (Table 7). Wheat germ oil is one of the richest natural sources of
TABLE 7. Unsaponifiable Content of Wheat

Germ Oil.1
Compounds % of Unsaponifiables
Tocopherols 18
Hydrocarbons 7
n-alkanols Triterpenols 9
Methylsterols 17
Phytosterols 35
Others 14
1
WHEAT GERM 203
TABLE 8. Tocol Content of Wheat Germ Oil.
SC-CO2 Extraction (550 Bar, 80 C) Soxhlet Extraction (Hexane)
Extraction time (min) 15 30 45 60

a- Tocopherol (ppm) 1353 1320 1365 1176 1377
a-Tocotrienol (ppm) n.d.1 9 n.d.1 n.d.1 n.d.1
b- Tocopherol (ppm) 1005 945 998 1277 1209
b-Tocotrienol
g-Tocopherol (ppm) 23 19 24 31 48
g- Tocotrienol (ppm) n.d.1 n.d.1 n.d.1 47 n.d.1
d- Tocopherol (ppm) 4 5 5 16 5
d- Tocotrienol (ppm) n.d.1 n.d.1 n.d.1 12 7
1
Not detected.
a-tocopherols. Synthetic Vitamin E, which is a dl-a-tocopherol, is commercially

available. The natural isomer of a-tocopherol occurs in the d-form and posseses
greater biological activity than that of the synthetic product. Tocopherols, particu-
larly a-tocopherol, are also known to have antioxidant activity (34). The composi-
tion of tocopherol isomers in WGO oil is shown in Table 8. Wheat germ oil is rich
in both a- and b-tocopherols. Bran and endosperm contain less tocopherol than that
of the germ; hence, their presence reduces tocopherol content in the WGO extracted
from germ contaminated with nongerm fractions. It is interesting to note that oil
from fresh wheat germ contained similar amounts of tocopherol as the oil from ran-
cid germ (32) indicating that conditions causing lipid deterioration did not have any
significant adverse effect on the tocopherols. A commercial form of Vitamin E
(a-tocopherol acetate) was detected in some commercial WGO (2). This might
be caused by synthetic Vitamin E fortification of these commercial oils because
no significant amount of a-tocopherol acetate has been detected either in any plant
lipid extracts or laboratory-extracted WGO (2). A very small amount of tocotrienols
was also detected in the WGO extracted in our laboratories (Table 8). According to
Barnes (1), only a- and b-tocopherols are present in the pure dissected wheat germ.
The presence of tocotrienols in the germ oil is attributed to the contaminations from
endosperm and bran.
2.5.3.2. n-Alkanols n-Alkanols are a group of high-molecular-weight primary

fatty alcohols present in many plants. Information on the composition of n-alkanols
in WGO is scarce. The major component of wheat wax is octacosanol, which is an
n-alkanol with chemical formula CH3(CH2)26CH2OH and molecular weight of
410.8 (35, 36). Barnes (1) reported that unrefined, unadulterated, solvent-extracted
wheat germ oil contains 80 mg/kg of octacosanol. A U.S. patent, which has expired,
describes separation of a mixture of octacosanol and triacontanol from a wheat
germ oil unsaponifiable fraction (37).
2.5.3.3. Sterols Phytosterols constitute a major fraction of the WGO unsaponi-

fiables (about 35%) (Table 7). According to Itoh et al. (38), WGO contains a
significantly higher amount of phytosterols than do the other common commercial

oils. Sitosterol (6070%) and campesterol (2030%) are the two major phytosterols
present in WGO (3840). The majority of the phytosterols in WGO are present in
an esterified form (41). According to Kiosseoglou and Boskou (41), the total
phytosterol content of WGO is about 34%. Esterified sterols constitute 23% of
the oil. The free to esterified phytosterol ratio in the WGO varies between 0.3
and 0.5.
2.5.3.4. Other Unaponifiable Compounds Cycloartenol, b-amyrin, and 24-

methylene cycloartanol are the major triterpenols, which constitute less than 1%
of WGO. Hydrocarbons are minor components in the WGO unsaponifiable fraction
(Table 7). According to Kuksis (42), 50% of the hydrocarbons was squalene and
the reminder consisted of n-C29 alkanes. The presence of lutein and cryptoxanthin
in WGO were first reported in 1935 and 1940, respectively (43, 44). Recently,
Panfili et al. (45), reported that petroleum-ether-extracted WGO contained
25 ppm lutein, 23 ppm zeaxanthin, and 8 ppm b-carotene.
2.6. Oil Production

Wheat germ oil can be extracted by mechanical expelling, organic solvent extrac-
tion, and supercritical fluid extraction. Mechanical expression and organic solvent
extraction are both being used for commercial extraction of wheat germ oil. To our
knowledge, supercritical fluid technology has not been commercialized for WGO
processing.
Hexane is commonly used for WGO extraction (46). Ethanol and 1,2-dichlor-
oethane are also used to a lesser extent for commercial WGO extraction (1).
Although hot solvents are preferred for vegetable oil extraction, it has been reported
that at least one company performs solvent extraction at temperatures around 38 C
for WGO recovery (1).
Solvent extraction is more efficient than mechanical pressing of wheat germ oil.
The residual oil content of solvent-defatted wheat germ can be as low as 1%, (w/w).
Solvent-extracted wheat germ is more stable than the mechanically expressed
wheat germ because of its lower lipid content. Pressing recovers only 50%,
(w/w) of the wheat germ oil, and residual wheat germ requires further stabilization
to avoid rancidity and shelf life extension. Mechanical pressing of wheat germ oil
can be successful only when the bran contamination is minimized during the
milling operation because the oil content of bran is much lower than that of the
germ fraction. Pressed wheat germ is perceived as natural and usually preferred
by consumers.
Barnes (2) reviewed the literature on wheat germ extract yields using various
solvents. Hexane and light petroleum ether extraction resulted in yields ranging
from 515% (w/w). This wide range of variation in oil yield was explained by
the degree of contamination of germ by bran, which contains only 5% oil. Diethyl
WHEAT GERM 205
ether, which is expected to extract more polar components, yielded 715% oil.
Acetone gave 1.9% more oil yield than light petroleum ether did.
Although most edible oils are refined to remove PL, free fatty acids (FFA), color
compounds, and volatile components, WGO is often used in the crude form. Phos-
phatidylcholine, color, and flavor are desired attributes for the products marketed as
natural in the health food stores. However, refining improves the stability of the
oil. The FFA content of crude WGO can be high, 525%, depending on the germ
separation conditions, storage, and oil-extraction method. Free fatty acids contri-
bute to bitter and soapy flavors in the product; hence, they are removed from
WGO by alkali treatment. However, the alkali deacidification process results in sig-
nificant losses in oil and more importantly in tocopherols. Wang and Johnson (30)
examined the effect of conventional oil-refining processes on the WGO quality.
According to this study, tocopherol content of WGO did not change significantly
during degumming, neutralization, and bleaching processes. However, deodoriza-
tion conditions reduced the tocopherol content of WGO significantly. Lower tem-
perature and longer residence time were effective in reducing FFA, peroxide value,
and color while retaining tocopherols in WGO during deodorization. Although
degumming did not reduce phosphorous content of the crude oil effectively, phos-
phorous concentration was reduced at every stage of WGO refining. Wang and
Johnson (30) have suggested that WGO refining should include acid degumming
at high temperatures and high shear for an extended time, as compared with that
for the typical vegetable oils, to maximize PL hydration. Although PL, specifically
PC, has beneficial health effects for humans, they are removed from the crude oil
during the degumming process. Phospholipids tend to precipitate out in the oil dur-
ing storage and have adverse effects on frying operations due to their emulsification
properties. Neutralization of FFA may need excess alkali treatment. Wheat germ
oil bleaching requires more bleaching earth than that of the typical vegetable oil
refining.
An expired U.S. patent describes molecular distillation of WGO (47). Initially,
WGO was degummed by using phosphoric acid and water. Bleaching was carried
out with activated clay followed by distillation using a centrifugal molecular distil-
lation unit. Free fatty acids were removed at 140200 C and below 50 mTorr. It was
claimed in the same patent that a Vitamin E concentrate was prepared from purified
WGO by a second-stage molecular distillation process carried out at 220300 C
and pressures less than 25 mTorr (47).
Supercritical fluid-extraction technology is an alternative method to conven-
tional hexane extraction. Supercritical fluid extraction of WGO has been reported
by several research groups (45, 4850). Wheat germ oil solubility in SC-CO2 at
40 C and 200 bar was 0.35% (w/w) (48). Oil extracted with SC-CO2 has a lighter
color and contains less phosphorus than that of the hexane-extracted oil. Although
oil extraction rates from the ground and flaked wheat germ were not significantly
different, use of flaked wheat germ is recommended for large-scale SC-CO2 extrac-
tion. Ground wheat germ particles can be difficult to handle because of dusting.
Furthermore, channeling of SC-CO2 flow through the ground wheat germ in
the extraction vessel because of compaction may reduce the mass transfer.
According to Dunford and Martinez (50) and Taniguchi et al. (48), the a- and
b-tocopherols content of SC-CO2-extracted oil were similar to those of hexane-
extracted oil. However, Gomez and Ossa (49) reported higher tocopherol content
in the SC-CO2-extracted WGO as compared with that of the hexane-extracted
oil.
Panfili et al. (45) characterized the composition of SC-CO2-extracted WGO and
defatted cake. According to this study, FFA content and PV of the oils collected
during the initial stages of SC-CO2 extraction (during the first 45 min) were higher
than that of the oil fractions collected at the later stages of the process. Similarly,
more tocopherols were detected in the oils collected during the first 75 min of 3 h
SC-CO2 extraction. Experiments also indicated that WGO collected during the
initial stages of SC-CO2 extraction had a higher tocopherol content (50). The
most abundant carotenoid in SC-CO2-extracted WGO was lutein, followed by
zeaxanthin and b-carotene. A larger amount of carotenoids was extracted toward
the end of SC-CO2 extraction (45).
Studies carried out with liquid and SC-CO2 (50400 bar) at relatively low tem-
peratures (1060 C) indicated that pressure had a significant effect on the oil yields,
whereas the effect of temperature was insignificant (48). Hence, the effect of pres-
sure and temperature on the SC-CO2-extraction yields and WGO composition was
studied in the range of 100550 bar and 4080 C (50). Yields of SC-CO2 extracts
[(weight loss from the sample during the extraction/initial weight of wheat germ
used for extraction) 100] varied significantly with temperature and pressure in
the 2% to 20% (w/w) range. The wheat germ oil yield was 11% (w/w) when hot
hexane (Soxhlet) was used for extraction. The higher SC-CO2-extraction yield
(>11%) indicates that SC-CO2 at high pressures extracted some of the wheat
germ components, which are not soluble in hexane. Moisture in the wheat germ
might be one of the compounds coextracted with oil resulting in higher extraction
yields. The highest SC-CO2-extraction yield was obtained at the highest pressure
used (550 bar). The temperature dependence of the extract yield was more pro-
nounced at higher temperatures (60 C and 80 C) and the lowest pressure examined
in this study (100 bar). This is a result of the significant change in SC-CO2 density
under those conditions.
The fatty acid composition of the extracts was not affected by temperature,
pressure, and the extraction method (Table 4). Supercritical carbon-dioxide-
extracted oil samples had similar fatty acid composition to that of the Soxhlet-
extracted oil (Table 4). All of the wheat germ extracts consisted of about 56%
linoleic acid (18:2 n-6), which is an essential fatty acid (Table 4). The total unsa-
turated and polyunsaturated fatty acid (PUFA) content of the wheat germ oil was
about 81% and 64%, respectively. The SC-CO2 extraction of wheat germ resulted in
extracts with similar tocopherol and tocotrienol compositions to those of the Soxh-
let extracts (Table 8) (50). These results indicate that SC-CO2 technology can be
used for extraction and fractionation of WGO components to obtain products
with high quality.
CORN GERM OIL 207
TABLE 9. Corn (Maize) Production.1
Production (x106 Metric Tons)
Year 2001 2002 2003

USA 242 229 257
India 13 11 15
Canada 8 9 10
World 615 604 638
1
FAOSTAT Database (http://apps.fao.org) (11).
3. CORN GERM OIL

Corn, also referred to as maize (Zea mays L.), is one of the most important cereal
grains that is a commercial source of vegetable oil. Maize originated in the
Americas and was introduced into Europe in 1492. About 600 million tons of maize
is produced annually worldwide (Table 9) (11). The United States is by far the
largest corn producer in the world followed by Mainland China.
Corn grown in the United States contains about 65% starch and 34% oil (5).
The corn germ contains 85% of the oil and 80% of the minerals present in the whole
grain (6). A large fraction of the maize crop is used as livestock feed. However, in
many developing countries, corn is used only for human consumption. Maize is also
used for numerous industrial products, in distillation and fermentation industries,
and for the production of starch and corn syrup. Corn oil is a byproduct of the
corn milling industry and accounts for a relatively small portion of the economic
value of the whole plant. Corn germ contains 5060% oil. Corn oil is used as salad
and cooking oil and in margarines.

Like the grain of other cereals, corn grain is a caryopsis. The maize grain is a one-
seeded fruit and fairly large with an average mass of 285 mg (51). The most impor-
tant variety of maize is Dent Maize or Indian Corn (Zea mays var. Indentata
Bailey). Dent refers to both tooth-like structure and indentation at the end of
the grain (Figure 2). Structure of a corn grain is quite similar to any other cereal
grain. The endosperm consists of an aleurone layer and horny and floury
endosperm. The embryo or germ consists of scutellum and embryonic axis. The
plumule, mesocotyl, and the radicle form the embryonic axis. In corn, the germ
is located inside the endosperm and a cap that is visible externally protects its
end.
Figure 2. Structure of corn grain (21).
3.3. Germ Separation

Maize milling process has been reviewed by several authors (5, 51, 52). Maize
milling can be done by semiwet (2023% moisture), wet (45% moisture), or dry
(17% moisture) milling operations. Semiwet and dry milling operations require
a tempering step, which refers to controlled addition of moisture to the grain before
the dehulling and degerming operation. The dry milling process is very similar to
the wheat degermination process described earlier in this chapter. The recovered
germ contains about 2025% oil.
The wet milling operation entails a steeping step that involves soaking of
cleaned corn in water containing 0.10.2% sulfur dioxide at 52 C (5). Then, the
whole grain is ground in roller mills. The ground product contains all components
of the grain; bran, hulls, starch, gluten, and germ. Various components of ground
corn are separated by screening, centrifugation/floatation, and filtration operations.
The germ fraction is separated from the other ground grain components either by
CORN GERM OIL 209
centrifugation or floatation operations. Centrifugal separation has replaced the

flotation method because it provides improved separation efficiency, cleaner
germ, higher operation capacity, and more sanitary conditions. Overflow from the
centrifugation process carries lighter weight germ and the underflow contains
heavier components such as starch, gluten, and hulls. The germ is then washed,
dewatered, and dried. The objective of the degerming operation is to obtain grits
with 1% fatty matter. The wet corn milling operation produces germ with
5060% oil.
3.4. Oil Content

An examination of the data on oil content of corn grown between the years 1917
and 1972 showed fairly small variations (4.0% to 4.9%) (53). Since then, plant
breeders and genetic engineers altered the starch, protein, and oil content of corn
significantly. In 1982, after 82 generation of selection for high and low oil content
corn, Illinois High and Low Oil varieties, which contained 19% and 0.3% oil,
respectively, were developed (6). A disadvantage of Illinois High Oil Corn was
its lower yield than its traditional counterparts. More effective breeding later
resulted in 68% oil and yields similar to those of a traditional corn crop. Waxy,
high amylose, high oil, and high lysine corn have been created to meet the needs
of livestock feeders, the food industry, and other industrial users of corn. High oil
corn contains 78% oil, which is a 23% increase over traditional corn. High amy-
lose corn, which was developed to meet the needs of the wet milling industry,
contains higher than 50% amylose. This variety also contains a higher amount of
oil. Germ of this variety is larger and contains higher quality protein than that of the
traditional varieties. High lysine corn contains higher levels of essential amino
acids lysine and tryptophane. These varieties are not grown as widely as traditional
dent type corn.
There are several agronomic factors, such as drought, leaf blight, and fertilizer
application, that affect oil content in the crop (5456). For example, fertilizer appli-
cation was shown to improve oil content of the corn grain only slightly. However,
an increase in the total grain yield was significant, consequently increasing the oil
produced per acre (57). Drought also increased protein content of the grain while
decreasing oil content (54). An epidemic of southern maize leaf blight did not affect
the starch and protein content of the grain, but the oil content was decreased
significantly (55).
Germ is particularly susceptible to mechanical damage, which causes oil dete-
rioration by enzymatically catalyzed fatty acid oxidation and TAG hydrolysis. Fun-
gi are the main factor responsible for the production of FFA in stored grains (6).
Damaged pericarp makes the grain susceptible to fungal invasion. Fungi grow
preferentially on the germ causing oil depletion. High drying temperatures have
been linked to difficult and incomplete grinding, poor germ separation leading to
inefficient oil recovery, and high FFA content in the oil (56, 58). When maize is
harvested at a moisture content of over 25%, it should be dried at temperatures
below 60 C to avoid adverse affects on the oil (59).
TABLE 10. Physicochemical Properties of Maize

Germ Oil.1
Property
Specific Gravity 0.9170.925 (20 /20 C)

1.4651.468 (40 C)
Iodine Value 107135
1

Physicochemical properties of maize oil are summarized in Table 10. Specific
gravity of maize oil varies between 0.917 and 0.925 at ambient temperature
(20 C/20 C). Unsaponifiable content of corn oil is lower (13%) than that of the
WGO.
3.5.1. Fatty Acid Composition Corn oil fatty acid composition can vary
depending on the seed type and the climatic conditions. Oil obtained from maize
grown in the northern hemisphere had higher iodine value than that of the maize oil
grown in the southern hemisphere. U.S. grown corn had higher linoleic acid and
lower oleic acid content as compared with South African corn (Table 11) (60).
The fatty acid compositions of various grain fractions also show differences
(Table 12). Corn germ oil contains more linoleic and oleic acids and a lower
amount of palmitic acid than those of the endosperm oil (61). There is a wide varia-
bility in fatty acid composition of maize oil. It was found that oleic and linoleic
acids content of 788 maize varieties ranged from 14% to 64% and from 19% to
TABLE 11. Fatty Acid Composition of Maize Oil.1
Fatty Acid South Africa (%) U.S.A. (%)
12:0 0.4 0.1

14:0 0.2 0.2
16:0 11.5 11.0
16:1 0.1
18:0 2.0 2.0
18:1 38.7 24.1
18:2 44.3 61.9
18:3 1.1 0.7
20:0 0.6
22:0 0.1 1.7
22.1 0.3
24.0 0.3
1
Adapted from Leibovitz and Ruckenstein (60).
CORN GERM OIL 211
TABLE 12. Fatty Acid Composition of Corn Grain Fractions.1
Grain Fraction Fatty Acid Composition (%, w/w)
16:0 18:0 18:1 18:2 18:3

Germ 11 1 24 63 1
Endosperm
(Non-starch lipids aleurone lipids) 17 2 19 58 4
1
Adapted from Tan and Morrison (61).
TABLE 13. Fatty Acid Composition of Different Corn Varieties.1
Fatty Acid High Oleic Maize Low Saturate
12:0 00.3
14:0 00.3
16:0 1016 9.216.5 68
16:1 00.4
18:0 2 03.3 1
18:1 4464 2042.2 2531
18:2 2038 39.465.6 5864
18:3 0.81.0 0.51.5 0.80.9
20:0 1 0.30.7 0.5
20:1 00.4
20:2 00.1
22:0 00.5
22:1 00.1
24:0 00.4
1
Adapted from Weber (64).
71%, respectively (62). Polyunsaturation of maize oil fatty acids increased 58%
over the past 20 years (63). Weber discussed the effect of breeding and genetic
modifications on the fatty acid composition of maize (6, 64). Maize, high oleic,
and low saturate corn oil fatty acid compositions are shown in Table 13 to illustrate
the potential of biotechnology for modification of oil composition in plants.
3.5.2. Acyl Lipids Acyl lipid composition of maize oil is shown in Table 14.
Triacylglycerols are the main components of maize oil (about 90%). Traditional
TABLE 14. Acyl Lipid Composition of Maize Oil.1
Lipid Type % of Total Acyl Lipid
Sterol Fatty Esters 2

Triacylglycerols 89
Free Fatty Acids 1
Glycolipids 2
Phospholipids 4
1
Adapted from Weber (63).
TABLE 15. Unsaponifiable Composition

of Commercial Corn Oils.1
Mazola Kroger
Hydrocarbons 0.039 0.039

Sterols 1.42 0.95
Sterol esters 0.37 0.32
1
Adapted from Worthington (67).
maize germ oil contains approximately 58% triunsaturated followed by about 40%
diunsaturated TAGs (60). The majority of the TAG species were PLL (palmitic-
linoleic-linoleic) (20%), POL (palmitic-oleic-linoleic) (15%), LLL (linoleic-lino-
leic-linoleic) (26%), OLL (oleic-linoleic-linoleic) (27%), and OOL (oleic-oleic-
linoleic (14%) (65).
Although FFA content of the oil varies considerably depending on the preharvest
and postharvest conditions, it is usually in the range of 11.5% of the total acyl
lipids. Wet milled corn germ oil contains higher FFA content than that of the dry
milled oil, 1.54% and about 2%, respectively (7, 66). The PL content of corn germ
oil varies with the extraction process. Expelled oil contains about 120 ppm PL as
compared with 670 ppm in hexane prepress commercial products (66).
3.5.3. Unsaponifiable Content The unsaponifiable content of maize oil ranges

from 1% to 3% of the crude oil. Tocopherols and sterols are the main components
of the corn oil unsaponifiables (Table 15). Worthington (67) examined the sterol
and hydrocarbon content of two commercial corn oils (Table 14). Squalene was
the major hydrocarbon found in the corn oil unsaponifiable fraction (3560%).
Weber (68) reported that 24% of the carotenoids present in corn kernel are loca-
lized in the germ. Lutein and zeaxanthin are the most abundant carotenoids in corn
kernels. The levels of carotenoids in corn germ oil are relatively low as a result of
low concentrations in the germ and losses during the oil bleaching process.
3.5.3.1. Tocols The tocol content of maize oil varies from 26 ppm to 102 ppm of
grain (6). Traditionally, g-tocopherol has been considered the predominant tocol in
maize oil. However, some varieties contain a-tocopherol as a main isomer. Weber
(6) reported that g-tocopherol content of maize inbreds ranged from 36% to 88%
(as a percent of total tocols), whereas a-tocopherol content varied from 2.1% to
45%. Milling is known to have a significant effect on the tocopherol content of
corn germ oil. Oil from wet milled corn germ contained only 18% of the tocopherol
present in whole grain; however, 73% of the tocopherols was recovered from the
dry milled corn germ (69).
3.5.3.2. Sterols The total phytosterol content (free and esterified phytosterols) of
corn germ oil is higher than in most of the other vegetable oils (3, 70). Corn germ
oil contains about 1.1% (w/w) phytosterol esters (70). Sitosterol is the major sterol
CORN GERM OIL 213
TABLE 16. Effect of Processing on Phytosterol Content of Corn Germ Oil.1
(mg/g)

Esterified Sterols Free Sterols Total Sterols
Crude corn oil 548 439 976

Physical refining
Degummed 480 448 910
Bleached 460 461 898
Deodorized 491 257 743
Chemical refining
Degummed 461 488 930
Neutralized 473 390 869
Bleached 447 388 845
Deodorized 455 336 793
1
Adapted from Verleyen et al. (72).
in corn germ oil. Sitosterol, campesterol, stigmasterol, and 5-avenasterol consist

of 9095% of all corn oil sterols (6, 71). Itoh et al. identified a new sterol,
24-methyl-E-23-dehydrolophenol, in maize germ oil. The effect of corn oil refining
process on the free, esterified, and total sterols is shown in Table 16 (72).
3.6. Oil Processing

Corn germ oil processing is similar to that for other vegetable oils. Dry germ is
initially subjected to high pressure in an expeller that breaks down the cell structure
and releases the oil. The expelling operation reduces the oil content of the corn
from about 50% to 15%. The residue germ cake is flaked then hexane extracted.
These processes reduce the oil content of the germ to less than 1.5%. The meal
(defatted germ) is used as animal feed. Solvent is removed from the miscella by
distillation. The oil is then filtered to remove suspended solid material. Recently,
extrusion of corn germ has been employed as a preparation step for solvent extrac-
tion to produce a crude corn oil of high quality and high yield (73). Germ quality
significantly affects the oil recovery. As an example, blight infection decreased oil
recovery from 53% in the uninfected corn to less than 30% in infected corn (74).
Also, FFA content of blight-infected crude oil was very high, about 12%, which
caused higher refining losses.
According to Mounts and Anderson (5), a few processors use water degumming,
otherwise PL are removed during alkali refining. The conventional caustic soda
refining has been used by the majority of refiners in the United States for neutrali-
zation of crude corn oil. Physical or steam refining can also be used for crude corn
oil neutralization. Physical refining is recommended only for high-quality oil,
otherwise the oil becomes dark during the physical refining process. A degumming
process should precede physical refining. Free fatty acids can also be removed by
liquid-liquid extraction or by a new method, which involves solvent extraction in a
perforated rotating disc column (75). Undesirable odor and flavor components of
TABLE 17. Effect of Processing on the Tocopherol, Phosphorous (P), Iron (Fe), and FFA
content of Corn Germ Oil.1
Tocopherols (ppm)
a g d Total P (ppm) Fe (ppm) FFA (%)
Crude corn oil 266 1009 54 1329 110 0.9 1.3

Physical refining
Degummed 261 968 52 1281 1 <0.1 1
Bleached 262 963 50 1275 0.3 <0.1 1
Deodorized 142 506 23 671 0.3 <0.1 0.08
Chemical refining
Degummed 219 842 47 1108 7.5 0.2 0.3
Neutralized 217 811 45 1073 0.6 <0.1 0.2
Bleached 198 798 42 1039 0.1 <0.1 0.1
Deodorized 83 458 19 560 0.1 <0.1 0.09
1
Adapted from Verleyen et al. (72).
corn oil are removed during the deodorization process at temperatures above 200 C
and under 210 mm Hg vacuum. The deodorization process removes a significant
amount of tocopherol (Table 17) and phytosterols (Table 16) from the oil (3). Pig-
ments are usually removed by treating the oil with acid-activated bleaching clay (4).
Dewaxing or winterization is usually carried out at 510 C. Wax precipitate that
forms during the winterization process is removed by filtration (60). Effect of refin-
ing process on the tocol, FFA, iron, and phosphorous content is summarized in
Table 17.
A process to extract whole flaked corn using ethanol was reported in 1992 (76,
77). This sequential extraction process (SEP) involved the following steps: (1)
extraction of crude oil and removal of water from ethanol and (2) extraction of
food-grade protein using an alkali-alcohol mixture. Corn was extracted with 95%
ethanol at 76 C in a countercurrent mode while simultaneously dehydrating the
ethanol. A small amount of zein was coextracted along with oil (76). The oil
extracted using the above explained SEP contained a higher concentration of
FFA, DAG, PL, and carotenoids and a smaller amount of TAG and had darker
red color than the hexane-extracted corn oil. A modified SEP used a 30% hexane
and 70% ethanol mixture at 56 C for corn extraction (77). The modified SEP
resulted in products with smaller amounts of FFA, DAG, and PL and larger concen-
tration of TAG and carotenoids than the original SEP oil. However, refining losses
for SEP oil was higher than that of the hexane-extracted oil. A U.S. patent also
describes recovery of oil and zein by ethanol extraction of dry milled corn followed
by a membrane process to fractionate oil and zein (78).
Karlovic et al. (79) examined the aqueous enzymatic extraction of corn germ oil.
Hydrothermal pretreatment, grinding, and enzymatic treatment of corn germ
improved extraction efficiency. Although the energy cost for enzymatic corn
germ oil extraction was lower than that of the conventional extraction, the enzyme
cost made the process more expensive (79).
RICE BRAN OIL 215
Supercritical fluid extraction of corn germ oil has been also reported (66, 80, 81).
Refining losses and FFA content of dry milled corn germ oil extracted with super-
critical carbon dioxide (SC-CO2) at 50008000 psi and 50 C oil was lower and it
had a lighter color than those of the commercial expeller-milled crude oil (80).
Total unsaponifiable and tocopherol contents were similar for both oils. Wet and
dry milled corn germs were also extracted with SC-CO2 at higher temperatures
and pressures 5090 C and 800012000 psi (66). Experimental data indicated
that increasing extraction temperature and pressure did not adversely affect the
oil quality, i.e., FFA, color, phosphorus refining loss, and unsaponifiable matter con-
tent of the extracts did not increase significantly. Although tocopherol content of
dry milled corn oil decreased with increasing SC-CO2 temperature, it was similar
to that of the expeller oil. Supercritical carbon-dioxide-extracted wet milled corn oil
contained less tocopherol than commercial prepress hexane-extracted oil. Unsapo-
nifiable content of SC-CO2 extracted wet and dry milled corn germ oil was similar
to that of the commercial corn oil, 1.21.4%. The SC-CO2 extraction process yields
more TAG in the oil (98.499%) than commercially extracted corn germ oil (95.8%
and 97.2% for wet and dry milled corn germ oil, respectively) as a result of very
low PL content of SC-CO2-extracted corn germ oil (15 ppm) (66). It was also
reported that SC-CO2-extracted wet milled corn germ oil had better flavor profile
than commercially extracted corn germ oil.
4. RICE BRAN OIL

China is the largest producer of rice grain, followed by India, Japan, and the United
States. About 600 million metric tons of rice is grown worldwide annually
(Table 18). The main consumers of rice bran oil (RBO) are Asian countries, such
as Japan, Korea, China, Taiwan, Thailand, Pakistan, and India (82). In the United
States RBO production started in the 1950s but was discontinued in 1980s because
it appeared uneconomic (8). Interest in RBO was renewed in 1990s due to export
opportunities and its nutritional benefits. Rice bran oil has been commercially
produced for food use in the United States since 1994.
TABLE 18. Rice (Paddy) Production.1
Production
(x106 Metric Tons)
Year 2001 2002 2003

India 140 108 132
Japan 11 11 10
USA 10 10 9
World 598 570 589
1
FAOSTAT Database (11).
Rice bran oil has several unique properties that make it very appealing as a spe-
cialty oil in niche markets. It has a nut-like flavor and is quite stable after extrac-
tion. Rice bran oil contains high levels of bioactive components, such as
phytosterols, tocopherols, and tocotrienols, which have nutraceutical value.
Although RBO can be used virtually in any application to replace other vegetable
oils, it is well suited for use where both functionality and health benefits are impor-
tant. For example, RBO exhibits excellent frying performance because of its good
storage stability and fry life and contributes a pleasant flavor to the fried foods. Rice
bran oil is also used in margarines. Its natural tendency to form stable b0 crystals
and its high palmitic acid content results in a good balance of plasticity, creaminess,
and spreading properties. Rice bran oil processed to retain high levels of tocols may
be used as a natural antioxidant source and can be used as coatings for a wide range
of products and snacks such as crackers and nuts. Rice bran oil can be blended with
other oils to improve their stability (83). Rice bran oil is an attractive ingredient,
which can be incorporated into functional foods and nutraceuticals to provide
health benefits. Industrial uses of RBO include an additive to animal feed, glycer-
ine, and soap production. Rice wax can be used in confectionery products,
cosmetics, shoe polish, and auto wax (84).

Rice grain structure is similar to that of most of the other cereal grains. The grain
consists of the edible portion, caryopsis or brown rice, and the hull (Figure 3).
Removal of the hull from the rice grain by dehulling exposes the caryopsis. Peri-
carp, seed coat, nucellus, and embryo comprise the bran portion of the rice grain.
The bran portion is 58% of the brown rice weight (85). The embryo alone is 12%
of the weight of brown rice. Commercial rice germ includes the outer layers enclos-
ing the embryo and represents about 3% of the grain. Commercial bran usually
includes the germ. Thus, commercial RBO contains both bran and germ oil. The
lipids are present in the grain in the form of lipid bodies or spherosomes (86).
The lipid bodies are very small, 0.7 mm in all tissues of the embryo. The lipid
content of rice germ is higher (3437%) than the bran fraction (1926%).
4.3. Grain Processing

The purpose of rice milling is to produce an edible, polished product (white rice)
from rough rice. Rice bran is a byproduct of rice milling industry. Huller type rice
mills, where dehulling and milling is a single processing step, produces rice mill
feed, whereas raw rice bran is produced during a process (cone-type mills)
where dehulling is a separate process, which separates germ and bran. Germ con-
tains about four times more oil than the rice mill feed (4%) (10). The total amount
of rice bran available from rice milling operations throughout the world is estimated
to be 30 million tons (10).
The rough rice or paddy undergoes a precleaning process prior to milling. For-
eign materials, such as stones, mud balls, straw, and weed, are separated with the
use of screens, aspirators, and gravity separators. Cleaned rough rice is then
RICE BRAN OIL 217
Figure 3. Structure of rice grain (85).
dehulled. Most modern rice mills use rubber roll shellers to remove the hull from
the kernel so that brown rice is obtained (87). Unshelled kernels are separated by
density from the brown rice stream in the paddy separators prior to bran separation.
In the United States, rice bran consists of pericarp, aleurone, subaleurone layer,
germ or embryo, and a small amount of endosperm. The germ can be removed
from the kernel as a relatively intact particle and typically large enough to be
retained on a U.S. 18 mesh screen (87). However, germ separation from the bran
is not a common practice in the United States Bran is removed from the milling
chamber by air suction and collected in cyclone separators and secondary cloth fil-
ters. Then bran is screened through 1618 mesh to remove fine particles. The
amount of lipids present in the bran is inversely proportional to the degree of
milling.
Ideally, bran should be stabilized within a few minutes after removal from the
kernel. Stabilization process inactivates enzyme lipase that causes rapid hydrolysis
of TAG. Three methods developed for brown rice stabilization are: (1) heat dena-
turation and inactivation of lipases, (2) extraction with an organic solvent to remove
TABLE 19. Rice Bran Proximate Composition.1
Compound % (w/w)
Protein 15
Oil 18
Carbohydrates 50
Ash 7
Crude fiber 7
Total dietary fiber
Soluble 2
Insoluble 26
1
Adopted from Orthoefer (8).
oil that serves as a substrate for lipase, and (3) ethanolic denaturation and inactiva-
tion of lipases and lipase-producing bacteria and mold. The rice bran stabilization
process has been discussed in detail by Champagne (88).
4.4. Oil Content

A typical proximate analysis for rice bran is shown in Table 19. Rice bran typically
contains 1632% oil. However, the oil content of the rice kernel and bran might
vary for several reasons. An early or late cropping season, crop variety, and degree
of milling are some of the factors affecting oil content of rice kernel and the bran
(9). Rao et al. (89) examined the oil content of rice bran from four varieties of
brown rice grown in India and polished to different degrees (110% of its weight
as bran). The oil content of bran did not change significantly up to 6% polishing,
after which there was a gradual decrease. This phenomenon was explained by the
presence of higher amounts of endosperm and inner aleurone material in the bran
after more polishing. Recently, 204 genetically diverse rice cultivars were evaluated
for oil content in the bran fraction (90). Genotype and environment had a significant
effect on the lipid content of rice bran. The variations suggest that breeding material
is available for modifying oil content and composition in rice bran (90).

The typical compositions of crude rice bran and germ are shown in Table 20. Bran
and germ oil compositions are very similar. Crude rice oil tends to contain higher
levels of non-TAG components, which causes higher oil refining losses than for
other vegetable oils. Physical properties of RBO are summarized in Table 21. In
the United States, characteristics of quality RBO are described as: maximum
0.1% FFA, maximum peroxide value 1 meq/kg, 0.05% moisture, iodine value
95110, saponification value 180195, and Lovibond color value of 3.5R (91).
Japanese standards for refined RBO are similar to those in the United States: cloud
point, 15 C; unsaponifiable matter, 5%; saponification value, 180195; iodine
RICE BRAN OIL 219
TABLE 21. Physicochemical Properties of Rice Bran

Oil.1
Property
Specific Gravity 0.9160.921 (25 C/25 C)

1.4651.468 (40 C)
Iodine Value 92108
Smoke Point 231 C
Fire Point 352 C
Cloud Index 17 C
1
value, 92115; specific gravity (25 C/25 C), 0.9130919; refractive index, 1.470
1.473; and cold test at 0 C for 1 h (9).
4.5.1. Acyl Lipids The lipids of rice bran and germ are quite similar and mainly
consist of TAG, FFA, ASG, PE, and PC (Table 22). TAGs are the main components
of the RBO (Table 23). The monoacylglycerol content of RBO is higher than in
other vegetable oils (67%). The wax content of RBO varies significantly with
the variety. Waxy varieties can produce an RBO that contains up to 8% wax.
The amount of wax in the RBO can be reduced to 0.5% by adjusting the extraction
conditions (i.e., low-temperature extraction) (92). Rice bran oil waxes are esters
of C16C26 fatty acids and saturated C24C30 fatty alcohols. Rice waxes are
TABLE 22. Lipid Composition of Rice Germ and Bran.1
Lipid composition (%) Germ Bran
NL 9192 8890
GL 23 45
PL 67 78
Lipid classes (% of total)
TAG 7779 7576
FFA 4 45
ASG 1 2
SG <1 <1
PE 34 3
PC 34 34
LPE <1 <1
Others 89 89
1
Neutral lipids (NL), Glycolipids (GL), Phospholipids (PL), Triacylgly-
cerols (TAG), Free fatty acids (FFA), Acylsterylglycosides (ASG),
Sterylglycosides (SG), Phosphatidylethanolamine (PE), Phosphatidyl-
choline (PC), Lysophosphatidylethanolamine (LPE). Adapted from
Juliano (9).
TABLE 23. Lipids Classes of Rice Bran Oil.1
Lipid Class Composition (%)
Triacylglicerols 8889
Diacylglicerols 34
Monoacylglycerols 67
Free Fatty Acids 24
Waxes 34
Glycolipids 67
Phospholipids 45
1
Adapted from Juliano (9).
classified as soft and hard waxes based on their melting points, 75 C and 80 C,
respectively. Two thirds of the waxes are present in polymeric form with the
remainder in monomeric form (92). Waxes are separated from the crude oil during
the refining process as a result of their adverse effects during frying and opaque
appearance in the oil.
4.5.2. Fatty Acid Composition The fatty acid content of RBO is mainly palmi-
tic, oleic, and linoleic acid (Table 24). The low linolenic acid content of RBO
makes it stable to oxidation. Several studies reported variations in fatty acid com-
position of RBO (90, 9596). Goffman et al. (90) studied the fatty acid composition
of 204 rice varieties. Genotype and environment significantly affected stearic, oleic,
linoleic, and linolenic acids but not palmitic acid content of the RBO. The ratio of
saturated to unsaturated acid ratio (S/U) was correlated to the palmitic acid content
of the oil. Japonica lines had low palmitic acid content and S/U ratio, whereas
Indica lines were characterized by high palmitic acid content and high S/U ratio
(90).
TABLE 24. Fatty Acid Composition of Rice

Bran Oil.1
Fatty Acid (%)
14:0 0.50.7
16:0 1628
16:1 0.5
18:0 2 4
18:1 38 48
18:2 1636
18:3 0.22.2
20:0 0.50.8
20:1 0.30.5
22:0 0.10.5
24:0 00.5
1
Adapted from Juliano (9).
RICE BRAN OIL 221
TABLE 25. Unsaponifiable Composition

of Rice Bran Oil.1
Component %
Hydrocarbons 18
Phytosterols 43
Sterol esters 10
Triterpene alcohols 28
Tocopherols 1
1
Adopted from Orthoefer (8).
4.5.3. Unsaponifiables Rice oil has a significantly higher concentration of

unsaponifiables compared with other vegetable oils (Table 25). The bioactive
components of rice bran are associated with the oil fraction (97) and concentrated
in the unsaponifiable fraction of the oil (8).
4.5.3.1. Tocols Along with the protein and edible oil, rice bran is also a rich
source of tocopherols (500 ppm-4% bran oil). About 30% of the total tocopherols
in RBO are a-tocopherols. Over 95% of the total tocopherols is contained in the
rice germ. The tocopherol content of commercially available rice bran varies
significantly (98). In general, increasing degree of brown rice milling resulted in
higher tocopherol content in the bran (98).
Rice bran oil and palm oil are the only readily available oils that contain signif-
icant levels (about 1000 ppm) tocotrienols (99). Tocotrienols belong to the Vitamin
E family and have similar chemical structures. According to Tomeo et al. (100),
tocotrienols are powerful antioxidants. Commercially available RBO may contain
980 ppm g-tocotrienol (101).
4.5.3.2. Sterols The unsaponifiable content of RBO is mainly phytosterols

(Table 25). These include free sterols, sterol esters, sterylglycosides, and acylster-
ylglycosides. b-Sitosterol is the main sterol present in RBO. Other sterols present in
the RBO include 4-desmethylsterol, 4-monomethylsterol, and 4,4-dimethylsterol
(10). Crude RBO contains about 1.5% oryzanol (102). Oryzanol, a major compo-
nent of rice unsaponifiables, is a group of compounds containing ferulate (4-hydro-
xy-3-methoxycinnamic acid) esters of triterpene alcohols and plant sterols.
Cycloartenol, 24-methylene cycloartenol, campesterol, and b-sitosterol are present
as ferulate esters in the RBO. Oryzanol was reported to contribute to the hypocho-
lesterolemic activity of RBO in rats (103) and hamsters (104). Studies carried out
with human subjects also support hypocholesterolemic activity of RBO (105). The
oryzanol content of the RBO varies with the technique used for oil refining. Crude
oil contains 2% oryzanol. The degumming process reduces the oryzanol content
to 1.7%. Physically refined oil contains 1.01.5% and alkaline refined oil has 0.1%
oryzanol (8).
4.6. Oil Production

A major problem with rice bran oil extraction is the high lipase activity, which
results in FFA formation within a few days of milling particularly at high tempera-
ture and humidity. Free fatty acid content in rice bran increases during storage, i.e.,
24% in a fresh crop, 58% in 1-year grain, and >10% in a 2-year-old crop (105).
Thus, lipase is inactivated to stabilize rice bran prior to oil extraction (88). Heat-
stabilized bran may be stored up to three months. However, oil extraction should be
carried out within the first month to obtain better efficiency and higher quality oil.
Mechanical expression of rice bran yields less oil, 1012%, than solvent extrac-
tion, 1618%. Rice bran is treated with steam and dried prior to pressure expres-
sion. Prepressing is usually carried out at 70 kg/cm2 followed by oil expulsion at
105316 kg/cm2 (9). As a result of the low yield of oil from mechanical extraction,
residual oil in the bran is recovered with hexane. Hexane extraction can be per-
formed by batch or a continuous operation. Continuous operation uses countercur-
rent flow to improve mass transfer. Solvent extraction at high temperatures results
in higher crude oil yield, but the crude oil is of lower quality. A new oil-extraction
process, which involves premolding of rice bran at 14% moisture content and
40 C followed by hexane extraction at 15 C, was reported to yield a light-
colored crude oil with no wax (9).
Rice bran oil refining involves very similar unit operations as for other vegetable
oils. Alkali refining is widely used for neutralization of RBO. However, some refi-
ners use molecular distillation at high temperature and low pressures. Rice bran oil
refining usually involves wax removal, which is usually achieved by gradual cool-
ing of the crude oil in settling tanks followed by filtration or centrifugation of the
sludge at low temperature (84). Water is used for phospholipid removal. Colored
compounds are removed by using either activated carbon or bleaching earth.
Although a common practice for oil deodorization has been steam stripping,
modern refiners use steam-vacuum deodorization. According to Orthoefer (8), total
losses for crude RBO refining might be as high as 1822%.
There have been several studies of SC-CO2 extraction of RBO. Taniguchi et al.
(107) noted the presence of oryzanol in SC-CO2-extracted RBO, which had a light-
er color and was less phosphorous than hexane-extracted oil. Zhao et al. (108)
showed that fractions obtained at high extraction pressures contained low FFA,
waxes, and unsaponifiables. Ramsay et al. (109), comparing yields and sterol con-
tent of the hexane- and SC-CO2-extracted RBO, showed that total sterol content of
the SC-CO2-extracted RBO was less than that found in the hexane-extracted
oil. Pilot scale SC-CO2 extraction of RBO was examined by Shen et al. (110).
The same research group also determined the apparent partition coefficients of
oil components between the oil and CO2 phases. A two-stage SC-CO2 process
that involved extraction of RBO in the first stage, followed by continuously feeding
the initial extract to a second stage expansion column to achieve further fractiona-
tion of the oil components, was also reported by the same group (110). The rate of
RBO extraction with SC-CO2 has been correlated with dimensionless Sherwood,
Schmidt, and Reynolds numbers by Kim et al. (111).
OAT AND BARLEY OIL 223
Dunford and King (102, 112, 114) developed a phytosterol-enrichment process

using supercritical fluid fractionation (SFF) technology. Enrichment of the phytos-
terol esters was achieved during oil processing rather than by isolation from the
byproducts and readdition to the oil. Using a high-pressure-packed fractionation
column, the researchers were able to obtain RBO fractions with a similar phytos-
terol ester content to that in commercially available phytosterol-enriched margar-
ines. Commercial phytosterol-enriched margarines contain mainly fatty acid
esters of phytosterols. However, the SFF product contained both fatty acid esters
of phytosterols and oryzanol. Higher oryzanol content of the SFF-processed oil
is an additional feature of the SFF process. Hexane-extracted RBO was used for
this study; however, oil extracted with SC-CO2 can also be used as a starting
material.
5. OAT AND BARLEY OIL
The annual productions of barley and oats average about 141 and 25 million metric
tons in the world (11), respectively. Like corn, most barley and oat grain is used for
animal feed (about 70% of world production) (115). Today, the use of oat and bar-
ley in human foods is very limited. However, recent interests in oat- and barley-
derived dietary fibers enriched in b-glucan create a great potential for functional
foods, nutraceuticals, and other value-added product development from these
grains.
5.1. Grain Processing

The barley kernel is comprised of the caryopsis and the enclosing hull or husk. Bar-
ley hull is strongly attached to the pericarp. Thus, it is very difficult to dehull,
instead barley is usually pearled. The caryopsis consists of the pericarp, integu-
ments, aleurone layer, endosperm, and germ or embryo (116). The embryo is
located at the attachment end of the caryopsis on its dorsal side. Barley is milled
to make blocked barley, pearl barley, barley groats, barley flakes, and barley flour.
A barley milling operation consists of the following operations: cleaning, condi-
tioning, bleaching (used in some countries), blocking (shelling), aspiration (husk
removal), sifting, cutting (on groat cutter), and pearling (rounding). Barley flakes
are made from groats or pearl barley on flaking rolls. Barley flour is produced in
a roller mill. Jadhav et al. (117) reported a processing scheme that is used for barley
processing by one manufacturer in the United States. This process involved the
following steps: carter disc separators, flat stone mill (removes hull tips that makes
up 5% of the grain), decorticator I (removes hulls and bran, which is 25% of the
grain), decorticator 2 (removes bran and germ), roller milling (grinding and sifting
to remove crease, 10% of the grain), and decorticator 3 (removes crease bran, aleur-
one, and outer endosperm, 10% of the grain). Commercially available barley
products are pot and pearled barley, grits, flakes, and malt flour (117).
The oat grain consists of the groat (actual caryopsis of the oat) and a surrounding
hull or husk. The use of whole, unmilled oats is limited due to the high cellulose
content in the hulls. Thus, the hull is separated and removed by dehulling (shelling)
(118). The oat groat contains an active lipase, which has to be inactivated before
milling to avoid lipid hydrolysis. A stabilization step to inactivate the enzyme is
an essential operation in an oat milling operation. There are two oat milling systems
(118). The traditional, or dry-shelling, system consists of the following steps: width
grading, stabilization, kiln-drying, length grading, and shelling on stones. The mod-
ern green-shelling system involves the following unit operations: width grading,
shelling by impact, stabilization, kiln-drying, and length grading. Both milling sys-
tems also employ cutting, grinding (for oatmeal, oat flour, and oat bran), steaming,
and flaking (for rolled oats). Steel-cut groats, old-fashioned flakes, quick flakes,
instant flakes, and flour are all common forms of commercial oat products (119).
Based on the American Association of Cereal Chemists definition, oat bran can be
up to 50% of the groat. The FDA also adopted this oat bran definition by some mod-
ifications: Oat bran can be up to 50% of the groat, but it has to provide at least 5%
b-glucan and 16% total dietary fiber (119).
5.2. Lipids
5.2.1. Barley The data on barley and oat lipids is very limited in the literature.
Barley differs in chemical characteristics, because of genotype and environment
and the interaction between the two. Large variations in chemical composition of
barley have been reported (120). Oil is a minor component of barley and constitutes
27% (w/w, based on dry matter) of the grain weight. The barley varieties Risoj 1580
and Hiproly contain more lipids than the average barley (117).
Ko et al. (121) reported the fractional proportions of barley grain as follows:
germ 0.3%, endosperm 72.2%, pearling flour 4.1%, bran 12.6%, and hull 10.1%
(w/w). Based on this fractionation, the oil content of germ, endosperm, pearling
flour, bran, and hull were 13%, 0.7%, 10.7%, 5.6%, and 2.6%, respectively. Similar
oil content in barley fractions (bran: 4.65.3%, germ: 14.7%) were also reported by
Seog et al. (122). However, the oil content of milled barley products may vary
depending on the milling system used. The oil contents of milled barley products
are shown in Table 26. Barley bran had the highest oil content among the milled
TABLE 26. Oil Content of Milled Barley Products.1
Oil Content
Products (% w/w, on dry matter basis)
Pearl barley 1.1

Barley flour 1.9
Barley husk 0.3
Barley bran 4.0
Barley dust 2.5
1
Adapted from Kent and Evers (118).
OAT AND BARLEY OIL 225
barley products (118). Wang et al. (123) used a Miag Multomat 8 roller dry mill and
a laboratory-type pearler to fractionate Waxbar (two-rowed) and Azhul (six-rowed)
waxy hulless barley cultivars. Fractions from the first break flour through the fourth
middling were combined, mixed, and designated as flour for the roller mill pro-
ducts. The six roller mill fractions obtained were flour, fifth middling, red dog,
reduction shorts, break shorts, and bran. Fifth middling (4.5%, w/w), followed by
red dog (3.8%), had the highest oil contents among the fractions obtained from the
roller mill. Pearling flour contained a substantially larger amount of oil (8.2%, w/w)
as compared with that of the whole grain (2.8%, w/w).
The greatest portion of the lipids in barley kernel is nonpolar lipids (6778%).
The compositions of lipids in the embryonic axis, bran endosperm, and hull
fractions of hulless barley caryopses were determined by Price and Parsons (124)
(Table 27). Neutral lipids were predominant in all fractions. Phospholipid
content of barley hull was lower than that of the bran endosperm and embryonic
axis. The hull fraction contained the highest glycolipid amount among the grain
fractions.
Linoleic acid (C18:2) was the predominant component of barley neutral lipids
(Table 28). Oleic (18:1) and palmitic acid (16:0) were the other major fatty acids
in all the barley fractions. A significant amount of polyunsaturated acid, linolenic
(18:3), was also detected in all the barley fractions. Arachidic acid (C20:0) was pre-
sent in measurable amounts in hull fraction of barley.
5.2.2. Oats The geographical location and weather can alter the total lipids of
oats, although the effect of environment is small compared with that of the varietal
effect (125). In general, the oil content of oats ranges from 2% to 11% (125). The
higher oil content is associated with Avena species other than Avena sativa. Oat
embryo contains significantly higher oil than that of bran endoperm and hull
(Table 27). Oat hull has relatively lower oil content. Practically all the lipid is in
TABLE 27. Lipid Content and Composition in Grain Fractions

of Barley and Oats.1
Lipid Composition (% of total lipid)

Grain Fraction/Lipid Type Barley Oats
Embryo axis NL 75.8 87.4

GL 6.4 3.8
PL 17.8 8.8
Bran-endosperm NL 64.4 56.9
GL 12.5 21.4
PL 23.1 21.7
Hull NL 75.9 66.9
GL 18.2 27.6
PL 5.9 5.5
1
Adapted from Price and Parsons (124).
NL: Neutral lipids, GL: Glycolipids, PL: Phospholipids.
TABLE 28. Fatty Acid Composition of Neutral Lipids of Barley and Oats.1
Barley Oats
Fatty Acid Embryonic Axis Bran-endosperm Hull Embryonic Axis Bran-endosperm Hull
C12:0 1.7 0.6

C14:0 0.2 0.2 5.3 0.1 0.3 3.4
C16:0 19.5 20.6 26.1 17.1 16.4 32.4
C16:1 0.1 2.0 0.6 0.1 1.7
C18:0 0.6 1.3 5.4 0.6 1.5 4.7
C18:1 20.4 16.6 21.5 32.1 36.1 18.2
C18:2 49.7 56.8 30.6 45.8 43.9 27.6
C18:3 9.6 4.4 5.0 3.7 1.7 8.8
C20:0 2.4 2.6
1
Adapted from Price and Parsons (124).
the dehulled grain. Although the oil content of the oat milling products varies with
the milling system used, a typical range is as follows: oatmeal, 7.5%; rolled oats,
7.6%; oat flour; 7.9%; oat husk; 0.4%; oat dust; 5.0%; meal seeds; 3.8%; and oat
feed meal; 1.5% (w/w) (118).
Oat lipids may contain 5590% neutral lipids depending on the grain fraction
(Table 27). Oat bran endosperm contains significantly higher amounts of both
glycolipids and phospholipids. The phospholipid content of oat hull is lower than
that of the other fractions. There is considerable variation in the reported propor-
tions of individual lipid classes, but the major component of the neutral lipids is
TAG (3585%) (125). The free fatty acid content of oat lipids is in the range of
211%.
The fatty acid composition of oat lipids is similar to that of the barley oil
(Table 28). Linoleic acid is the major fatty acid in all the grain fractions. Oat lipids
are also rich in oleic acid. Palmitic acid is the third major fatty acid in oat oil.
5.3. Lipophilic Bioactive Components

5.3.1. Barley Barley oil is a good source of tocols. All eight tocol isomers were
detected in barley germ (134 ppm), endosperm (12 ppm), bran (117 ppm), and hull
(34 ppm) (121). Tocols were uniformly distributed in the grain. Although the level
of total tocols was highest in the pearling flour (240 ppm), g- and d-tocopherols
and tocotrienols were not detected in this fraction. Wang et al. (123) also reported
that the barley pearling flour contained the highest amount of tocopherols and
tocotrienols. a-Tocopherol was the predominant isomer in the barley germ
(92 ppm) and pearling flour (170 ppm), whereas a-tocotrienol was the most abun-
dant isomer in endosperm, bran, and hulls (121). Peterson (126) could not detect
any a-tocotrienol in the barley germ; however, Ko et al. (121) reported a presence
of a-tocotrienol in the barley germ (8 ppm). The disparity was attributed to different
genotypes of barley and different milling processes used for these studies.
REFERENCES 227
The data on phytosterols in barley oil is not available in the literature. The most
recent data on phytosterols in barley grain was reported by Piironen et al. (127).
Total phytosterol content of barley varies between 4080 mg/100 g grain. Sitosterol
consists about 50% of the total sterols.
5.3.2. Oats Oat grain and oat oil contain both tocopherols and tocotrienols. The
major tocols are a-tocopherol and a-tocotrienol. Significant variations occur in
tocol composition of oat oil due to the oat variety and milling process used for
the study. Total tocol content of oat oil varies from 175 ppm to 640 ppm (125).
a-Tocopherol (2082% of the total tocols), a-tocotrienol (1651% of total tocols),
d-tocotrienols (012% of total tocols), and g-tocopherol (036% of total tocols) are
the predominant tocol isomers reported in the oat oil.
Presence of phytosterols in oat oil was first reported by Idler et al. in 1953 (128).
Oat grain contains 3560 mg phytosterols/100 g grain (127). Phytosterol content of
oat oil varies between 0.19% to 0.32% (125). b-Sitosterol (4070% of total sterols)
is the major phytosterol in oats. 5- and 7-Avenasterol are the two other phytos-
terols that present in significant quantities in oats. Campesterol, stigmasterol, 7-
stigmasten-3b-ol, 7-cholesten-3b-ol, and cholesterol were also present in oat
grain.
6. CONCLUSIONS
Plant germ oils of high nutritional quality can be obtained from cereal grains. Germ
oil processing presents a challenge because of the high content of biologically
active heat labile components. Physical refining methods improve retention of these
compounds in the final product. Currently plant germ oils, except corn germ oil, are
produced for specialty markets in the United States. However, increasing consumer
demand for healthy food products may change the market trends in the future.
REFERENCES
1. P. J. Barnes, in P. J. Barnes, ed., Lipid in Cereal Technology, Academic Press, New York,
1983, p. 389.
2. P. J. Barnes, Fette Seifen Anstrichmittel, 84, 256 (1982).
3. F. T. Orthoefer and R. D. Sinram, in S. A. Watson and P. E. Ramstad, eds., Corn:
Chemistry and Technology, American Association of Cereal Chemists, St. Paul,
Minnesota, 1987, p. 535.
4. L. R. Strecker, M. A. Bieber, G. Mazza, T. Grossberger, and W. J. Doskoczynski, in
T. H. Hui, ed., Baileys Industrial Industrial Oil and Fat Products, Edible Oil and Fat
Products; Oils and Oilseeds, 5th ed., vol. 2, John Wiley and Sons, New York, 1996,
p. 125.
5. T. L. Mounts and R. A. Anderson, in P. J. Barnes, ed., Lipid in Cereal Technology,
Academic Press, New York, 1983, p. 373.
6. E. J. Weber, in P. J. Barnes, ed., Lipids in Cereal Technology, Academic Press, New York,
1983, p. 353.
7. R. A. Reiners and C. M. Gooding, in G. E. Inglett, ed., Corn:Culture, Processing,
Products, Avi Publishing Co., Westport, Connecticut, 1970, p. 241.
8. F. T. Orthoefer, Food Technol, 50, 62 (1996).
9. B. O. Juliano, in P. J. Barnes, ed., Lipids in Cereal Technology, Academic Press,
New York, 1983, p. 305.
10. F. T. Orthoefer, in T. H. Hui, ed., Baileys Industrial Industrial Oil and Fat Products,
Edible Oil and Fat Products; Oils and Oilseeds, 5th ed., vol. 2, John Wiley and Sons,
New York, 1996, p. 393.
11. FAO. http://apps.fao.org.
12. N. O. V. Sonntag, in D. Swern, ed., Baileys Industrial Oil and Fat Products, 4th ed.,
vol. 1, John Wiley & Sons, New York, 1979, p. 289.
13. P. H. Rao, G. V. Kumar, R. G. C. P. Rao, and S. R. Shurpalekar, Lebensmittel Wissenschaft
und Technologie, 13, 302 (1980).
14. D. B. Jones and K. D. Widness, J. Nutr., 31, 675 (1946).
15. T. S. Kahlon, Cereal Foods World, 34, 872 (1989).
16. N. T. Dunford, Food Technol., 55, 38 (2001).
17. M. Berdick, J. Am. Oil Chem. Soc., 49, 406 (1972).
18. T. Schoenberg, Soaps Cosmetics Chem. Specialties, 61, 37 (1985).
19. J. M. Nara, R. C. Lindsey and W. E. Burkholder, J. Agric. Food Chem., 29, 68 (1981).
20. J. M. Nara and W. E. Burkholder, Environ. Entomol., 12, 703 (1983).
21. R. E. Angold, in P. J. Barnes, ed., Lipids in Cereal Technology, Academic Press,
New York, 1983, p. 1.
22. B. Godon, in B. Godon and C. Willm, eds., Primary Cereal Processing, VCH Publishers
Inc., New York, 1994, p. 61.
23. W. R. Morrison and K. D. Hargin, J. Sci. Food Agric., 32, 579 (1981).
24. J. J. C. Hinton, Biochem. J., 38, 214 (1944).
25. W. R. Morrison, S. L. Tan, and K. D. Hargin, J. Sci. Food Agric., 31, 329 (1980).
26. D. Firestone, Physical and Chemical Characteristics of Oils, Fats and Waxes, AOCS
27. N. T. Dunford and M. Zhang, Food Research International, 36, 905 (2003).
28. A. P. Simopoulos, Amer. J. Clin. Nutr., 70, 560S (1999).
29. F. B. Hu, Nutrition, 17, 741 (2001).
31. J. H. Nelson, R. L. Glass, and W. F. Geddes, Cereal Chem., 40, 343 (1963).
32. P. J. Barnes and P. W. Taylor, J. Sci. Food Agric., 31, 997 (1980).
33. K. D. Hargin and W. R. Morrison, J. Sci. Food Agric., 31, 877 (1980).
34. A. Azzi and A. Stocker, Progress In Lipid Research, 39, 231 (2000).
35. A. Pollard, A. C. Chibnall, and S. H. Piper, Biochem. J., 27, 1889 (1933).
36. A. P. Tulloch and L. L. Hoffman, Phytochemistry, 12, 2217 (1973).
37. E. Levin, V. K. Collins, D. S. Varner, J. D. Mosser, and G. Wolf, U.S. Patent 3,031,376,
1962.
REFERENCES 229
39. T. Itoh, J. Amer. Oil Chem. Soc., 50, 122 (1973).
40. R. J. Anderson, R. L. Shriner, and G. O. Burr, J. Amer. Oil Chem. Soc., 48, 2987 (1926).
41. B. Kiosseoglou and D. Boskou, Oleagineux, 42, 169 (1987).
42. A. Kuksis, Biochemistry, 3, 1086 (1964).
43. J. C. Drummond, E. Singer, and R. J. MacWalter, Biochem. J., 29, 456 (1935).
44. O. Gisvold, J. Amer. Pharm. Assoc., 29, 312 (1940).
45. G. Panfili, L. Cinquanta, A. Fratianni, and R. Cubadda, J. Amer Oil Chem. Soc., 80, 157
(2003).
47. L. Singh and W. K. Rice, U.S. Patent 4,298,622, (1981).
48. M. Taniguchi, T. Tsuji, M. Shibata, and T. Kobayashi, Agric. Biol. Chem., 49, 2367
(1985).
49. A. M. Gomez and E. M. de la Ossa, J. Amer. Oil Chem. Soc., 77, 969 (2000).
50. N. T. Dunford and J. Martinez, 6th International Symposium on Supercritical Fluids,
Versailles, France, 2003.
51. C. Willm, in B. Godon and C. Willm, ed., Primary Cereal Processing, VCH Publishers,
Inc., New York, 1994, p. 319.
52. R. A. Anderson, in G. E. Inglett, ed., Corn: Culture, Processing, Products, AVI
Publishing Company, Inc., Westport, Connecticut, 1970, p. 253.
53. F. R. Earle, Cereal Chem., 54, 70 (1977).
54. S. K. Jurgens, R. R. Johnson, and J. S. Boyer, Agron. J., 70, 678 (1978).
55. J. E. Freeman, Trans. Amer. Soc. Agric. Engr., 16, 671 (1973).
56. C. Vojnovich, R. A. Anderson, and E. L. J. Griffin, Cereal Foods World, 20, 333 (1975).
57. L. F. Welch, Agron. J., 61, 890 (1969).
58. M. M. MacMaster, M. D. Finkner, M. M. Holzapfel, J. H. Ramser, and G. H. Dungan,
Cereal Chem., 36, 247 (1959).
59. R. B. Brown, G. N. Fulford, L. Otten, and T. B. Daynard, Cereal Chem., 58, 75 (1981).
60. Z. Leibovitz and C. Ruckenstein, J. Amer. Oil Chem. Soc., 60, 347 (1983).
61. S. L. Tan and W. R. Morrison, J. Amer. Oil Chem. Soc., 56, 531 (1979).
62. M. D. Jellum, J. Agric. Food Chem., 18, 365 (1970).
63. E. J. Weber, Cereal Chem., 55, 572 (1978).
64. E. J. Weber, Biochemical Genetics, 21, 1 (1983).
65. N. K. Andrikopoulus, Food Reviews International, 18, 71 (2002).
66. G. R. List, J. P. Friedrich, and D. D. Christianson, J. Amer. Oil Chem. Soc., 61, 1849
(1984).
67. R. E. Worhington, J. Amer. Oil Chem. Soc., 59, 317A (1982).
68. E. J. Weber, in S. A. Watson and P. E. Ramstad, eds., Corn: Chemistry and Technology,
American Association of Cereal Chemists, St. Paul, Minnesota, 1987, p. 311.
69. G. W. Grams, C. W. Blessin, and G. E. Inglett, Cereal Chem., 48, 356 (1971).
70. R. A. Ferrari, W. Esteves, K. D. Mukherjee, and E. Schulte, J. Agric. Food Chem., 45,
4753 (1997).
71. T. Itoh, N. Shimizu, T. Tamura, and T. Matsumoto, Phytochemistry, 20, 1353 (1981).
72. T. Verleyen, U. Sosinska, S. Ioannidou, R. Verhe, K. Dewettinck, A. Huyghebaert, and

W. De Greyt, J. Amer. Oil Chem. Soc., 79, 947 (2002).
73. A. Maza, U.S. Patent 6,201,142, (2001).
74. R. A. Anderson, J. J. Ellis, and E. L. J. Griffin, Cereal Sci. Today, 17, 41 (1972).
75. C. G. Pina and A. J. A. Meirelles, J. Amer. Oil Chem. Soc., 77, 553 (2000).
76. M. P. Hojilla-Evangelista, L. A. Johnson, and D. J. Myers, Cereal Chem., 69, 643
(1992).
77. F. Feng, D. J. Myers, M. P. Hojilla-Evangelista, K. A. Miller, L. A. Johnson, and S. K.
Singh, Cereal Chem., 79, 707 (2002).
78. M. Cheryan, U.S. Patent 20,020,183,490, (2002).
79. D. J. Karlovic, M. Bocevska, J. Jakovlevic, and J. Turkulov, Acta Alimentaria, 23, 389
(1994).
80. D. D. Christianson, J. P. Friedrich, G. R. List, K. Warner, E. B. Bagley, A. C. Stringfellow,
and G. E. Inglett, J. Food Sci., 49, 229 (1984).
81. E. Ronyai, B. Simandi, S. Tomoskozi, A. Deak, L. Vigh, and Z. Weinbrenner,
J. Supercritical Fluids, 14, 75 (1998).
82. T. S. Kahlon, R. M. Saunders, R. N. Sayre, F. I. Chow, M. M. Chiu, and A. A. Betschart,
Cereal Chem., 69, 485 (1992).
83. J. B. Taylor, T. M. Richard, C. L. Whelm, M. M. Chrysam, M. Otternburn, and G. A.
Leveille, U.S. Patent 15,552,167, (1996).
84. S. C. Chang, R. M. Saunders, and B. S. Luh, in B. S. Luh, ed., Avi Publishing Co.,
Westport, Connecticut, 1980, p. 764.
85. B. O. Juliano, and D. B. Bechtel, in B. O. Juliano, ed., Rice: Chemistry, and Technology,
American Association of Cereal Chemists, St. Paul, Minnesota, 1985, p. 17.
86. D. B. Bechtel, and B. O. Juliano, Ann. Bot., 45, 503 (1980).
87. K. L. J. Hargrove, in W. E. Marshall, and J. I. Wadsworth, eds., Rice Science and
technology, Marcel Dekker, Inc., New York, 1993, p. 381.
88. E. T. Champagne, in W. E. Marshall and J. I. Wadsworth, eds., Rice Science and
Technology, Marcel Dekker, Inc., New York, 1993, p. 17.
89. S. N. R. Rao, T. K. Ananthachar, and H. S. R. Desikachar, J. Food Sci. Technol., 2, 115
(1965).
90. F. D. Goffman, S. Pinson, and C. Brergman, J. Amer. Oil. Chem. Soc., 80, 485 (2003).
91. D. R. McCaskill and F. Zhang, Food Technol., 53, 50 (1999).
92. L. Gingras, Inform, 11, 1196 (2000).
93. J. C. Lugay and B. O. Juliano, J. Amer. Oil Chem. Soc., 41, 273 (1964).
94. H. Taira, M. Nakagahra, and T. Nagamine, J. Agric. Food Chem., 36, 45 (1988).
95. H. Taira, J. Amer. Oil Chem. Soc., 66, 1326 (1989).
96. S. C. Deka, D. R. Sood, and S. K. Gupta, J. Food Sci. Technol., 37, 217 (2000).
97. R. Sharma and C. Rukmini, Lipids, 21, 715 (1986).
98. A. G. G. Krishna, J. V. Prabhakar, and D. P. Sen, J. Food Sci. Tech., 21, 222 (1984).
99. R. R. Eitenmiller, Food Technol., 51, 78 (1997).
100. A. C. Tomeo, M. Geller, T. R. Watkins, A. Gapor, and M. L. Bierebbaum, Lipids, 30, 1179
(1995).
REFERENCES 231
101. E. J. Roger, S. M. Rice, R. J. Nicolasi, D. R. Carpenter, C. A. McClelland, and

L. J. Romanczyk, J. Amer. Oil Chem. Soc., 70, 301 (1993).
102. N. T. Dunford and W. J. King, J. Food Sci., 65, 1395 (2000).
103. G. S. Seetharamiah, T. P. Krishnakantha, and N. Chandrasekhara, J. Nutr. Sci. Vitaminol.,
36, 291 (1990).
104. R. J. Nicolasi, E. J. Roger, D. Kritchevsky, J. Scimeca, and P. Huth, Artery, 22, 266
(1997).
105. G. Yoshino, Current Therapeutic Res., 45, 543 (1989).
106. K. Yokochi, Rice By-Products Utilization, International Conference, Valencia, Spain,
1974.
107. M. Taniguchi, E. Tsuji, H. Morimoto, M. Shibata, and T. Kobayashi, Nippon Shokuhin
Kogyo Gakkaishi, 34, 102 (1987).
108. W. Zhao, A. Shishikura, K. Fujimoto, K. Arai, and S. Saito, Agric. Biol. Chem., 51, 1773
(1987).
109. M. E. Ramsay, J. T. Hsu, R. A. Novak, and W. J. Reightler, Food Techol., 30, 98 (1991).
110. Z. Shen, M. V. Palmer, S. S. T. Ting, and R. J. Fairclough, J. Agric. Food Chem., 44, 3033
(1996).
111. H. Kim, S. Lee, K. A. Park, and I. Hong, Sep. Purif. Technol., 15, 1 (1999).
112. N. T. Dunford, J. A. Teel, and J. W. King, Food Research Intenational, 36, 175 (2002).
113. N. T. Dunford and J. W. King, J. Amer. Oil Chem. Soc., 78, 121 (2001).
114. N. T. Dunford and J. W. King, U.S. Patent 6,677,469 B1, (2004).
115. S. Kim and B. E. Dale, Biomass and Bioenergy, 26, 361 (2004).
116. N. L. Kent, Technology of Cereals, Pergamon Press, Oxford, United Kingdom, 1983,
pp. 154.
117. S. J. Jadhav, S. E. Lutz, V. M. Ghorpade, and D. K. Salunkhe, Crit. Rev. Food Sci., 38, 123
(1998).
118. N. L. Kent and A. D. Evers, Kents Technology of Cereals, Elsevier Science, Ltd., Oxford,
United Kingdom, 1994, pp. 334.
119. K. Salisbury, Grains for the Health of It: Increasing Grains and Whole Grains
Consumption as a Strategy for Health, Minneapolis, Minnesota, 2001.
120. A. A. M. Andersson, C. Elfverson, R. Andersson, S. Regner, and P. Aman, J. Sci. Food
Agric., 79, 979 (1999).
121. S-N. Ko, C-J. Kim, H. Kim, C-T. Kim, S-H. Chung, B-S. Tae, and I-H. Kim, J. Amer. Oil
Chem. Soc., 80, 585 (2003).
122. H-M. Seog, M-S. Seo, Y-S. Kim, and Y-T. Lee, Food Sci. Biotechnol., 11, 623 (2002).
123. L. Wang, Q. Xue, R. K. Newman, and C. W. Newman, Cereal Chem., 70, 499 (1993).
124. P. B. Price and J. Parsons, J. Agric. Food Chem., 27, 813 (1979).
125. E. G. Hammond, in P. J. Barnes, ed., Lipids in Cereal Technology, Academic Press Inc.,
London, United Kingdom, 1983, p. 331.
126. D. M. Peterson, Cereal Chem., 71, 442 (1994).
127. M. L. Piironen, J. Toivo, and A. M. Lampi, Cereal Chem., 79, 148 (2002).
128. D. R. Idler, S. W. Nicksic, D. R. Johnson, V. W. Meloche, H. A. Schuette, and C. A.
Baumann, J. Amer. Oil Chem. Soc., 75, 1712 (1953).
9
Oils from Herbs, Spices,
and Fruit Seeds
Liangli (Lucy) Yu, John W. Parry, and Kequan Zhou
University of Maryland
College Park, Maryland
1. INTRODUCTION
Edible seed oils are important common food ingredients. Fatty acids are primary
nutritional components found in edible seed oils. Growing evidence has suggested
that individual fatty acids may play different roles in human health. Diets rich in a
specific fatty acid may provide potential prevention of a number of health problems
or diseases. For instance, o3 (n-3) unsaturated fatty acids may have health benefits
including the prevention of cancer, heart disease, hypertension, and autoimmune
disorders. Currently, consumers growing interest in improving their dietary nutri-
tion is driving the development of novel seed oils having unique fatty acid profiles
and other beneficial components, including phytosterols and natural antioxidants. It
is the purpose of this chapter to summarize the edible fruit, spice, or herb seed oils
with unique fatty acid profiles. Physicochemical properties and other beneficial
components of these oils, such as phytosterols and tocopherols, may also be
included. The seed oils are presented according to their primary or distinguishing
fatty acid (s), including oleic, linoleic, a-linolenic, and g-linolenic acids. Seed oils
containing only small amounts of beneficial fatty acids but significant quantities of
other valuable components (natural antioxidants) are also included.
233
234 OILS FROM HERBS, SPICES, AND FRUIT SEEDS
2. EDIBLE SEED OILS RICH IN a-LINOLENIC ACID (18:3n3)
Alpha-linolenic acid (18:3n-3) is an 18-carbon fatty acid with three double bonds at
carbons 9, 12, and 15. It is an essential n-3 fatty acid that is a required nutrient for
human beings and can be obtained through diets including both plant and animal
sources. Alpha-linolenic acid can be converted by elongases and desaturases to
other beneficial n-3 fatty acids, such as eicosapentaenoic acid (EPA) and docosa-
hexaenoic acid (DHA), which are implicated in normal brain development, normal
vision, and a decreased risk of heart disease. Novel dietary sources of n-3 fatty
acids are desired for those who do not consume adequate amounts of fish or fish-
based food products rich in long-chain n-3 fatty acids. This section summarized
fruit, spice, and herb seed oils rich in a-linolenic acid (18:3n-3). These include
black raspberry, red raspberry, boysenberry, marionberry, blueberry, cranberry,
sea buckthorn, basil, and hemp seed oils.
2.1. Black Raspberry Seed Oil (Rubus occidentalis L., cv Jewel)

Black raspberry is a member of the genus Rubus from the Roseacea family, which is
also known as caneberries. The majority of black raspberry crops are located in the
Northwest region of the United States, predominantly in Oregon. The annual
harvests for black raspberries in Oregon in 2002 and 2003 were 3.02 million pounds
and 2.70 million pounds, respectively. Nearly 99.5% of the total crop goes into
postharvest production (http://www.nass.usda.gov/or/berries03.pdf), and seeds are
a major byproduct thereof.
The fatty acid profile of two cold-pressed black raspberry seed oils demonstrated
high concentrations of both n-3 and total unsaturated fatty acids. The concentration
of a-linolenic acid (18:3n-3) was 35% of total fats, and unsaturated fatty acids com-
prised 9899% (Table 1). Linoleic acid was the predominant fatty acid (Table 1);
however, the ratios of n-6 to n-3 fatty acids were very low at 1.6:1. The other mea-
surable fatty acids included oleic (18:1n-9) and palmitic (16:0) acids (Table 1). The
overall fatty acid composition of black raspberry seed oil was very similar to red
raspberry seed oil (1) (Table 1).
2.2. Red Raspberry Seed Oil (Rubus ideaus)

Red Raspberry is a production crop grown throughout the world, and the
total worldwide annual production is typically around 250,000 metric tons
(http://www.oregon-berries.com). The majority of commercial raspberries are
grown in Eastern Europe, followed by Northern and Western Europe, the
United States, and Chile. Like black raspberries, red raspberries are also grown
in the Northwest region of the United States, and total production in the years
2002 and 2003 was 42.2 metric tons (MT) and 38 MT, respectively (http://
www.nass.usda.gov/or/berries03.pdf).
Red raspberry seed oils, extracted by either hexane (2) or cold-pressing (3), were
examined for their fatty acid compositions. Both methods detected very similar
TABLE 1. Fatty Acid Compositions (g Fatty Acid/100-g Oil) of Fruit Seed Oils Relatively High in a-Linolenic Acid (18:3n-3).*
Black Red Buckthorn(6) Buckthorn(6) Buckthorn(7)

Fatty Acid Raspberry(1) Raspberry(2, 3)
Boysenberry(3) Marionberry(3) Blueberry(3) Cranberry(35) sinensis rhamnoides mongolica
16:0 1.21.6 1.22.7 4.2 3.3 5.7 3.07.8 7.79.6 6.78.2 8.6
18:0 trace 1.0 4.5 3.1 2.8 0.21.9 2.13.3 2.34.1 3.3
18:1 6.27.7 12.012.4 17.9 15.1 22.8 20.027.8 12.926.1 13.720.0 17.9
18:2n-6 55.957.9 53.054.5 53.8 62.8 43.5 35.044.31 38.243.6 36.743.0 38.6
18:3n-3 35.235.3 29.132.4 19.5 15.7 25.1 22.335.0 20.236.3 25.436.0 29.1
others nd nd nd nd nd 2.58 1.92.5 1.83.8 2.1
n-6/n-3 1.591.63 1.641.87 2.75 3.99 1.73 1.162.0 1.072.00 1.021.62 1.33
*
Black raspberry, Red raspberry, Boysenberry, Marionberry, Blueberry, Cranberry, Buckthorn sinesis, Buckthorn rhamnoides, and Buckthorn mongolica, stand for black raspberry,
red raspberry, boysenberry, marionberry, blueberry, cranberry, buckthorn sinesis, buckthorn rhamnoides, and buckthorn mongolica seed oil, respectively. Numbers correspond to
the references cited. nd stands for not detected.
fatty acid profiles and high concentrations of a-linolenic acid, an n-3 fatty acid
(Table 1). The crude oil from the hexane extract contained 29.1% a-linolenic
acid and the extra virgin cold-pressed seed oil had 32.4% a-linolenic acid. Both
of these samples were also very comparable in their fatty acid compositions
compared with the black raspberry seed oil discussed above (Table 1). In addition
to its a-linolenic acid content, red raspberry seed oil may contain a significant
level of tocopherols and other natural antioxidants (2, 3). Total tocopherol was
97-mg/100-g oil and 61-mg/100-g oil in the hexane-extracted and the cold-pressed
oils, respectively (2, 3), whereas the antioxidant activity, measured as the oxygen
radical absorbing capacity (ORAC), was 48.8-mmoles trolox equivalents per gram
of oil (3). Trolox, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, is a
water-soluble analog of a-tocopherol and widely used as a standard antioxidant
compound.
2.3. Boysenberry Seed Oil (Rubus hybrid)

Like the other caneberries (black raspberry, red raspberry, marionberry), boysenberry
also prefers the growing conditions found in the Northwest region of the United
States. However, aside from Oregon, boysenberry is also grown in Northern
California as a production crop. In 2002 and 2003, the total boysenberry production
in the United States was 2665 tons and 2350 tons, respectively.
Also, like the other cold-pressed caneberry seed oils, boysenberry seed oil had a
high percentage (19.5%) of n-3 a-linolenic acid and a low n-6 to n-3 ratio of 2.8:1.
Total unsaturated fatty acids constituted over 91% of the seed oil and polyunsatu-
rated fatty acids (PUFA) were very high at 73.3%, but stearic, palmitic, and total
saturated fatty acids were higher than all other caneberry seed oils (Table 1). Inter-
estingly, the boysenberry seed oil demonstrated the best antioxidative potential
using the oxygen radical scavenging capacity (ORAC) test compared with eight
other seed oil samples, including blueberry, black raspberry, and red raspberry
seed oils, which are known to be rich in antioxidants (3).
2.4. Marionberry (Rubus hybrid) Seed Oil

Marionberry is a blackberry hybrid. It is another member of the caneberry
family and is also grown in the Northwest United States, specifically in Oregon.
The production in 2002 was 15,000 MT and in 2003 it was 12,910 MT. Marionberry
comprises almost one-half of the total caneberry production in Oregon.
In 2004, Parry et al. (3) examined the chemical composition and physico-
chemical properties of cold-pressed marionberry seed oil. The oil was shown to
contain a relatively high percentage of n-3 fatty acids in the form of a-linolenic
acid (15.7%) (Table 1). This amount was lower than that of other caneberry seed
oils, including black raspberry, red raspberry, and boysenberry seed oils, tested
under the same conditions. The n-6 to n-3 fatty acid ratio was 4:1, which was
the highest among the tested caneberry group.
EDIBLE SEED OILS RICH IN a-LINOLENIC ACID 237
2.5. Blueberry Seed Oil (Vaccinium corymbosum)

Blueberries are grown in temperate climates throughout the world; however, the lar-
gest producers are the United States and Canada. Approximately 42,000 MT are pro-
duced annually outside of the United States and Canada. In 2002 and 2003, the United
States harvested 87.3 MT and 86.200 MT, respectively. (http://usda.mannlib.cornell.
edu/reports/nassr/fruit/pnf-bb/ncit0104.pdf, http://www.ushbc.org/blueberry.htm).
The cold-pressed blueberry seed oil investigated by Parry and Yu. (3) demonstrated a
high concentration of n-3 fatty acids. Alpha linolenic acid was the sole source of the n-3
and comprised 25.1% of the total fatty acids (Table 1). The ratio of n-6 to n-3 fatty acids
was 1.7:1. Linoleic acid (18:2n-6) was the most prevalent fatty acid in the blueberry
seed oil followed by a-linolenic, oleic, palmitic (16:0), and stearic (18:0) acids
(Table 1). The blueberry seed oil also showed a significantly higher antioxidant capacity
compared with marionberry, black raspberry, cranberry, and pumpkin seed oils using
the oxygen radical absorbance capacity (ORAC) test. Therefore, blueberry seed oil may
serve as an excellent dietary source of n-3 fatty acids and natural antioxidants.
2.6. Cranberry (Vaccinum macrocarpon) Seed Oil

The North American cranberry, Vaccinum macrocarpon, is best adapted to grow at
higher latitudes and in bog terrains. It is grown for production in Wisconsin, Maine,
New Jersey, Oregon, and Washington in the United States, and British Columbia
and Quebec in Canada. Cranberries are also grown in Europe, but are a different
species of Vaccinum. The total production in the United States for the year 2002
was 284,200 tons and was projected to be 291,500 tons in 2003 (http://usda.mann-
lib.cornell.edu/reports/nassr/fruit/zcr-bb/cran0803.pdf).
Several studies have confirmed that the seed oil from the North American variety of
cranberry contains significant levels of a-linolenic acid. In a U.S. patent, Heeg et al.
(4) reported the a-linolenic acid content of cranberry seed oil to be between 30% and
35% of total fatty acids. In 2003, Parker et al. (5) found 22.3% a-linolenic acid in the
cold-pressed cranberry seed oil, and in 2004, Parry et al. (3) determined the oil to
contain 32.0% a-linolenic acid from two different lots of the seed oil. The ratios of
n-6 to n-3 fatty acids in all were low from 1.2:1 to 2:1. Also, all of the studies docu-
mented similar ratios among the rest of the common fatty acids found in cranberry
seed oil, including, in order of higher amount present: linoleic, oleic, palmitic, stearic,
and eicosadienoic (20:2) acids (Table 1). In addition to a-linolenic acid, cranberry
seed oil is rich in natural antioxidants (8). These antioxidants may directly react
with free radicals and prevent lipid oxidation in human low-density lipoprotein.
2.7. Sea Buckthorn (Hippophae rhamnoides L.) Seed Oil

Sea buckthorn is native to Asia and Europe. It is a hardy plant that is also being
considered as a major commercial crop in Canada. It has been used in Tibetan,
Mongolian, and Chinese traditional medicine for more than 1000 years, and
has demonstrated many beneficial health attributes (6). The fruit has a good
flavor and is rich in nutrients. The whole berries contain a higher concentration of
Vitamin C than strawberry, kiwi, orange, and tomato, and the fruit also contains a
higher concentration of Vitamin E than wheat embryo, safflower, maize, and
soybean (9).
In 2001, Yang and Kallio (9) investigated the lipid compositions of two sub-
species of Hippophae rhamnoides L. The subspecies were H. rhamnoides L.
sinensis and H. rhamnoides L. rhamnoides. Twelve samples of sinensis and
nine samples of rhamnoides were grown at different locations in China and Finland,
respectively. Among the twenty-one samples there was some variation in the
compositions of the seed oils; however, they all had relatively high percentages
of a-linolenic acid, g-linolenic acid (18:3n-6), and oleic acid (18:1n-9). All seed
oil samples also had an n-6 to n-3 fatty acid ratio under 2:1 (Table 1). Other
constituent fatty acids included palmitic, stearic, and vaccinic (18:1n-7) acids.
Kallio et al. (7) examined the fatty acid composition of the subspecies sinensis,
and mongolica of Hippophae rhamnoides L. Both displayed fatty acid profiles
very similar to those found in the previous study. The fatty acid composition of
Hippophae rhamnoides L. mongolica is shown in Table 1.
2.8. Basil (Ocimum sp.) Seed Oil

Basil is a popular herb grown throughout the world and is an ingredient in many
recipes. There are more than 50 species, but sweet basil, Ocimum basilicum, is
the most common variety (10).
In 1996, Angers et al. (10) investigated the fatty acid composition of the seed
oils of four species of basil, including Ocimum basilicum, Ocimum canum, Ocimum
gratissimum, and Ocimum sanctum. Also, four total different varieties of Ocimum
basilicum were tested. All samples were compared with flaxseed oil and had similar
fatty acid profiles in regard to a-linolenic, palmitic, and stearic acids. The flaxseed
oil had 52% a-linolenic acid, and the basil seed oils had 57.462.5% a-linolenic
acid (Table 2). The n-6 to n-3 fatty acid ratio of the flaxseed oil was 1:3.2, and
TABLE 2. Fatty Acid Compositions (g Fatty Acid/100-g

Oil) of Herb Seed Oils and Other Oils with Relatively High
Concentrations of a-Linolenic Acid (18:3 n-3).*
Fatty Acid O. Basilicum (basil)(10) Hemp(5, 11, 12)
16:0 6.88.8 5.86.7

16:1 0.20.3 00.2
18:0 2.02.8 2.63.2
18:1 8.711.6 9.915.6
18:2 18.321.7 53.460.0
18:3 n-3 57.462.5 15.119.4
18:3 n-6 0.10.3 03.6
20:0 < 0.2 0.81.0
20:1 trace nd
others trace 01.8
n-6/n-3 0.30.4 2.83.5
*
O.basilicum (basil) and Hemp stand for O.basilicum (basil) and hemp
seed oil, respectively. Numbers correspond to the references cited.
nd stands for not detected.
EDIBLE SEED OILS RICH IN g-LINOLENIC ACID 239
the basil seed oils were 1:1.61:3.6. Complete fatty acid profiles are shown in
Table 2.
2.9. Hemp (Cannabis sativa) Seed Oil

Hemp is an ancient crop that is still cultivated by many societies and has many
different uses. The fiber from hemp has been used to make rope, paper, and
clothing; hemp is used for medicinal purposes, and its seed oil is commercially
available (11).
Recent investigations of hemp seed oil (5, 11, 12) reported similar findings in
fatty acid compositions. The n-3 fatty acid, a-linolenic acid, was determined to
constitute between 15.1% and 19.4% of total fat (Table 2). Gamma-linolenic acid
(18:3n-6) was also detected in two of the studies, and comprised up to 3.6% of total
fatty acids (11, 12) (Table 2). The most prevalent fatty acid was linoleic in all of the
studies, which was between 53.4% and 60.0% of total fatty acids and was followed
by a-linolenic, oleic, palmitic, g-linolenic, and stearic acids. Eicosadienoic, arachi-
dic (20:0), and behenic (22:0) acids were also detected in small quantities.
3. EDIBLE SEED OILS RICH IN c-LINOLENIC ACID (18:3n6)
Gamma-linolenic acid (18:3n-6) is an important unsaturated fatty acid. It is the

precursor for biosynthesis of arachidonic acid that is a precursor for prostaglandin
formation. Recently, g-linolenic acid has been recognized for its potential health
benefits in prevention and treatment of cardiovascular disorders, premenstrual
syndrome, atopic eczema, rheumatic arthritis, and alcoholism (13, 14). Seed oils
of blackcurrant and other Ribes species, as well as evening primrose seed oils,
are rich sources of natural g-linolenic acid.
3.1. Black Currant and Other Ribes Seed Oils

Blackcurrant (Ribes nigrum) is cultivated for the production of its berries (15).
It is rich in ascorbic acid and exhibited high levels of antioxidant activity (16).
Blackcurrant is mainly consumed in the form of juice and the seeds are the
byproduct of juice production. Blackcurrant seed oils were analyzed for fatty
acid composition, tocopherols, and their prostaglandin E2 production reduction
potential (15, 1719). Blackcurrant seed oil is an excellent dietary source of
both g-linolenic (18:3n-6) and a-linolenic (18:3n-3) acids. Gamma-linolenic acid
constituted 1225% of the total fatty acids, whereas a-linolenic acid comprised
the other 1013% (Table 3). The fatty acid composition was dependent on genotype
and growing conditions. The seed oils also had significant levels of tocopherols
(18). The total tocopherol content was 1.22.5-mg/g oil, with a mean value of
1.7-mg/g oil for ten oil samples. The major tocopherol in the blackcurrant seed
oil was g-tocopherol, but b-tocopherol was not detected in the blackcurrant seed
oil. In 1999, Wu et al. (17) investigated the effect of dietary supplementation
with blackcurrant seed oil on the immune response of healthy elderly subjects.
TABLE 3. Fatty Acid Composition (g Fatty Acid/100-g Oil) of Seed Oils Relatively High in g-Linolenic Acid.*
Evening primrose Evening primrose Evening primrose

Fatty Acid Blackcurrant(21) Blackcurrant(19) (Oenothera Spp.)(20) (Oenothera biennis)(14) (Oenothera lamarckiana)(13)
16:0 5.3 6.06.3 710 9.1 5.87.2

18:0 1.5 1.31.6 1.53.5 3.1 1.53.1
18:1 14.7 8.99.6 611 17.7 9.220.1
18:2n-6 47.0 42.743.5 6580 64.3 62.074.6
18:3n-6 12.2 22.024.6 814 4.9 5.59.6
18:3n-3 13.2 10.011.5 nd trace nd
n-6/n-3 4.5 5.76.8 N/A N/A N/A
*
Blackcurrant and Evening primrose stand for Blackcurrant and Evening primrose seed oil, respectively. Numbers correspond to the references cited. nd stands for not
detected, whereas N/A stands for not applicable.
EDIBLE SEED OILS RICH IN LINOLEIC ACID 241
The seed oil moderately enhanced immune function through reducing the produc-
tion of prostaglandin E2, suggesting that blackcurrant seed oil may have potential in
preventing cancer, cardiovascular disease, and other health problems.
Other Ribes species, including R. grossularia (red-black gooseberries), R.
grossularia (yellow gooseberries), R. nigrum (blackcurrants), R. rubrum (red
currants), and R. nigrum R. hirtellum ( jostaberries), were also examined for
g-linolenic acid concentration and tocopherol content in the seed oils. Among
the tested samples, blackcurrant seed oil had greatest level of g-linolenic acid,
and all three species of currant had of total concentration of tocopherols over
1.0 mg/g oil (18).
3.2. Evening Primrose Seed Oil

Evening primrose (Oenothera spp.) belongs to the Onagraceae family and produces
a large number of highly fertile seeds. The roots of evening primrose are used in
human diet, whereas its bark, leaves, and essential oil are used for medicinal pur-
poses (http://botanical.com/botanical/mgmh/p/primro70.html) (13, 20). Evening
primrose seed oil is a natural source of g-linolenic acid (18:3n-6). Hudson (20)
evaluated 192 evening primrose (Oenothera spp.) seed oil samples for their fatty
acid compositions. The normal range of g-linolenic acid concentration was 8
14% and the extreme range was 220% of total fatty acids, with a median of
10.4% (Table 3) (20). Linoleic acid normally accounted for 6580% of total fatty
acids and the median was 73%, which was as high as that of any known vegetable
oil. Another study showed that common evening primrose (Oenothera biennis)
seed oil contained 4.9% of g-linolenic acid, along with 64% linoleic acid (14). In
addition, the growing conditions were found to alter the g-linolenic acid content in
the seed oil. The concentration of g-linolenic acid ranged from 5.59.6% of total
fatty acids (Table 3) (13). Ratnayake et al. (14) reported that evening primrose seed
oil from Canada contained 64.3% of linoleic acid and 4.9% of g-linolenic acid
(Table 3). These previous studies indicate that evening primrose (Oenothera spp.)
seed oil contains a significant level of natural g-linolenic and oleic acids, and
linoleic acid is the predominant fatty acid. This topic is discussed in another chapter
in this series.
4. EDIBLE SEED OILS RICH IN LINOLEIC ACID (18:2n6)
Linoleic acid (18:2n-6) is an essential fatty acid that must be obtained through
diets. In this section, fruit, spice, and herb seed oils rich in linoleic acids are
summarized. These seed oils include watermelon, melon (Cucumis melo and
Colocynthis citrullus), goldenberry, grape, rose fruit, paprika, red pepper, onion,
black cumin, and Onagraceae seed oils. Several seed oils may be listed in other
sections if they contain significant level of a special fatty acid. For example, pump-
kin seed oils rich in both oleic acid and linoleic acid, are listed under the section
named, Edible seed oils rich in oleic acid (18:1n-9).
4.1. Watermelon (Citrullus vulgaris) Seed Oil

Watermelon (Citrullus vulgaris) is taxonomically classified as a member of the Cur-
cubitaceae family, which is also known as the gourd family. Other gourds include
pumpkins, cucumbers, squash, and other melons. It prefers warm climate growing
conditions and is produced worldwide where conditions permit. Over 1200 varieties
have been cultivated and about 250 varieties are grown in North America, http://
www.watermelon.org/index.asp?adsp&htypeabout&pid39. The worlds pro-
duction of watermelons in 2002 was over 81 million metric tons (MMT) and
approximately 71% of that was grown in China. The U.S. production in both
2002 and 2003 was over 1.7 MMT (http://usda.mannlib.cornell.edu/reports/nassr/
fruit/pvg-bban/vgan0104.txt). Watermelon seeds are consumed as snack food world-
wide and are used to prepare edible oil in some countries.
Watermelon seed oil was prepared and evaluated for its physicochemical pro-
perties (22, 23). The seed oil consisted of 59.6% linoleic acid (18:2n-6) and
78.4% total unsaturated fatty acids (Table 4). The predominant fatty acid in the
oil was linoleic acid, which was followed by oleic, palmitic, and stearic acids.
Linolenic, palmitoleic, and myristic acids were minor constituents. The refractive
index, acid value, peroxide value, and free fatty acids of watermelon seed oil were
determined to be 1.4696 (25 C), 2.82 (mg KOH/g oil), 3.40 (mequiv oxygen/kg oil), and
1.41 (% as oleic acid), respectively. The saponification value of watermelon seed oil
was 201 (mg KOH/g oil), and its iodine value was 115 (g iodine/100-g oil), which
was significantly higher than pumpkin at 109 (g iodine/100-g oil) (22, 23).
4.2. Melon (Cucumis melo) Seed Oil

Melon, Cucumis melo, is a member of the Cucurbitaceae family and grows best in
tropical regions. The pulp of the fruit has pleasant flavor and taste, and the seeds are
generally treated as waste; however, medicinal effects have been reported for
the seeds (24, 25). Hexane-extracted seed oil of Cucumis melo hybrid AF-522
was determined to contain 64 g of linoleic acid per 100 g of total fatty acids
(Table 4) (24). Significant amounts of oleic, palmitic, and stearic acids were also
detected in the melon seed oil. The specific gravity (28 C), refractive index
(28 C), and iodine value of the seed oil were 0.9000, 1.4820, and 112, respectively,
under the experimental conditions (24). Earlier in 1986, Lazos (25) extracted the
oil from Cucumis melo seeds and examined its physicochemical properties (25).
Linoleic acid was the primary fatty acid and accounted for 64.6% of the total fat
(w/w), along with 20.1% oleic acid, and 14.7% total saturated fatty acids (Table 4).
Iodine value and refractive index (40 ) of the seed oil were 124.5 and 1.4662,
respectively.
4.3. Melon (Colocynthis citrullus L.) Seed Oil

Colocynthis citrullus L (melon) is a tropical vine that is native to West Africa. The
flesh from the fruit of this melon is bitter and inedible; the edible part of the fruit
is the seed (33). Nwokolo and Sim (26) examined the fatty acid composition of
TABLE 4. Fatty Acid Composition (g Fatty Acid/100-g Oil) of Fruit Seed Oils Relatively High in Linoleic Acid.*
Melon Melon
Fatty Acid Watermelon(22, 23)
(cucumis melo)(24, 25)
(Colocynthis citrullus L.)(26, 27)
Goldenberry(28) Grape(29, 30)
Rose(31) Paprika(22, 23, 32)
16:0 11.3 9.09.5 11.812.1 7.3 5.814.2 1.73.1 11.213.8

18:0 10.2 4.95.6 9.010.7 2.5 8.6 1.72.5 3.23.7
18:1 18.1 19.420.1 13.514.5 11.7 13.731.9 14.718.4 9.814.6
18:2n-6 59.6 64.164.6 57.765.4 76.4 50.177.8 48.654.4 67.874.4
18:3n-3 0.4 0.20.3 2.1 0.3 5.0 16.418.4 nd
Total saturated 21.5 14.715.2 21.125.3 11.9 8.414.4 11.618.1 15.017.6
Total unsaturated 78.4 84.485.1 74.677.5 88.1 85.591.5 81.888.3 82.584.9
*
Watermelon, Melon (Cucumis melo), Melon (Colocynthis citrullus L.), Goldenberry, Grape, Rose, and Paprika stand for Watermelon, Melon (Cucumis melo), Melon (Colocynthis
citrullus L.), goldenberry, grape, rose, and paprika seed oil, respectively. Numbers correspond to the references cited. nd stands for not detected.
Colocynthis citrullus seed oil and found that it contained a relatively high percen-
tage of linoleic acid that accounted for 57.7% of total fatty acids (Table 4) (26).
Oleic acid was the second major fatty acid (14.5%). The seed oil contained
about 25.3% saturated fatty acids (Table 4). Moussata and Akoh (27) also reported
a similar fatty acid profile of Colocynthis citrullus L. seed oil. The primary
fatty acid was linoleic acid, contributing 65.4% of total fats. The other significant
fatty acids included oleic (13.5%), palmitic (12.1%), and stearic (9.0%) acids
(Table 4).
4.4. Goldenberry (Physalis peruviana L.) Seed Oil

Goldenberry, (Physalis peruviana L.), also known as cape gooseberry, is a perennial
native to the Andes. It is also cultivated in the United States, South Africa, East
Africa, India, New Zealand, Australia, and Great Britain (34). It is related to
both tomatoes and chile peppers and prefers growing in well draining soils like
tomatoes. Goldenberry has a pleasant flavor that is similar to tomatoes and is eaten
in many ways, including in salads, cooked dishes, chocolate covered desserts, jams,
preserves, and natural snacks (28). The fruit is an excellent source of Vitamins A
and C as well as minerals. Goldenberry seed oil was prepared by extracting lyophi-
lized ground seed meal with chloroform-methanol and was characterized for fatty
acid composition (28). The fatty acid composition of the seed oil is shown in Table
4. Linoleic acid was the predominant fatty acid and constituted 76.1% of total fat.
Combined monounsaturated fatty acids were 12.2%, linolenic acid was 0.33%, and
total polyunsaturated fatty acids were 76.1%. These data suggest that goldenberry
seed oil may serve as an excellent dietary source for linoleic acid, the essential n-6
fatty acid, and may be a good choice for consumers seeking a greater intake of total
unsaturated fatty acids.
The fat-soluble Vitamins E and K, carotene, and phytosterols were also detected
in the goldenberry seed oil (28). Total tocopherols were 29.7 mg/g oil, including
0.9-mg a-, 11.3-mg b-, 9.1-mg g-, and 8.4-mg d-tocopherols. The total Vitamin K
content was 0.12-mg/g oil, and the b-carotene concentration was 1.30-mg/g
oil. In addition, significant levels of phytosterols were also detected. The major
phytosterol in the goldenberry seed oil was campesterol, having a concentration
of 6.5-mg/g oil. Other phytosterols, including ergosterol, stigmasterol, lanosterol,
b-sitosterol, 5-avenosterol, and 7-avenosterol, were also detected in the seed
oil.
4.5. Grape Seed Oil (Vitis spp.)

World grape production was 61.2 million tons in 2001 (http://www.winetitles.
com.au/awol.overview/world.asp). Grape seeds are byproducts from the manufac-
turing of grape juice, jam, jelly, and wine. In 1998, Abou Rayan et al. (29) inves-
tigated the characteristics and composition of Egyptian-grown Cabarina red grape
seed oil. Crude grape seed oil was extracted with hexane at room temperature.
Linoleic acid was the major fatty acid detected and comprised more than 50% of
the total fatty acids (Table 4) (29). Oleic acid was the second major fatty acid in the
seed oil, along with significant levels of palmitic and stearic acids. This finding is
consistent with a previous observation in which linoleic acid accounted for 62% of
the total fatty acids in grape seed oil (Table 4) (30). Iodine value (IV) and peroxide
value (PV) were also determined according to the methods described in AOCS,
1983. The measured IV was 130-g iodine/100-g oil, and the PV was determined
to be 2.92-mequiv peroxide/kg oil.
4.6. Rose Fruit (Rosa canina L.) Seed Oil

Rose, Rosa canina L., also known as dogberry or hop fruit, is in the Rosaceae
family. The fruit of this particular species of rose is generally used to prepare a
stew. The seeds from Rosa canina L. were investigated for their chemical compo-
sition and nutritional values for medicinal purposes. Seed oils were prepared from
fruits grown at three locations in Turkey and evaluated for their fatty acid composi-
tion (31). Linoleic acid was the primary fatty acid detected, which ranged from
48.654.4% of total fatty acids, followed by a-linolenic acid (16.418.4%) and
oleic acid (14.718.4%) (Table 4). The seed oil contained approximately 85% total
unsaturated fatty acids, indicating that Rosa canina L. seed oil may be an excellent
source for unsaturated and essential fatty acids.
4.7. Paprika (Capsicum annuum) Seed Oil

Paprika (Capsicum annuum) is a commonly used flavor enhancer, and following
production, the seeds are treated as waste. Paprika seed oils have been evaluated for
their physicochemical properties (22, 23, 32). Paprika seed oil contained more
than 82% of total unsaturated fatty acids, with polyunsaturated fatty acids com-
prising 67.8% of total fatty acids (Table 4) (22, 23). Oleic acid was the second
major fatty acid at approximately 15% of the total. This fatty acid profile was
consistent with a previous observation by Domokos et al. (32) on the fatty acid
profile of Hungarian paprika seed oils. Linoleic acid comprised 74.4% of the total
fat, whereas oleic and palmitic acid made up 9.8% and 11.2% of total fat, respec-
tively (32). The paprika seed oil was determined to contain 870 mg/kg oil total
tocopherols, 380 mg/kg oil carotenoids, and 0.92% phytosterols (32).
4.8. Apple Seed Oil

In 1997, the production of apples was 44.7 MMT worldwide, and 84% of that was
processed (http://www.geocities.com/perfectapple/prod.html). In 20002001, the
worldwide apple production reached a record high of 48 MMT (http://www.fas.usda.
gov/htp/circular/2003/3-7-03%2520Web%2520Art.%Updates//World%2520Apple%
2520Situation%25202002-03.pdf). Apple seeds are a byproduct of processing. In
1971, Morice et al. (35) investigated the seed oils from three different varieties
TABLE 5. Fatty Acid Composition (g/100-g Fatty Acids) of Apple Seed Oils.*(35)
Fatty Acid Granny Smith Sturmer Dougherty Golden Delicious
16:0 6.87.1 4.86.4 5.76.8 8.5

16:1 0.10.2 0.1 0.10.2 0.5
18:0 1.02.1 1.52.5 1.32.1 nd
18:1 24.427.4 32.836.6 34.642.1 31
18:2 62.164.1 52.158.3 48.256.1 59
18:3 0.20.4 0.5 0.6 0.5
20:0 0.61.1 0.71.7 0.60.9 0.5
20:1 0.20.3 0.20.4 0.20.3 nd
20:2 0.10.7 0.10.7 0.00.3 nd
22:0 0.10.2 0.10.3 0.1 trace
*
Granny Smith, Sturmer, Dougherty, and Golden Delicious stand for seed oil of four varieties of apple.
Numbers correspond to the references cited. nd stands for not detected.
of apples: Granny Smith, Sturmer, and Dougherty, and compared them with the
seed oils prepared from other apple varieties. The results showed similarities in
the fatty acid profiles among the varieties (Table 5). Oleic and linoleic acids con-
sisted of 8595% of the total fatty acids in all tested samples (35). The investigators
also examined other physicochemical properties of apple seed oils. The Granny
Smith apple seed oil had an iodine value of 127-g iodine/100-g oil; the Sturmer
had an IV of 122.4-g iodine/100-g oil, and the Doughertys IV was 119-g iodine/
100-g oil. Apple seed oils may be useful as a dietary source for linoleic and oleic
acids.
4.9. Red Pepper Seed Oil

Chili peppers, Capsicum sp. are members of the Solanaceae family. They originated
in South America but are now grown worldwide. They are eaten in many dishes
throughout the world and provide the feeling of hotness when eaten. There is
a very large range of hotness among the Capsicum species, which depends on their
concentration of capsaicin. Some selected peppers ranging from lowest in concen-
tration of capsaicin include bell, Anaheim, jalapeno, Hungarian wax, serrano,
cayenne, and habanero. Red pepper seeds are byproducts from the production of
red pepper powder. For centuries in South America, peppers have been used to treat
such ailments as gastrointestinal disorders, and it is also thought to benefit those
suffering from circulatory diseases. Capsaicin, the spicy chemical in peppers, has
also been shown to be a potent anti-inflammatory in vivo (36).
The roasted red pepper seed oil contained an extremely high concentration of
linoleic acid, approximately 74%, and a high total unsaturated fat level (Table 6)
(37). The fatty acid profile was very similar to that of both goldenberry seed
(Physalis peruviana L.) and safflower oils (36). The iodine value of roasted red
pepper seed oil was determined to be 137-g iodine/100-g oil.This shows that there
is a high degree of unsaturation in the oil. Oxidative stabilities of the roasted red
TABLE 6. Fatty Acid Composition (g Fatty Acid/100-g Oil) of Herb Seed Oils Relatively
High in Linoleic Acid.*
Fatty Acid Red Pepper(37) Onion(38) Onion(3) Black Cumin(39) Onagraceae(40)
16:0 13.4 9.1 5.66.2 13.013.1 8.010.9

18:0 2.1 4.4 0.63.5 3.2 2.43.5
18:1 10.2 34.3 26.730.1 24.0 8.713.1
18:2n-6 73.9 44.6 63.7 57.057.3 71.580.0
18:3n-3 0.4 0.3 nd nd nd
*
Red pepper, Onion, Black cumin, and Onagraceae stand for red pepper, onion, black cumin, and Onagraceae
seed oil. Numbers correspond to the references cited. nd stands for not detected.
pepper seed oil were tested at different roasting times. As roasting time increased,
the oxidative stability of the oil increased significantly.
4.10. Onion Seed Oil

Onion (Allium cepa) seeds contained about 23.6% crude fat. The seed oil was ana-
lyzed for its chemical composition. The onion seed oil contained 44.6% linoleic
acid and 34.3% oleic acid (Table 6) (38). The total unsaturated fatty acids com-
prised of 79% of the oil. A greater concentration of linoleic acid was determined
in the cold-pressed onion seed oil obtained from Botanical Oil Co. (Spooner, WI).
Linoleic acid accounted for 63.7% of total fatty acids, and oleic acid ranged
from 26.730.1%. The total unsaturated fatty acids were about 90% (3). In summary,
onion seed oil may serve as a dietary source of essential n-6 fatty acid and oleic
acid.
4.11. Black Cumin (Nigella sativa L.) Seed Oil

Black cumin (Nigella sativa) is an annual spicy herb. It has been used for many
years as a food preservative and a traditional medicine for protection against and
a therapeutic remedy against a number of health disorders (41). Black cumin
seed and its oil have also been used for medicinal purposes (39). According to
Ramadan and Morsel (39), the seed contained about 2835% oil. Black cumin
seed oil contained a relatively high level of unsaturated fatty acids ( 84%) of
the total fatty acids (Table 6). The major fatty acid in the seed oil was linoleic
acid, which accounted for about 57% of the total fatty acids, followed by oleic
acid from 23.924.1%, along with a small amount of palmitic and stearic acids.
However, the iodine value of the seed oil was only 48.4-g iodine/100-g oil, which
is much lower than the expected value relating to the level of unsaturation.
4.12. Seed Oils of Some Onagraceae Rich in Linoleic Acid

Onagraceae contains about 36 genera and 300 species; it can grow in mud, sand,
rocks, or on grassy plains and occurs chiefly in the temperate zone of the New
World, primarily on the Pacific coast. Family members include evening primrose,
fushia, suncups, willowherb, and clarkia (http://1.1911encyclopedia.org/O/ON/
ONAGRACEAE.htm).
The seeds of selected Onagraceae, including Oenothera picensis, O. indecora,
Ludwigia longifolia, and O. L. peruviana were analyzed for their physicochemical
characteristics (40). Linoleic acid was the predominant fatty acid in all tested seed
oils, comprising 71.580.0% of the total fatty acids (Table 6) (40). These
Onagraceae seed oils may be excellent dietary sources of the essential n-6 fatty
acid (18:2n-6).
5. EDIBLE SEED OILS RICH IN OLEIC ACID (18:1n-9)
Oleic acid is an n-9 monounsaturated fatty acid (MUFA). Growing evidence

suggests that diets rich in oleic acid may serve as an alternative choice to a low-fat
blood cholesterol reducing diet, modulate immune function, and may delay the
development of atherosclerosis (4244). Oleic acid is the predominant fatty acid
in olive, canola, peanut, and specially produced sunflower seed oils. Oleic acid is
not an essential fatty acid; it is synthesized in vivo through the desaturation of
stearic acid (18:0). Oleic acid is also rich in a number of other edible oils, including
mango, cherry, date, pumpkin, naked seed squash, fluted pumpkin, carob bean
germ, American gingseng, Khaya senegalensis, and Moringa oleifera seed oils.
5.1. Mango Seed Kernel Oil

Mango seed kernels contain about 412% total fat (4547). Mango seed kernel oil
is rich in oleic acid (Table 7), and exhibited 42% (47), 3459% (45), and 4144% of
total fatty acids (46). Stearic acid is the other major fatty acid in mango seed kernel
TABLE 7. Fatty Acid Composition (g/100-g Fatty Acids) of Fruit Seed Oils Rich in Oleic
Acid.*
Fluted Carob Bean

Fatty Acid Mango(4547) Cherry(48) Date(49) Pumpkin(26) Germ(50)
12:0 nd nd 16.917.8 nd nd
14:0 nd nd 10.812.1 0.6 < 0.1
16:0 618 6.89.4 10.210.4 17.1 7.814.2
16:1 nd 0.40.6 0.2 nd nd
18:0 2657 1.62.1 2.82.8 15.0 3.010.0
18:1 3459 23.937.5 43.545.0 35.4 20.438.5
18:2n-6 113 40.048.9 8.78.2 27.1 43.659.2
18:3n-3 nd <1 0.6 1.2 0.31.3
20:0 <4 < 1.3 0.50.6 1.7 nd
Other FA nd 10.313.3 3.64.1 nd nd
*
Mango, Cherry, Date, and Fluted pumpkin stand for mango, cherry, date, and fluted pumpkin seed oil,
respectively. Carob bean germ stands for carob bean germ oil. Numbers correspond to the references cited.
nd stands for not detected.
EDIBLE SEED OILS RICH IN OLEIC ACID 249
oil and may account for up to 57% of the total fat. In addition, palmitic and linoleic
acids were detected in the oil along with trace amounts of a-linolenic acid (45, 47).
5.2. Cherry Seed Oil

The cherry tree (Prunus avium L.) is a member of the Rosaceae family. Cherry seed
contains about 18% oil on a dry weight basis (48). Significant levels of oleic acid
were detected in the cherry seed oils prepared by hexane extraction using a Soxhlet
apparatus. Oleic acid comprised 2438% of the total fatty acids from three different
varieties of cherry fruits (Table 7) (48). Linoleic acid was the major fatty acid in
the cherry seed oil, and ranged 4049% in the seed oil, along with a-eleostearic
(18:n-5), palmitic, stearic, arachidonic, and a-linolenic acids (Table 7). alpha-
eleostearic acid (9c,11t,13t), comprising 1013% of cherry seed oil, is a conjugated
isomer of a-linolenic acid (9c,12c,15c). alpha-eleostearic acid was not detected in
other previously studied seed oils from prunoids including peach, apricot, and plum
seed oils.
5.3. Date Seed Oil

Dates (Phoenix dactylifera L.) are popular in most Middle Eastern countries and
serve as a major source of food and nutrients (51, 52). Oil contents and fatty
acid profiles of date seeds may vary among individual varieties. Date seeds con-
tained 2024% total fat (49). Oleic acid was the primary fatty acid in the date
seed oil and had a concentration of 43.545% of total fatty acids. This was followed
by lauric (12:0), myristic (14:0), palmitic (16:0), linoleic (18:2n6), capric (10:0),
and stearic (18:0) acids along with trace amounts of other fatty acids (Table 7).
Date seed oil may serve as an excellent dietary source of oleic acid with a minor
amount of linoleic acid.
5.4. Fluted Pumpkin (Telfaria occidentalis) Seed Oil

The fluted pumpkin (Telfaria occidentalis) is a tropical gourd native to West Africa.
It is taxonomically classified as a member of the Curcubitaceae family. The fruits
are very large and weigh up to 13 kg, but only the seeds are edible (33). The seeds
are very rich in both protein and fat, containing approximately 28% and 55%,
respectively, from whole oven-dried fluted pumpkin seeds (26). The fatty acid
profile of fluted pumpkin seeds demonstrated a high oleic acid content of 35.4%
and a total saturated fatty acid concentration over 34% (Table 7) (26). Significant
level of linoleic acid (18:2n-6) was also detected in the seed oil.
5.5. Carob Bean Germ Seed Oil

Carob (Ceratonia siliqua L), a tree in the Leguminosae family, is widely cultivated
in the south and east of the Mediterranean region. The pulp tastes sweet and is
used as a food source (50). The seed germ contains about 5% oil. Oleic acid
contributed to 2039% of the total fatty acids, but only a small portion of that was
on the sn-2 position of the triacylglycerol. Linoleic acid ranged 4459%, whereas
palmitic and stearic acids accounted for 814% and 310% of the total fatty acids,
respectively, along with minor amount of linolenic and myristic acids (Table 7). In
addition, b-sitosterol was the primary sterol compound and contributed to 74% of
the total sterols. Other sterol compounds included stigmasterol (17% of total sterol),
campesterol (6%), and cholesterol (4.4%).
5.6. Pumpkin (Curcubita sp.) Seed Oil

Pumpkin, Curcubita sp., is a member of the gourd family, Curcubitaceae, that also
includes melons, cucumbers, squash, and gac. In 2003, the United States produc-
tion of pumpkins was approximately 335,000 MT (http://usda.mannlib.cornell.edu/
reports/nassr/fruit/pvg-bban/vgan0104.txt). In some mid-eastern African countries,
dried pumpkin seeds have been used to treat tapeworm when eaten on an empty
stomach (53). Also, for many years in Europe, pumpkin seeds have been used as
a remedy for micturition. Pumpkin seed oil has also shown possible beneficial
affects in retarding the progression of hypertension (54), potential anti-inflamma-
tory activity in arthritis (55), and may be effective in reducing the risk of blad-
der-stone disease (56).
The fatty acid compositions of the seed oils prepared from a combination of two
different pumpkin species (pepo and mixta) are shown in Table 8 (25). The seed oil
contained 37.8% oleic acid and 43.1% linoleic acid and was fairly high in unsatu-
rated fats (81%). Among the four pumpkin seed oil samples analyzed by Spangen-
berg and Orgrinc (57), oleic acid content was consistent and ranged 30.233.9% of
total fatty acids (57), along with 24.547.9% linoleic acid (Table 8). In addition, the
oil of unroasted pumpkin seed kernel (Cucurbita mixta) contained 21.4% oleic acid
and 58.9% linoleic acid (58). A recent study of pumpkin seed oil detected 55.6%
linoelic acid and 20.4% oleic acid in the total fatty acids, with a total unsaturated
fatty acid concentration of 76.5% (22, 23). The iodine value for the pumpkin seed
oil was 103-g iodine/100-g oil (25), and the refractive index was 1.4616 (40 C)
(25), 1.4706 (25 C) (22, 23), and 1.4615 (60 C) (58). The ORAC value of roasted
pumpkin seed oil was determined to be 1.1-m moles TE/g oil, and was the lowest in
comparison with six other fruit seed oils tested that included blueberry, red raspber-
ry, black raspberry, boysenberry, and cranberry seed oils (3). Phytosterols were also
detected in pumpkin seed oil (22, 23).
The seed oils from African pumpkin (squash) (Cucurbita pepo L.) were evalu-
ated for their fatty acid profiles and the presence of other phytochemicals (53). The
seed oils contained 2836% oleic acid. The primary fatty acid was linoleic acid,
along with palmitic and stearic acids, with a total unsaturated fatty acid concentra-
tion of 7783%. Alpha-tocopherol was determined at a concentration up to 3.0 mg/
100 g. These data suggest that pumpkin seed oil may be a better choice for consu-
mers who prefer high unsaturation, or both linoleic and oleic acids. Seed oils from
species of Cucurbita with minimal pericarp, called naked seed squash, are dis-
cussed below.
TABLE 8. Fatty Acid Composition (g Fatty Acid/100-g Oil) of Cucurbita Seed Oils Relatively High Oleic Acid.*
Pumpkin Pumpkin African

Fatty Acid Pumpkin(25) (C. Pepo)(57) Pumpkin(58) (Cucurbita sp)(22, 23)
Pumpkin(53) Pumpkin(3) Squash(60) Squash(59)
16:0 12.7 13.649.2 13.8 13.4 11.114.0 6.6 6.624.4 12.815.8

18:0 5.4 4.811.2 5.9 10.0 4.88.2 3.5 1.210.2 5.28.3
18:1 37.8 30.233.9 21.4 20.4 28.335.5 42.0 1031.6 46.660.4
18:2n-6 43.1 24.547.9 58.9 55.6 43.051.9 47.8 39.377.2 9.627.9
18:3n-3 0.3 0.40.9 nd nd 0.3 nd nd 0.10.8
*
Pumpkin and Squash stand for pumpkin and squash seed oil, respectively. Numbers correspond to the references cited. nd stands for not detected.
5.7. Naked Seed Squash (Curcubita pepo L.) Seed Oil

Squash (Curcubita foetidissima, C. pepo, and C. lagenaria) has been consumed
worldwide for thousands of years. The fruits are used as vegetables and desserts,
and seeds are consumed as nuts or used to prepare edible oils (61). Normally,
each plant produces 19 fruits weighing 0.4712.67 kg, and each fruit may have
16393 seeds (59). The weight of the individual naked seeds ranged from 46
223 mg. Seeds of the naked seed squash varieties are preferred for direct consump-
tion and further processing. In 1996, Idouraine et al. (59) reported that the crude oil
content in the nine selected naked seed squash lines (C. pepo L.) ranged 3444% on
a dry weight basis. The seed oil of the selected naked seed squash was rich in oleic
acid, which made up 4760% of the total fatty acids. Other major fatty acids were
linoleic, palmitic, and stearic acids, along with a small amount of other long-chain
fatty acids. The total unsaturated fatty acids ranged from 7079% for the nine tested
seed oils (59). These were in consistent with earlier observations that oleic acid was
the major fatty acid and constituted 4650% of the seed oils for all progeny lines
(62, 63). In summary, the seed oil of Curcubita pepo may serve as edible oil rich in
unsaturated fatty acid and oleic acid.
In addition, other studies reported that linoleic acid was the predominant fatty
acid in the seed oils of squash varieties (Curcubita foetidissima) with a range of
3977%, whereas the level of oleic acid was 1032% (25, 60, 64). Interestingly,
a significant level of conjugated dienoic acids was detected in the seed oils of
Curcubita foetidissima. It is widely accepted that conjugated linoleic acids may
have a number of health benefits, such as anticarcinogenesis, reducing body
weight, antiatherosclerotic activity, and antioxidant activity (65). Future research
is required to further confirm the presence of the conjugated dienoic acids in the
seed oils of Curcubita foetidissima and identify their chemical structures. In
summary, seed oil of Curcubita foetidissima may serve as edible oil rich in unsa-
turated fatty acid and linoleic acid.
5.8. American Ginseng Seed Oil

American ginseng (Panax quinquefolium L.), native to North America, is one of
the most widely used medicinal herbs. American ginseng seed oil prepared by
hexane or methylene chloride extraction at ambient temperature was analyzed
for fatty acid profile and phytosterol content (66). The seed oils contained about
86.887.5% oleic acid, 10.010.5% linoleic acid, and 2.6% total saturated fatty
acids (Table 9). Significant levels of phytosterols were observed in the American
ginseng seed oils prepared with hexane and CH2Cl2 (66). Squalene was the major
phytosterol compound with a concentration of 502514-mg/100-g oil, followed
by b-sitosterol and stigmasterol at levels of 164177-mg/100-g oil and
9395-mg/100-g oil, respectively. Phytosterols are thought to benefit human health
through lowering cholesterol and increasing antioxidant activity (71, 72). American
ginseng seed oil would therefore be a desirable dietary source for oleic acid,
squalene, and total phytosterols.
EDIBLE SEED OILS RICH IN OLEIC ACID 253
TABLE 9. Fatty Acid Composition (g/100-g Fatty Acids) of Seed Oils Rich in Oleic Acid.*
Fatty Acid American Ginsengo(66) K. Senegalensis(67) M. Oleifera(68, 69) M. Oleifera Mbololo(70)
14:0 nd nd < 0.1 0.1

16:0 2.22.3 21.4 5.96.5 5.76.0
16:1 nd nd 1.01.8 0.1
18:0 0.30.4 10.4 5.77.2 1.421.57
18:1n-9 86.887.5 64.6 66.976.0 73.675.4
18:2n-6 9.9510.5 nd 0.61.3 0.7
18:3n-3 nd nd < 0.2 0.2
20:0 nd nd 3.04.0 2.52.7
*
American ginsengo, K. senegalensis, M. Oleifera, and M. oleifera Mbololo stand for American ginsengo,
K. senegalensis, M. oleifera Mbololo, and M. Oleifera seed oil, respectively. Numbers correspond to the
references cited. nd stands for not detected.
5.9. Khaya senegalensis Seed Oil

Khaya senegalensis is a dry land mahogany that grows to about 30 m in height and 3
m in girth; it is widely distributed in the savanna region of Nigeria. The seed of
Khaya senegalensis contains about 53% oil. The seed oil has been used to treat cer-
tain local ailments. Oleic acid was the predominant fatty acid in the seed oil and
accounted for 65% of the total fatty acid, along with 21% palmitic and 10% stearic
acids (Table 9) (67). The seed oil had a saponification value of 186-mg KOH/100-g
oil, a density of 0.962, a refractive index of 1.4690 (20 C), and an iodine value of
68.0-g iodine/100-g oil (67).
5.10. Moringa oleifera Seed Oil

Moringa oleifera, native to the western and sub-Himalayam tracts of India and
other countries in Asia, Africa, the Middle East, Central America, and the
Caribbean islands, is the most common and broadly distributed species of the
Moringaceae family (73). The leaves, flowers, fruits, and roots of the tree are
used as vegetables (70). Each fruit generally contains about 20 seeds, which
have an average weight of 0.3 g, with the kernel responsible for 7075% of the
weight. The seed oils of Moringa oleifera were extracted and analyzed for chemical
compositions (68, 69). The seed oil was a rich source of monounsaturated fatty
acid, especially 18:1 that contributed to 74.476.0% of the total fatty acids
(Table 9) (68, 69). A significant amount of phytosterols was detected in the seed
oil (68). The major sterols were b-sitosterol (46.65% of total sterol), stigmasterol
(19%), campesterol (16%), and 5-avenasterol (10.7%), along with a number of
others.
The seeds of M. oleifera variety Mbololo yielded 26%, 31%, and 36% crude oil
by cold-pressing, hexane extraction, and chloroform-methanol extraction (1:1, v/v),
respectively (70). The seed oil was rich in total monounsaturated fatty acids
and contained 7475% oleic acid (Table 9). The total saturated fatty acids were
1921% in the seed oil, with small amounts of linoleic and a-linolenic acids. The
density of the seed oils prepared by solvent extraction and cold-pressing ranged
from 0.88090.9182 g/mL at 24 C, which is similar to that of olive oil at 0.915 g/
mL at the same temperature. The refractive index (ND40 C) was 1.45491.4591
and the smoke point was 198202 C for the seed oils. The seed oil prepared by
cold-pressing had a greater viscosity of 103 mPa s, whereas the oil prepared by
solvent extraction exhibited a viscosity of 5766 mPa s, which is more comparable
with that of olive oil (74 mPa s) (70). In addition, the seed oil of M. oleifera variety
Mbololo contained greater concentrations of total sterols and tocopherols than olive
oil. The primary sterol in the Mbololo seed oil was b-sitosterol with significant
levels of stigmasterol, campesterol, and 5-avenasterol, as well as small amounts
of other sterol compounds (70). Thus, seed oil of M. oleifera may serve as a dietary
source of oleic acid, sterols, and tocopherols.
6. OTHER SPECIAL SEED OILS OF FRUIT, SPICE, AND HERB
6.1. Gac (Momordica cochinchinensis)

Gac (Momordica cochinchinensis Spreng) is another member of the gourd family
(Curcubitaceae). The gac fruit is round and about 1822 cm in diameter. It is native
to Asia and is both consumed as food and used in traditional medicine. The seeds
comprise about 16% of the total fresh weight of the fruit.
Ishida et al. (74) examined gac seed oil for its fatty acid composition. The oil
was determined to contain an average of 60.5% stearic acid. Palmitic and arachidic
acids were two other saturated fats found in the gac seed oil, and the total percent of
saturated fatty acids ranged from 60.5% to 79.2%. The most prevalent unsaturated
fatty acid was linoleic acid at 20.3%, with an interesting variety of others including
palmitoleic (16:1), oleic (18:1n-9), cis-vaccenic (18:1n-11), a-linolenic acid
(18:3n-3), eicosa-11-enoic acid (20:1n-11), and eicosa-13-enoic acid (20:1n-13)
(Table 10).
TABLE 10. Fatty Acid Composition (g /100-g Fatty Acids)

of Gac and Pomegranate Seed Oils.*
Fatty Acid Pomegranate(75) Gac(74)
16:0 4.8 5.26.2

18:0 2.3 54.571.7
18:1 6.3 4.811.2
18:2n-6 6.6 11.225.0
18:3n-3 nd 0.40.6
18:3n-5 65.3 nd
others nd 3.04.1
*
Pomegranate and Gac stand for pomegranate and gac seed oil,
respectively. Numbers correspond to the references cited. nd stands
for not detected.
SUMMARY 255
6.2. Pomegranate Seed Oil

Pomegranate (Punica granatum), of the Punicaceae family, is a small tree grown
in Iran, India, and the United States, as well as in most Near and Far East countries
(75). Pomegranate is used as a table fruit and is also processed into juice. Pomegra-
nate preparations, including the juice of the fruit, the dried pericarp, the bark, and
the roots, have been used in folk medicines to treat colic, colitis, dysentery, diar-
rhea, menorrhagia, oxyuriasis, parasis, headache, vermifugal, carminative, antispas-
modic, taenicidal, and emmenagogue (75). Seeds are byproducts form the juice
manufacture. Cold-pressed pomegranate seed oil was prepared and analyzed for
fatty acid composition, inhibitory effects against both cyclo-oxygenase and lipox-
ygenase, antioxidant properties, and total phenolic content (75). The major fatty
acid was punicic acid (18:3n-5), which comprised 65% of total fatty acids, along
with linoleic, oleic, palmitic, and stearic acids (Table 10). The seed oil contained
about 150 ppm total phenolics on an oil weight basis. The oil extract, at a concen-
tration of 5 mg/mL total phenolics, exhibited 37% inhibition of the sheep
cyclo-oxygenase activity under experimental conditions (75). On a same weight
concentration basis, the oil extract resulted in 75% inhibition of the soybean lipox-
ygenase activity, whereas butylated hydroxy anisole (BHA) had a 92% inhibition
under the same experimental conditions. The oil extract also showed strong antiox-
idant activity in the coupled oxidation system of carotene and linoleic acid, and the
antioxidant capacity was comparable with that of BHA and green tea extract on the
same weight concentration basis (75). These data suggest the potential application
of pomegranate seed oil as anti-inflammatory agent and for general health promotion.
7. SUMMARY
There is an increasing demand for edible oils with special fatty acid profiles
and other beneficial components for improving nutritional status. A number of
studies have been conducted to screen for and evaluate the chemical composition
and potential nutraceutical applications of fruit, spice, and herb seed oils. Among
the discussed edible seed oils, some have unique fatty acid compositions, such
as black raspberry and hemp seed oils rich in a-linolenic acid and date and naked
seed squash seed oils rich in oleic acid, whereas blackcurrant seed oil is rich in
g-linolenic acid. The oils of selected fruit, spice, and herb seeds may also contain
significant levels of phytosterols, tocopherols, carotenoids, and natural antioxidants.
The chemical composition of edible seed oil determines the potential health
benefit and applications for the oil. Individual edible seed oils may be preferred
by special groups of consumers for preventing and treating a selected health
problem or for general health promotion. Great opportunities are available in the
research and development of specialty seed oils and the oil-based nutraceutical
products from fruit, spice, and herb seeds for improving human health. More
research is required to screen and characterize the fatty acids and bioactive
components in the fruit, spice, and herb seeds to develop novel edible seed oils
for optimum human nutrition.
REFERENCES
1. J. W. Parry and L. Yu, J. Food. Sci., 69, 188192 (2004).

2. B. D. Oohmah, S. Ladet, D. V. Godfrey, J. Liang, and G. Benoit, Food. Chem., 69, 187193
(2000).
3. J. W. Parry and L. Yu, Abstracts of Papers, 228th National Meeting of the American
Chemical Society, Philadelphia, Pennsylvania, August 2226, 2004.
4. H. Heeg, T. Lager, and G. Bernard, U.S. Patent 6,391,345. 2002.
5. T. D. Parker, D. A. Adams, K. Zhou, M. Harris, and L. Yu, J. Food. Sci., 68, 12401243
(2003).
6. B. Yang and H. P. Kallio, J. Agric. Food Chem., 49, 19391947 (2001).
7. H. Kallio, B. Yang, P. Peippo, R. Tahvonen, and R. Pan, J. Agric. Food Chem., 50, 3004
3009 (2002).
8. L. Yu, K. Zhou, and J. Parry, Food Chem., Accepted for publication.
9. T. S. C. Li, in J. Janick, ed., Perspectives on New Crops and New Uses, ASHS Press,
Alexandria, Virginia, 1999, pp. 335337.
10. P. Angers, M. R. Morales, and J. E. Simon, in J. Janick, ed, Progress in New Crops, ASHS
Press, Arlington, Virginia, 1996, pp. 598601.
11. B. D. Oohmah, M. Busson, D. V. Godfrey, and J. C. G. Drover, Food Chem., 76, 3343
(2002).
12. D. Pierce and G. Brooks, U.S. Patent 6,063,369, 2000.
13. Z. Yaniv, C. Ranen, A. Levy, and D. Palevitch, J. Exp. Bot., 40, 609613 (1989).
14. W. M. N. Ratnayake, D. G. Matthews, and R. G. Ackman, J. Amer. Oil Chem. Soc., 66,
966969 (1989).
15. M. L. Ruiz Del Castillo, G. Dobson, R. Brennan, and S. Gordon, J. Agric. Food Chem., 52,
948952 (2004).
16. C. E. Lister, P. E. Wilson, K. H. Sutton, and S. C. Morrison, Act. Hor., 585, 443449 (2002).
17. D. Wu, M. Meydani, L. S Leka, Z. Nightingale, G. J. Handelman, J. B. Blumberg, and S. N.
Meydani, Amer. J. Clin. Nutr., 70, 53643 (1999).
18. F. D. Goffman and S. J.Galletti, J. Agric. Food Chem., 49, 349354 (2001).
19. M. L. Ruiz Del Castillo, G. Dobson, R. Brennan, and S. Gordon, J. Agric. Food Chem., 50,
332335 (2002).
20. B. J. F. Hudson, J. Amer. Oil Chem. Soc., 61, 540543(1984).
21. U. Ollson, P. Kaufmann, and B. G. HerslOf, J. Chromatogr., 505, 385394 (1990).
22. T. A. El-Adaway and K. Taha, J. Agric. Food Chem., 49, 12531259 (2001).
23. T. A. El-Adaway and K. Taha, Food Chem., 74, 4754 (2001).
24. M. L. S. De Melo, N. Narain, and P. S. Bora, Food Chem., 68, 411414 (2000).
25. E. S. Lazos, J. Food Sci., 51, 13821383 (1986).
26. E. Nwokolo and J. S. Sim, J. Sci. Food Agric., 38, 237246 (1987).
27. C. O. Moussata and C. C. Akoh, J. Amer. Oil Chem. Soc., 74, 177179 (1997).
28. M. F. Ramadan and J. T. Morsel, J.Agric. Food Chem., 51, 969974 (2003).
29. M. A. Abou Rayan, A. A. Abdel-Nabey, O. R. Abou Samaha, and M. K. Mohamed, Alex. J.
Agric. Res., 43, 6779 (1998).
REFERENCES 257
30. M. G. Massanet, J. A. Montiel, E. Pando, and F. L. Rodriguez, Grasas Aceites., 37, 233
236 (1986).
31. M. Ozcan, J. Med. Food., 5, 137140 (2002).
32. J. Domokos, J. Peredi, and J. Bernath, Acta. Hort., 331, 4952 (1993).
33. G. I. D. Ogunsua and A. O. Ogunsua, Plant Foods Hum. Nutr., 41, 3544 (1991).
34. R. McCain, in J. Janick and J. E. Simon, eds., New Crops, Wiley, New York, 1993, pp. 479
486.
35. I. M. Morice, F. B. Shorland, and E. Williams, Sci. Food Agric., 22, 186188 (1971).
36. R. Sancho, C. Lucena, A. Macho, M. A. Calzado, M. Blanco-Molina, A. Minassi,
G. Appendino, and E. Munoz, Eur. J. Immunol., 32, 17531763 (2002).
37. M. Y. Jung, J. Y. Bock, S. O. Baik, J. H. Lee, and T. K. Lee, J. Agric. Food Chem., 47,
17001704 (1999).
38. P. U. Rao, Food Chem., 50, 379382 (1994).
39. M. F. Ramadan and J. T. Morsel, Nahrung., 46, 240244 (2002).
40. J. A. Zygadlo, R. E. Morero, R. E. Abburra, and C. A. Guzman, J. Amer. Oil Chem. Soc.,
71, 915916 (1994).
41. M. L. Salem and M. S. Hossain, Int. J. Immunopharmaco., 22, 729740 (2000).
42. R. L. Hargrove, T. D. Etherton, T. A. Pearson, E. H. Harrison, and P. M. Kris-Etherton,
J. Nutr., 131, 17581763 (2001).
43. R. J. Nicolosi, T. A. Wilson, G. Handelman, T. Foxall, J. F. Keaney, and J. A. Vita, J. Nutr.
Biochem., 13, 392402 (2002).
44. P. Yaqoob, Eur. J. Clin. Nutr., 56(Suppl. 3), S9S13 (2002).
45. G. Lakshminarayana, T. Chandrasekhara Rao, and P. A. Ramalingaswamy, J. Amer. Oil
Chem. Soc., 60, 8889 (1983).
46. M. A. Ali, M. A. Gafur, M. S. Rahman, and G. M. Ahmed, J. Amer. Oil Chem. Soc., 62,
520523 (1985).
47. Rao P. Udayasekhara, Food Chem., 50, 379382 (1994).
48. F. Comes, M. Farines, A. Aumelas, and J. Soulier, J. Amer. Oil Chem. Soc., 69, 12241227
(1992).
49. W. N. Sawaya, J. K. Khalil, and W. J. Safi, J. Food Sci., 49, 617619 (1984).
50. M. P. Maza, R. Zamora, M. Alaiz, F. J. Hidalgo, F. Millan, and E. Vioque, J. Sci. Food
Agric., 46, 495502 (1989).
51. S. A. Salem and S. M. Hegazi, J. Sci. Food Agric., 22, 632 (1971).
52. W. N. Sawaya, H. A. Khatchadourian, J. K. Khalil, W. M. Safi, and A. Al-Shalhat,
J. Food Sci., 47, 1489 (1982).
53. Y. M. H. Younis, S. Ghirmay, and S. S. Al-Shihry, Phytochem., 54, 7175 (2000).
54. H. A. Zuhair, A. A. Abd El-Fattah, and M. I. El-Sayed, Pharmacol. Res., 41, 555563
(2000).
55. A. T. Famhim, A. A. Abd-EL Fattah, A. M. Agha, and M. Z. Gad, Pharmacol. Res., 31,
7379 (1995).
56. V. Suphakarn, Amer. J. Clin. Nutr., 45, 115121 (1987).
57. J. E. Spangenberg and N. Ogrinc, J. Agric. Food Chem., 49, 15341540 (2001).
58. B. S. Kamel, J. M. DeMan, and B. Blackman, J. Food. Technol., 17, 263269 (1982).
59. A. Idouraine, E. A. Kohlhepp, and C. W. Weber, J. Agric Food Chem., 44, 721724 (1996).
60. J. A. Vasconcellos, J. M. Berry, C. W. Weber, and W. P. Bemis, J. Amer. Oil Chem. Soc., 57,
310313 (1980).
61. E. S. Lazos, Plant Foods Hum. Nutr., 43, 257273 (1992).
62. W. P. Bemis, J. W. Berry, M. J. Kennedy, D. Woods, M. Morah, and A. J. Deutchman,
63. M. S. El-Gharbawi, Libyan J. Agric., 6, 199203 (1977).
64. J. A. Vasconcellos, W. P. Bemis, J. M. Berry, and C. W. Weber, J. Amer. Oil Chem. Soc., 59,
5568 (1981).
65. L. Yu, D. Adams, and M. Gabel, J. Agric. Food Chem., 50, 41354140 (2002).
66. T. H. J. Beveridge, Th. S. C. Li, and J. C. G. Drover, J. Agric. Food Chem., 50, 744750
(2002).
67. F. E. Okieimen and C. O. Eromosele, Bioresource Technol., 69, 279280 (1999).
68. F. Anwar and M. I. Bhanger, J. Agric. Food Chem., 51, 65586563 (2003).
69. G. Vlahov, P. K. Chepkwony, and P. K. Ndalut, J. Agric. Food Chem., 50, 970975 (2002).
70. J. Tsaknis, S. Lalas, V. Gergis, V. Dourtoglou, and V. Spiliotis, J. Agric. Food Chem., 47,
44954499 (1999).
71. W. H. Ling and P. J. Jones, Life Sci., 57, 195206 (1995).
72. G. S. Kelly, Altern. Med. Rev., 4, 2936 (1999).
73. J. F. Morton, Econ. Bot., 45, 318333 (1991).
74. B. Ishida, C. Turner, M. Chapman, and T. McKeon, J. Agric. Food. Chem., 52, 274279
(2004).
75. S. Y. Schubert, E. P. Lansky, and I. Neeman, J. Ethnopharmacol., 66, 1117 (1999).
10
Marine Mammal Oils
1. INTRODUCTION
Marine oils are obtained from the flesh of fatty fish, liver of lean white fish, and
blubber of marine mammals. Although lipids from marine fish have been used as
food and medicine, traditional uses of blubber lipids of marine mammals were
mostly industrially oriented, except for Innus and Eskimos. Marine mammal oils
were lubricants or train oils as well as fuel and used for lighting (1). However,
recent research findings on the importance of long-chain polyunsaturated fatty acids
(LC PUFA) in human health have opened new channels for their value-added use in
food and pharmaceutical industries (2). During the last three decades, it has been
established that Greenland Eskimos living on their traditional diet have a lower
incidence of coronary heart disease than do Danes living on a western-style diet
(3, 4). It has been recognized that polyunsaturated fatty acids could be useful in
controlling serum triacylglycerols, but the fatty acids provided by the food industry
were often of the o6 family (1).
Lipids from marine mammals such as seal, whale, and walrus are primarily
stored as subcutaneous fat or blubber. Seal blubber comprises 29% of the carcass
259
260 MARINE MAMMAL OILS
TABLE 1. Lipid Content (g/100 g tissue) of Selected Tissues from Four Species of Seals.1
Tissue Harp Gray Ringed Hooded
Blubber 93.88 1.64 91.93 1.07 93.55 1.98 89.43 1.82

Muscle 1.92 0.03 1.82 0.03 1.85 0.53 2.36 0.74
Brain 8.10 0.32 10.25 0.10 6.86 1.01 7.40 0.79
Kidney 2.97 0.18 3.42 0.04 3.58 0.07 3.14 0.05
Heart 2.19 0.31 1.81 0.38 2.32 0.01 2.04 0.01
Liver 3.83 0.19 5.60 0.94 3.71 0.07 3.66 0.03
Lung 2.24 0.46 2.04 0.03 2.05 0.07 1.76 0.01
1
Data from Ref. (6).
weight and is considered a valuable component of it (5). Blubber lipids are mobi-
lized in times of energy need and replenished when food is in excess (2). Although
the blubber is the major site of lipid in the body of marine mammals, lipid is also
found in the muscles, liver, kidney, heart, lung, brain, and other organs. Lipid con-
tent of different tissues in different species of seals varies (Table 1) (6). In addition,
milk contains a high content of fat. A study carried out on hooded seals showed that
females may secrete up to 10 kg of milk on a daily basis, which contains 60% fat
(7). The milkfat of marine mammals resembles the composition of the depot fats
of these animals (8). The mobilization of maternal fat reserves and transfer of milk-
fat from mother to pup occurs at very high rates (9). Lipids in marine mammals
function as a source of energy, structural components of cells and tissues, and pro-
vide buoyancy (10). The blubber of marine mammals, especially harp seal, because
of its economic importance, has been the subject of numerous studies on marine
oils (6).
2. LIPID CLASSES
The oils from marine mammals are of very different composition. In the character-
ization of marine oils, many chromatographic techniques have been employed.
These techniques include thin layer chromatography (TLC), gas chromatography
(GC), high-performance liquid chromatography (HPLC), and supercritical fluid
chromatography (SFC). These techniques have advantages and disadvantages
depending on the goal of the analysis. TLC has been an excellent tool for qualitative
analysis of components present in marine oils (11).
The various lipid classes of marine mammals include triacylglycerols, diacylgly-
cerols, monoacylglycerols, free fatty acids, wax esters, cholesterol, cholesterol
esters, hydrocarbons, vitamins, and ether lipids. Some marine oils are very simple,
containing almost exclusively triacylglycerols (TAGs), whereas others contain a
variety of lipid classes (11), as shown in Table 2. Seal blubber, the depot lipid, is
mostly composed of neutral lipids (98.9%), in contrast to intramuscular lipids
(78.8% neutral and 21.1% polar lipids) (13). TAGs are the predominant components
of seal blubber, whereas organ lipids include both TAGs and phospholipids, and
TABLE 2. Composition of Arterial Lipid, as wt% of Total, for Three Species of Whales.1
Wax and Sterol Wax Alcohol Free Fatty

Species Tissues Hydrocarbons Ester Triacylglycerol and Sterol Monoacylglycerol Phosphatide Acid
Sperm whale Aortic lesions 16.5 6.4 13.8 57.8 5.5

Aortic intima 13.1 4.4 13.0 0.4 66.1 2.7
Aortic media 3.3 2.8 14.8 0.3 76.9 1.9
Killer whale Fibrous aortic atheroma 17.3 11.4 12.6 58.7
Aortic intima 0.3 7.4 29.5 17.0 0.2 45.5
Aortic media 0.6 4.2 23.9 16.4 0.6 54.3
Fatty coronary atheroma 56.4 20.9 4.1 18.6
Coronary intima media 0.7 1.7 50.4 17.3 0.9 25.8
Pilot whale Aortic lesions 2.8 17.1 9.2 28.6 0.4 35.9 6.0
Aortic intima media 0.4 13.5 4.6 20.5 0.6 57.5 3.1
1
From Ref. (12).
differences exist that originate from their varying proportions in different tissues. In
marine mammals such as whales and seals that have enormous layers of fat under
skin, TAGs serve as insulating material, which permits survival even in the cold
waters of the Arctic and the Antarctic (14).
In addition to TAGs, wax esters (long-chain alcohols esterified to fatty acids) are
another important group of neutral lipids found in marine mammals. Most species
of marine mammals have C32, C34, C36, and C38 (total of alcohol plus acid) as
major components (15). Whale oils are especially interesting because some contain
fatty acids that are largely in the form of wax esters (16). The oils from the blubber
of the Physeteridae may consist mainly of wax esters. The sperm whale blubber oil
consists of a mixture of about 79% wax esters and 21% TAGs (17). The dwarf
sperm whale (K. simus) blubber oils consist of 42% wax esters and 58% TAGs
(18). The blubber fat of beaked whales (Berardius, Hyperoodon, and Ziphius) is
composed almost entirely of wax esters (9499%) along with low levels of TAGs
(26%) (19). Several of possible functions for wax esters in marine mammals has
been proposed; these functions include their role as a reserve energy store, buoy-
ancy, metabolic water, thermal insulation, and biosonar (2022).
Among unsaponifiable matters, hydrocarbons, especially long-chain hydrocar-
bons, are found in detectable amounts in marine mammal oils. Some marine oils
contain less than 0.1% hydrocarbons, whereas others contain as much as 90%
(23). In the liver of the seal, Arctocephalus (Pinnipedia), liver squalene was
0.50% of the oil (24). High squalene contents (90%, 91%, and 92.8%) occur in
shark liver oils (2325). The 16:0, 16:1, and pristine were found in the bottlenose
whale (Berardius bairdi); (26) pristine is a highly unsaturated long-chain hydrocar-
bon (C49) occurring in the liver oil of sei whale (Balaenoptera borealis) (27) and
sperm whale (Physeter catodon) (28). In the blubber of the sei whales, pristine was
present at 11.3% and squalene at 13.1% of the total unsaponifiable fraction (29).
Total hydrocarbons were present at 0.3% of dry matter weight of the blubber,
1.6% in liver, and 1.3% in the muscle (30). Among cetaceans, limited data for
two dolphins have been published: In Delphinus longirostris liver, very long-chain
hydrocarbons (C44) were detected (31), and zamene was present in Langenor-
ynchus acutus (32).
3. FATTY ACID COMPOSITION
The fatty acid composition of marine lipids varies significantly, especially when
compared with vegetable oils. The fatty acid composition of blubber oil of marine
mammals is generally similar to fish oils as it contains a large proportion of long-
chain highly unsaturated fatty acids. However, the proportion of fatty acids in fish
and marine mammals varies considerably (2).
A marine oil typically contains some 40 different fatty acids, with carbon num-
bers varying from 10 to 24, which results in many different TAGs with the same
carbon number, but with different levels of unsaturation (11). The fatty acids
FATTY ACID COMPOSITION 263
present in marine mammal oils can be classified as saturated and unsaturated fatty
acids. The fatty acids C12:0, C14:0, C16:0, and C18:0 are among the common satu-
rated fatty acids. In addition, marine oils usually contain detectable (0.2%)
amounts of C20:0, and sometimes recognizable C24:0, but very little C22:0; the
total is normally 0.5% or less for these three fatty acids (33). Even numbered car-
bon fatty acids make up about 97% of the total fatty acids, with a few notable
exceptions (17). Some fatty acids with odd-numbered carbon chain such as
C15:0 and C17:0, along with traces of C13:0 and C19:0, have also been found in
marine oils (33). Besides, monomethyl branched fatty acids have been isolated
from marine oils, such as 3-methyldodecanoic acid from blubber of the sperm
whale physeter catodon (34).
In contrast to relatively small amounts of saturated fatty acids, marine mammal
oils have been characterized by high amounts of monounsaturated fatty acids
(MUFAs) and o3 polyunsaturated fatty acids (PUFAs) (35, 36). For instance, the
content of MUFAs in neutral and polar lipids in seal blubber is more than 60%
and 46%, respectively (37). Most fatty acids are long-chain with 20 to 22 carbon
atoms and have o3 configurations. Ackman et al. (38) have pointed out that the
total C20 and C22 monounsaturated and polyunsaturated fatty acids in each layer
of whale blubber is nearly constant, but the ratios of the monounsaturated to poly-
unsaturated fatty acids change significantly. The most common long-chain PUFAs
in marine lipids are eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)
as well as a smaller amount of docosapentaenoic acid (DPA), all of which belong to
the o3 family (10). The high content of o3 fatty acids in marine lipids is suggested
TABLE 3. Fatty Acid Composition (g/100 g) of Blubber of Various Species of Seal.1
Fatty Acid Bearded Gray Harbor Harp Hooded Ringed
14:0 3.05 3.83 0.03 4.52 0.13 4.66 0.49 4.40 0.38 3.36 0.66
16:0 DMA ND ND ND ND ND ND
16:0 10.14 6.61 0.08 8.03 0.38 6.24 0.44 9.81 1.57 4.82 2.07
16:1 o7 17.77 12.77 0.09 19.26 0.53 14.93 0.46 10.09 0.35 23.12 0.18
18:0 DMA ND ND NNND ND ND ND
18:1 o9 DMA ND ND ND ND ND ND
18:1 o7 DMA ND 0.45 0.01 ND 0.46 0.00 ND ND
18:0 2.15 0.94 0.02 0.85 0.02 0.95 0.03 1.83 0.31 0.42 0.19
18:1 o9 16.76 24.50 0.44 18.61 0.55 18.59 1.01 22.77 2.66 19.72 1.33
18:1 o7 9.49 4.95 0.09 5.16 0.44 3.57 0.36 3.75 0.47 5.03 0.46
18:2 o6 2.30 1.28 0.00 1.27 0.04 1.36 0.20 1.63 0.20 2.58 0.02
20:1 o9 5.08 12.50 0.43 9.06 0.33 12.56 2.92 13.00 1.86 6.71 2.17
20:4 o6 0.94 0.51 0.00 0.44 0.00 0.36 0.96 0.31 0.03 0.30 0.02
20:5 o3 8.28 4.85 0.13 9.31 0.21 6.82 0.69 5.21 1.65 8.72 1.06
22:0 0.63 <0.3 1.19 0.02 <0.3 <0.3 0.75 0.67
22:1 o11 0.27 0.62 0.03 0.31 0.01 0.77 0.61 0.86 0.33 0.34 0.01
22:5 o3 4.26 5.06 0.05 4.22 0.14 4.78 0.25 2.29 0.08 5.46 0.47
22:6 o3 7.22 8.91 0.29 7.76 0.98 10.48 1.98 9.56 2.36 9.45 1.74
DMA: dimethyl acetal.

ND: not detected.
1
From Refs. (6) and (40).
TABLE 4. Fatty Acids Composition of Blubber Expressed as a Percentage by Weight

of Fatty Acids Present from both Sexes of Four Species of Phocid Seals.1
P. vitulina largha P. fasciata P. hispida P. barbatus

Fatty Acids Male Female Male Female Male Female Male Female
12:0 0.03 0.04 0.02 0.02 0.03 0.02 0.06 0.05

12:1 tr tr tr tr 0.01 tr tr 0.01
13:0 tr 0.01 0.01 0.01 0.01 0.01 0.01 0.01
13:1 o9 tr tr tr tr tr tr tr tr
14:0 1.89 2.45 2.34 3.00 3.36 1.68 1.77 2.76
14:1 0.38 0.42 0.43 0.39 0.83 0.56 0.29 0.26
15:0 0.31 0.41 0.33 0.33 0.25 0.22 0.39 0.40
15:1 o6 0.07 0.05 0.06 tr 0.01 0.01 0.06 0.05
16:0 6.82 7.96 5.38 7.67 5.25 3.23 7.27 9.96
16:1 o7 10.96 10.63 9.26 8.51 19.04 9.52 12.46 14.05
16:2 o4 0.24 0.39 0.33 0.36 0.45 0.32 0.32 0.44
16:3 o6 0.06 0.04 tr
16:3 o3 0.11 tr 0.16 0.25 0.13
17:0 0.54 0.47 0.42 0.35 0.47 0.41 1.88 0.80
17:1 o8 0.39 0.32 0.25 0.27 0.39 0.27 0.57 0.33
18:0 0.93 1.10 1.05 1.75 0.66 0.73 1.66 2.30
18:1 o9 29.75 24.15 21.31 20.91 15.79 16.21 21.25 21.69
18:2 o6 0.89 1.10 0.82 1.07 0.81 0.59 0.62 0.82
18:2 o4 0.21 0.22 0.23 0.16 0.20 0.17 0.21 0.16
18:3 o6 tr 0.05 tr 0.27 0.07 tr 0.11
18:3 o3 0.13 0.32 0.30 0.88 0.52 0.22 0.11 0.21
18:4 o3 0.52 0.83 0.80 0.73 1.03 0.75 0.72 0.93
19:0 tr 0.27 0.26 tr 0.73 0.35 0.35 0.52
19:1 0.02 0.07 0.15 0.11 0.10 0.08 0.41 0.32
20:0 0.06 0.09 tr 0.36 tr 0.07 tr 0.23
20:1 3.75
20:1 o9 13.83 8.94 14.99 13.58 5.01 8.17 3.94 8.99
20:2 o6 0.15 0.16 0.20 0.29 0.13 0.13 0.24 0.24
20:2 o4 0.11 tr tr tr 0.04 0.10 0.08 0.12
20:3 o6 tr tr tr 0.09 0.08 0.06 0.11 0.11
20:3 o3 tr tr tr tr tr tr tr tr
20:4 o6 0.54 0.57 0.63 1.21 0.33 0.39 1.31 0.76
20:4 o3 0.26 0.37 0.56 0.68 0.34 0.37 0.38 0.51
20:5 o3 6.54 10.56 8.47 8.24 11.96 10.57 14.01 9.27
21:0 0.08 tr tr 0.32 0.29
21:5 o3 0.06 0 0.21 0.58 0.51
22:0 0.05
22:1 o11 2.69 2.42 5.07 6.49 0.68 1.57 0.82 3.26
22:2 o6 0.16 0.20
22:4 o6 0.68 0.81 0.81 0.64 1.16 1.44 0.83
22:4 o3 0.21 0.22
22:5 o3 7.81 7.67 7.74 6.51 11.62 14.55 9.82 4.76
22:6 o3 12.62 16.66 17.06 14.54 17.79 26.19 12.05 13.38
24:1 0.42 0.43 0.59 0.57 1.12 0.76 0.59 0.56
1
From Ref. (44).
TABLE 5. Fatty Acid Composition of Lipids (g/100-g lipid) from Different Tissues of Harp Seal.1
Fatty Acid Blubber Muscle Brain Kidney Heart Lung Liver
14:0 4.66 0.49 2.46 0.71 0.48 0.07 1.15 0.88 0.91 0.03 2.40 0.06 1.21 0.20
16:0 DMA ND 1.86 0.15 1.66 0.19 3.09 0.67 3.74 0.10 1.93 0.02 0.15 0.01
16:0 6.24 0.44 12.29 0.91 15.45 0.64 12.71 0.19 10.56 0.81 24.45 0.67 13.63 0.76
16:1 o7 14.93 0.46 7.30 0.79 1.33 0.22 4.05 0.09 4.32 0.12 2.54 0.08 5.34 0.13
18:0 DMA ND 0.78 0.08 4.38 0.43 1.65 0.28 2.69 0.33 1.68 0.08 0.16 0.04
18:1 o9 DMA ND 0.84 0.09 1.82 0.08 0.74 0.06 2.58 0.05 0.92 0.02 0.19 0.06
18:1 o7 DMA 0.46 0.00 0.97 0.11 2.69 0.18 0.79 0.07 1.74 0.03 1.01 0.00 0.13 0.04
18:0 0.95 0.03 5.93 0.19 18.08 0.69 12.30 0.59 11.09 0.34 8.72 0.26 19.55 0.87
18:1 o9 18.59 1.01 18.48 0.68 14.02 0.89 12.52 0.45 16.09 0.17 13.41 0.28 11.80 1.45
18:1 o7 3.57 0.36 4.88 1.39 4.60 0.28 5.50 0.35 4.56 0.25 3.96 0.27 6.10 0.84
18:2 o6 1.36 0.20 1.54 0.12 0.15 0.09 3.33 0.03 3.86 0.08 1.14 0.02 2.04 0.38
20:1 o9 12.56 2.92 11.75 1.58 1.83 0.24 2.20 0.57 4.51 0.05 2.68 0.62 3.40 0.58
20:4 o6 0.36 0.96 3.87 0.63 5.29 0.22 10.11 1.08 9.90 0.13 5.81 0.14 9.50 0.57
20:5 o3 6.82 0.69 5.56 0.60 0.70 0.60 9.86 0.53 9.74 0.13 4.86 0.31 8.07 1.74
22:0 <0.3 1.81 0.32 0.25 0.09 0.99 0.67 0.58 0.32 0.56 0.27 0.58 0.04
22:1 o11 0.77 0.61 0.93 0.23 0.21 0.13 0.21 0.04 ND 0.81 0.00 0.23 0.13
22:5 o3 4.78 0.25 2.21 0.32 2.92 0.26 2.13 0.05 1.17 0.06 2.35 0.04 2.16 0.05
22:6 o3 10.48 1.98 6.73 1.30 15.56 0.64 3.29 0.83 4.11 0.59 5.80 1.07 8.17 1.83
DMA: dimethyl acetal.

ND: not detected.
1
From Refs. (6) and (40).
as a consequence of cold temperature adaptation, because at lower habitat tempera-

tures, o3 PUFA remain liquid and oppose any tendency to crystallize (33). Most
long-chain PUFAs are formed in unicellular phytoplankton and multicellular sea
algae and eventually pass through the food web and become incorporated into
the body of fish and other higher marine species, including marine mammals, which
often eat fish (39). The fatty acid composition of oils from most species of marine
mammals has been summarized (1). Seal oils, because of the increasing interest in
seal fishery and product development, have been in focus and frequently studied by
researchers. The fatty acid composition of oils from different species of seal has
been reviewed (2). Table 3 shows the fatty acids and their contents in blubber lipid
from six species of seals.
The fatty acid composition of blubber of marine mammals such as seals is regu-
lated by the diet (41), location (42), season, as well as physiological conditions such
as age (43) and sex (44) of the animal. Table 4 presents the fatty acid compositions
of seals of different species and sexes. In some marine mammals, the depot fats are
largely dietary fatty acids laid down with a minimum change, but the fatty acids of
the lipids of the essential organs have terrestrial characteristics (1). Fatty acid com-
position also depends on tissue and species of the animal. However, differences are
most apparent among tissues (Table 5). Seal blubber, for example, had a high con-
tent of monounsaturated fatty acids, but it was low in arachidonic acid, dimethyl
acetals, and DHA. Lung tissue lipids were high in palmitic acid, and heart tissue
lipids had a higher content of linoleic acid. The proportions and fatty acid consti-
tuents in defferent tissues are different, most probably because of their varying
functional requirements (6). The lipids of vital organs of seals and whales contain
high proportions of fatty acids of the o6 family, similar to those of terrestrial ani-
mals. The distinction between the fatty acids of functional organs such as liver,
heart, and other organs with depot fat has been discussed (6, 45).
The fatty acid distribution in the TAG molecules in blubber oil of marine mam-
mals is different from that of fish oils. The o3 fatty acids (EPA, DPA, DHA) in
blubber oil of marine mammals tend to be located primarily in the sn-1 and sn-3
TABLE 6. Positional Distribution of Fatty Acids in Triacylglycerols from Blubber Fat

of Marine Mammals.1
Fatty Acids (mol %)
Position 14:0 16:0 16:1 18:0 18:1 18:2 20:1 22:1 20:5 22:5 22:6 18:4 20:4
Harbor seal 1 4 11 15 1 29 1 18 8 3 2 3
2 11 13 30 1 30 3 3 1 1 1 1
3 1 4 14 1 26 1 16 7 8 6 10
Harp seal 1 1 7 9 1 27 1 17 4 6 4 15
2 6 9 27 2 36 5 4 1 2 1 3
3 1 5 11 1 20 2 7 1 12 11 26
Sei whale 1 3 13 3 4 14 1 33 10 3 1 6 1 5
2 12 6 12 1 29 5 10 2 5 1 3 4 6
3 4 6 2 2 7 1 28 16 6 3 16 1 3
1
From Ref. (14).
OXIDATIVE STABILITY 267
TABLE 7. Fatty Acid Distribution in Different Positions of Triacylglycerols of Harp Seal

Blubber Oil.1
Fatty Acid sn-1 sn-2 sn-3
Total saturates 6.34 25.56 4.32

Total unsaturates 90.51 73.25 94.32
Total monounsaturates 62.91 65.98 51.09
Total polyunsaturates 27.60 7.27 43.23
Eicosapentaenoic acid (EPA) 8.36 1.60 11.21
Docosapentaenoic acid (DPA) 3.99 0.79 8.21
Docosahexaenoic acid (DHA) 10.52 2.27 17.91
Total omega3 25.65 5.56 38.87
Total omega6 0.75 1.58 3.34
1
From Ref. (2).
positions of TAGs (Tables 6 and 7), whereas in fish oils, they are located abundantly
in the sn-2 position of TAGs (46). In harp seal oil, as measured by 13C NMR, only
3.2% of DHA and 4.6% of EPA were esterified to the sn-2 position of the TAG (47).
During digestion, the fatty acids are liberated from sn-1 and sn-3 positions of the
TAG by a position-specific enzyme such as pancreatic lipase, whereas the fatty
acids attached to the sn-2 position are distributed in the body in the form of chy-
lomicron (2).
4. OXIDATIVE STABILITY
Marine oils, like other highly unsaturated oils, are susceptible to oxidation and are
more sensitive to oxidative deterioration than vegetable oils. Furthermore, they con-
tain only insignificant amounts of natural antioxidants, such as tocophorols (48).
The long-chain PUFAs (EPA, DPA, DHA) contain five or six double bonds that ren-
der them prone to atmospheric oxidation accompanied by the development of a
fishy taste and smell (49). The secondary products of oxidation may give rise to
unacceptable flavors and odors in the oil, impair digestibility of the oil, and can
damage or destroy the bodys cells, as a result of free radical attack (48). Although
antioxidants are generally available for prevention of lipid oxidation in foods, nat-
ural antioxidants, such as tocopherols, unfortunately are not effective in inhibiting
oxidation of marine oils (50). The autoxidation rate of PUFA depends on the type
and structures of fatty acids in lipids (51). For instance, seal oil is more stable than
fish oil and less vulnerable to the natural process of oxidation because of its fatty
acid composition and distribution and location of fatty acids in the triacylglycerol
molecules as well as because of its minor components (48). Removal of minor com-
ponents from oil may result in lower oxidative stability of oils (5255). Interester-
ification may change the oxidative stability of marine oils. Oxidative stability of
minke whale blubber oil was reduced after redistribution of fatty acids with lipase
or NaOCH3 (50). The lower oxidative stability of interesterified whale oil seemed to
TABLE 8. Changes of Tocopherol and Acid Value of Seal Blubber Oil during
Processing.1
Acid Value (mg KOH/g oil)

Sample a-Tocopherol (mg/100g) (as % Oleic Acid)
Crude 2.8 0.18 2.72 0.10

(1.367%)
Alkalirefined 3.20 0.11 0.08 0.00
(0.040%)
Refinedbleached (RB) 3.1 0.12 0.03 0.01
(0.015%)
Refinedbleached and deodorized (RBD) 2.4 0.09 0.04 0.01
(0.020%)
1
From Ref. (56).
be caused by displacement of PUFAs located at the sn-1 and sn-3 positions in TAG
to the sn-2 position. PUFAs located at the sn-1 and sn-3 positions in whale oil might
become more susceptible to free radical oxidation when they are transferred to 1,2-
or 2,3-positions (50). Furthermore, overprocessing of marine oils may adversely
affect their keeping quality by the removal of their endogenous natural antioxidants
(Table 8), and it is recommended that processing be minimized or that important
minor components with antioxidant activity be returned to the oil to improve their
quality (56).
Control of autoxidation of unsaturated fatty acids is vital to preserve integrity,
nutritional value, and functionalities of marine oils. Several reports in the literature
have reviewed these matters in detail. Controlling of the availability of essential
reactants in the oxidation process, such as oxygen, light, and other factors, may pro-
vide a means of retarding autoxidation. The level of available oxygen for reactions
can be controlled by reducing the partial pressure of oxygen (57, 58) or replacing
the headspace of the container with a nonreactive gas such as nitrogen. Proper
packaging of lipid is also necessary to prevent contact of oxygen with unsaturated
fatty acids. Microencapsulation that has been practiced as a packaging technique
for oils can coat oil droplets and prevent their contact with atmospheric oxygen
(59). In addition, hydrogenation controls autoxidation by reducing the reactivity
of lipid molecule, but at the cost of reducing or eliminating PUFA and compromis-
ing the nutritive value of the oil. Moreover, antioxidants may be added at very low
concentrations to control oxidation without changing the color and flavor to pre-
serve the nutritive value of oils (2).
5. PROCESSING
The basic processing steps for the manufacturing of marine oils for human con-
sumption involve cooking or rendering to release the oil followed by possible
degumming, alkali refining, bleaching, and finally deodorization as well as possible
PRODUCTION OF o3 FATTY ACID CONCENTRATES 269
addition of antioxidants (2). During these processing steps, free fatty acids, mono-
and diacylglycerols, phospholipids, sterols, vitamins, hydrocarbons, pigments, pro-
teins and their degradation products, suspended mucilaginous and colloid-like mat-
ter, and oxidation products of fatty acids are removed from the oil (60). Processing
steps of marine oils are similar to those for vegetable oils; however, the quality of
crude marine oils is less uniform than that for crude vegetable oils. To obtain high-
quality crude marine oils, proper handling and chilling of raw material to minimize
oxidative damage after landing is vital (2).
The rendered, crude oil from blubber of marine mammals can also be treated
with silica at low temperature under vacuum followed by bleaching and deodoriza-
tion, as described by Mag (48). The resulting oil, which is completely bland, is
essentially free of proteinaceous materials, phosphatides and mucilage, and proox-
idant metals and very low in colored compounds, peroxides, and secondary oxida-
tion products. This method avoids the use of acids and bases that are required in
conventional degumming and alkali refining of marine oils, thus eliminating the
risk of contamination as well as reducing the number of processing steps. The
method is also environmentally friendly because it does not require soapstock
and waste water processing (48). Another approach for preparing and stabilizing
food-grade marine oil has been proposed by Kendrick and Macfarlane (49). This
method includes treating the oil with silica, optionally in the presence of carbon,
and with vacuum steam deodorization at 140210 C in the presence of antioxidants
(49). Regardless of the processing method employed, the resultant product must be
stabilized by addition of food-grade antioxidants, particularly mixed tocopherols.
6. PRODUCTION OF x3 FATTY ACID CONCENTRATES
Supplementaion of o3 fatty acids has been recommended in addition to making

attempts to substitute saturated fatty acids with PUFA in dietary lipids. Marine
oils serve as a rich source of o3 fatty acids and may be used as the raw material
for preparation of o3 fatty acid concentrates. It has been suggested that PUFA con-
centrates devoid of more saturated fatty acids are much better than marine oils
because they allow keeping the daily intake of total lipids as low as possible
(61). Enriched fatty acids or esters can be produced by fractional vacuum distilla-
tion (62), low-temperature crystallization (63), chromatographic separation, includ-
ing HPLC (6466), silver resin chromatography (67), supercritical fluid extraction
(68), urea complexation (69, 70), and enzyme hydrolysis (71), among others.
Fractional vacuum distillation takes advantage of differences in the boiling
points of fatty acids under vacuum. This method is a an old one and requires
high temperature. The fractionation of marine oil esters is difficult because
separation of such components becomes less effective with increasing molecular
weight (72).
Low-temperature crystallization is based on the fact that the melting point of
fatty acids changes considerably with the type and degree of unsaturation (73).
At low temperatures, long-chain saturated fatty acids that have higher melting
points crystallize out and PUFAs remain in the liquid form. Briefly, the process con-
sists of cooling the oil or fatty acids in a solvent, which holds for a specified period
of time, and of removing the crystallized fraction by filtration. This method requires
the least amount of equipment and the simplest apparatus and has been an indispen-
sable method for preparing pure fatty acids (74, 75).
Fatty acids could also be separated according to their carbon number or degree
of unsaturation with appropriate adsorbents (63). Chromatographic techniques such
as HPLC and silver resin chromatography have been successfully employed to pre-
pare o3 fatty acid concentrates. However, these methods have certain shortcom-
ings, including use of organic solvents, loss of resolution of the column upon
repeated use, and difficulties in scaling up the process for commercial production
(76).
Supercritical fluid extraction (SFE) is a method that circumvents some problems
associated with conventional separation techniques. Carbon dioxide, as an inert,
inexpensive, nonflammable, and environmentally acceptable gas is the solvent of
choice because of its moderate critical temperature and pressure (76). SFE has
been used effectively to refine marine oils and remove cholesterol, polychlorinated
biphenyls (PCB), Vitamin E, and other components (77). The disadvantages of this
process include the use of extremely high pressures and the high capital cost.
The simplest and most efficient technique for obtaining o3 PUFA concentrates
in the form of free fatty acids is urea complexation. This technique is well estab-
lished for elimination of saturated and monounsaturated fatty acids (70). In this
method, the saturated and monounsaturated fatty acids can easily complex with
urea after hydrolysis of TAG with alkaline, and crystallize out on cooling and
may subsequently be removed by filtration (70). This method is favored by many
researchers because complexation depends on the configuration of the fatty acid
moieties because of the presence of multiple double bonds, rather than of pure phy-
sical properties such as melting point or solubility (10).
It is generally considered that PUFA in the acylglycerol form is nutritionally
more favorable than the corresponding methyl or ethyl esters as impaired intestinal
absorption of alkyl esters of o3 fatty acids has been observed in laboratory animals
(7881). Although most methods produce PUFA concentrates in the form of free
fatty acids or their corresponding alkyl esters, enzyme hydrolysis is a technique
proposed to produce o3 fatty acids concentrates in the form of acyglycerols by
hydrolyzing the TAG with lipase. Saturated and monounsaturated fatty acids can
be easily hydrolyzed because they do not present any barriers to lipases such as
the commercial microbial lipases (71).
Commercial marine mammal oils, such as seal blubber oil products, are avail-
able in form of soft gel capsules as nutritional supplements (82). The quality para-
meters of three commercial seal oil capsules are listed in Table 9.
Marine mammal oils or their o3 concentrates can also be modified for different
applications. Modifications include the changing of the fatty acid composition and/
or their location in the glycerol backbone. Structured lipids containing both o3
long-chain PUFAs, possibly from seal blubber oil, or their o3 concentrates, and
medium-chain fatty acids (MCFAs), which are saturated fatty acids with 612
APPLICATIONS 271
TABLE 9. Quality Comparison of x3 Seal Oil Capsules (Laboratory Analysis).1
Quality Indicators Brand A Brand B
EPA w/w % 6.89 0.07 6.43 0.07

DPA w/w % 4.11 0.02 3.76 0.00
DHA w/w % 8.16 0.01 8.50 0.10
18:3 o3 w/w % 0.48 0.01 0.48 0.00
18:4 o3 w/w % 1.10 0.04 1.35 0.06
Tocopherols mg/100 g Up to 400 Up to 1400
Acid value mg KOH/g oil 3.21 0.05 1.69 0.03
Conjugated dienes 5.80 0.25 7.63 0.13
TBARS value mmol/g oil 4.20 0.05 4.50 0.15
1
From Ref. (56).
carbon atoms, have been produced. Enzyme-catalyzed synthesis of structured lipids

has been proposed, with commercial lipase preparations (83). The final products,
with reduced calorie, exhibit the combined health benefits of long-chain PUFAs
and MCFAs, which are believed to possess many unique nutritional and metabolic
characteristics (83).
7. APPLICATIONS
Marine oils have been widely used in food and pharmaceutical industries as well as
in nonedible applications. The nonedible uses of marine oils primarily exploit their
highly unsaturated nature. In leather manufacturing, sulfated marine oils are used to
treat leather to prevent its brittleness and dryness. Oleochemicals (fatty acids, fatty
alcohols, esters of methyl and other alcohols, nitrogen derivatives) derived from
marine oils find a wide range of industrial applications, including use in lubricants,
corrosion inhibitors, plastic and rubber compounding, floatation agents, personal
care products, cleaners, textile and paper additives, asphalt additives, and tableting,
among others. In addition, marine oils have long been used as an alternative fuel to
petroleum-based products (2). Other industrial uses of marine oils are in the man-
ufacturing of polyurethane resins, cutting oils, caulks and sealants, printing ink for-
mulations, insecticides, and buffing compounds (84). Refined marine oils have also
been used in skin and hair care products. Marine oils may be used in the manufac-
turing of animal and poultry feed. Traditionally, marine oils have been used as an
economic source of calories to stimulate growth of farmed animals. However, cur-
rent knowledge about successful inclusion of EPA and DHA to mitochondria,
microsomes, and lipoprotein membranes of chicken by feeding marine oil supple-
mented diets has provided novel uses for marine oils in the animal feed industry (2).
It has been demonstrated that chicken has a natural predisposition to accumulate
EPA and DHA from the precursor C18:3 (o3) (85). Thus, inclusion of fish meal
in chicken diet enhanced the accumulation of EPA and DHA in chicken flesh
(85). Feeding seal blubber oil at 1.25% to hens was found to increase long-chain o3
PUFA and decrease arachidonic acid in the egg yolk lipids without any detriment to
their sensory properties (82). This result is not surprising because deodorized oil
was used in this study and the level of inclusion of seal oil was modest. In the
food industry, the major global use of marine oils has been in the manufacturing
of margarine and other edible oil products. In this, hydrogenated marine oils
were a low-priced alternative to vegetable oils. However, hydrogenation reduces
the unsaturation of fatty acids and negates the potential health benefits of PUFA;
if not fully hydrogenated, introduction of trans-fats to product formulation is also
of concern. Therefore, incorporation of long-chain fatty acids into the diet con-
tinues to be a topic of interest for food manufacturers, scientists, and consumers (2).
8. HEALTH BENEFITS AND DISEASE PREVENTION
Recognition of the health benefits associated with consumption of seafoods (o3

fatty acids) is one of the most promising developments in human nutrition and dis-
ease prevention research in the past three decades. According to our current knowl-
edge, long-chain o3 PUFAs play an important role in the prevention and treatment
of coronary artery disease (86), hypertension (87), diabetes (88), arthritis and other
inflammatory (89), autoimmune disorders (90), as well as cancer (91, 92) and are
essential for normal growth and development, especially for brain and retina (93).
The most direct and complete source of o3 oils is found in the blubber of certain
marine mammals, especially in the harp seal. Among its advantages is that the
bodys absorption of o3 fatty acids from marine mammal blubber may be faster
and more thorough than is the case with flaxseed and fish oils (48). As marine mam-
mal oils contain a high concentration of MUFAs, it is possible that some of their
beneficial effects may be ascribed to their MUFAs or to the combined effect of
MUFA and o3 PUFA (94). A pilot study indicated that a low dose of seal oil sup-
plementation can reduce atherogenic risk indices in young healthy persons, and the
effects are strongly dependent on the integrated o3 fatty acids dose (95,96). The
essential fatty acids found in seal oil include a high level of DPA (up to ten times
that of fish oils). There is growing evidence that DPA is the most important fatty
acid that keeps artery walls soft and plaque free (48). Marine oils are also attractive
from a nutritional point of view because they are thought to provide specific phy-
siological functions against thrombsis, cholesterol buildup, and allergies (50). Oils
from the blubbers of seal and whale have beneficial effects on selected parameters
that play a role in cardiovascular disease; it has been hypothesized that the effect of
whale oil is not mediated by its o3 fatty acids alone (97). The difference in the
beneficial effects of whale and seal oils on cardiovascular disease may argue against
the distribution of o3 fatty acids in TAG as being relevant to the superiority of
whale oil, because the o3 fatty acids are mainly in the sn-1 and sn-3 positions of
both oils. The effect of whale oil is probably not mediated by o3 fatty acids alone
as the content of these fatty acids is relatively low in whale oil. Thus, in addition to
HEALTH EFFECTS OF DPA 273
o3 fatty acids, other dietary factors may play a role in the protective effects against
atherosclerosis and thrombosis in Greenland Eskimos (97).
The beneficial effects of PUFA have also been ascribed to their ability to lower
serum TAG, to increase membrane fluidity, and to reduce thrombosis by conversion
to eicosanoids (98). Both EPA and DHA induced increases in the serum concentra-
tions of the corresponding fatty acids as well as in their relative contents in platelets
(99). However, distribution of o3 PUFA in TAG molecules influences glycerolipid
metabolism and arachidonic acid contents of serum and liver phospholipids as well
as thromboxane (TX) A2 production. In rats that were fed marine oils, for instance,
plasma and liver TAG concentrations were more effectively reduced by dietary seal
oil than by fish oil. Furthermore, dietary seal oil reduced arachidonic acid content
in liver phosphatidylcholine and phosphatidylethanolamine and serum phos-
phatidylcholine more effectively than fish oil. Activities of fatty acid synthase
(FAS), glucose-6-phosphate dehydrogenase (G6PDH), and the malic enzyme
were significantly lowered when hamsters were fed seal oil (100). The predominant
effect of seal oil was caused by the suppression of fatty acid synthesis in the liver
(101). In addition, reduction of TX A2 production of platelets and whole blood
platelet aggregation by seal oil has been observed (102, 103).
9. HEALTH EFFECTS OF DPA
Docosapentaenoic acid (DPA) is an elongation product of EPA. Unlike other o3

fatty acids, DPA has not been studied in any detail. Because of its availability,
DPA is present in a much lower concentration compared with that of EPA and
DHA in marine oils, and because of the difficulty in purifying it from mixtures con-
taining EPA and DHA with similar physicochemical properties (104). Although
only less than 1% DPA can be found in most fish oils, it is relatively more abundant
in seal oil. Harp seal oil contains 46% of DPA. DPA is almost as important as
either EPA or DHA. About one-third of the long-chain o3 fatty acids circulating
in human blood are attributed to DPA as the effective agent (105). DPA may
have pharmacological effects different from those of EPAs and DHAs, and it has
recently appeared as a focus topic (104).
It has been demonstrated that angiogenic activity in endothelial cells induced by
vascular endothelial growth factor (VEGF) can be suppressed by o3 PUFA treat-
ment. Among LC PUFA, DPA was the most potent inhibitor of angiogenesis; the
inhibitory activity by DPA pretreatment was approximately six-fold in comparison
with that of EPA and DHA, which indicates that DPA could be developed as a novel
drug or supplement against angiogenesis-related diseases (106). Angiogenesis plays
a major role in tumor growth and metastasis, and blocking angiogenesis can restrict
tumor growth. In addition, the stimulative effect of EPA on endothelial cells migra-
tion may be caused by DPA. In vitro studies have revealed that the activity of DPA
to stimulate endothelial cell migration is ten times higher than EPA (107). There-
fore, it is possible that the antiarteriosclerotic function of seal oil is mainly caused
by DPA rather than by EPA and DHA (104). Evidence suggests that DPA is the
most important fatty acid in keeping artery walls soft and plaque-free (105). A
recent study published by Tokyo Medical and Dental University indicates that
DPA can be more than ten times as effective as EPA in helping to heal damaged
blood vessels (105). Moreover, arachidonic acid-stimulated blood platelet aggrega-
tion was inhibited by o3 fatty acids in a dose-dependent manner, among which
DPA was the most potent inhibitor (105). DPA exhibits considerable activity for
interfering with the cyclooxygenase pathways, thus inhibiting platelet aggregation
most effectively (108). In addition, it has been suggested that the DPA concentra-
tion in platelet is inversely associated with coronary artery disease in women (109),
and a high proportion of DPA in serum is associated with a decreased risk of acute
coronary events in middle-aged men (110).
10. COMPARISON OF FISH OIL AND MARINE MAMMAL OIL
Fish oils and marine mammal oils are generally characterized by a large group of
saturated and unsaturated fatty acids, which are commonly associated with their
mix of TAGs (16). However, differences exist between the oils from fish and marine
mammal sources.
Fish oils may generally be described as flesh oil, liver oil, or oil of the whole fish
(111). The livers of white lean fish are known to be high in oil content. The fish
livers of cod, halibut, and shark contain approximately 50% oil and serve as an
important source of Vitamin A and Vitamin D (112). TAG is the major component
of depot fats of fish. Ether lipids, however, are restricted to the liver oils of deep sea
sharks (113). Like marine mammal oils, fish oils are rich in MUFA and PUFA, and
they are good sources of o3 PUFA such as EPA and DHA. However, DPA, which is
abundant in blubber of marine mammals, especially seals, is found in much lower
level or is absent in fish oils. Furthermore, the molecular configurations of EPA and
DHA in fish oil vary slightly from that found in marine mammal oils (48). Research
has shown that seal oil may be more beneficial than fish oil in reducing the risk of
heart disease and diabetes, Which is likely because of the relative absence of DPA
in fish oil and possibly the slower rate at which the body can use EPA and DHA
from fish oil (105). Fish oils vary considerably in the type and level of their fatty
acids depending on the particular species and their diets. For example, fish species
raised by aquaculture often have a lower level of o3 fatty acids than those in the
wild (48), and freshwater fish contain higher levels of the o6 fatty acids than do
marine fish (112). The fatty acid distribution in TAG of fish oil is also different
from that of marine mammal oil. In fish oil, PUFAs occupy the sn-2 position of
TAG, saturated and MUFAs the sn-1 position, and MUFAs the sn-3 position. In
marine mammals, however, the sn-1 and sn-3 positions are occupied by LC
PUFA such as EPA and DHA, and especially the sn-3 position, as noted earlier.
The sn-2 position is esterified to saturated fatty acids and especially to C16 and
C18 MUFAs (14, 112). The different distribution of fatty acids might be a factor
for lower oxidative stability of fish oils compared with seal oil (48).
REFERENCE 275
REFERENCE
1. R.G. Ackman, and F. Lamothe, in R. G. Ackman, ed., Marine Biogenic Lipids, Fats, and
Oils, vol. 2, 1989, CRC Press, Boca Raton, Florida.
2. F. Shahidi, in F. Shahidi, ed., Seal Fishery and Product Development, Science Tech, NF,
Canada, 1988.
3. P. Bjerregaard and J. Dyerberg, Int. J. Epidemiol., 17, 514519 (1988).
4. P. Bjerregaard, G. Mulvad, and H. S. Pederson, Int. J. Epidemiol., 26, 11821190 (1997).
5. R. Gales, D. Renouf, and E. Noseworthy, Can. J. Zool., 72, 545551 (1994).
6. E. Durnford and F. Shahidi, J. Amer. Oil Chem. Soc., 79, 10951102 (2002).
7. O. T. Oftedal, W. D. Bowen, and D. J. Boness, Physil. Zool., 66, 412436 (1993).
8. R. G. Ackman and R. D. Burgher, Can. J. Biochem. Physiol., 41, 25012505 (1963).
9. S. J. Iverson, Symp. Zool. Soc. Lond., 66, 263291 (1993).
10. U.N. Wanasundara, Ph.D. dissertation, Memorial University of Newfoundland, 1997.
11. C. Borch-Jensen, and J. Mollerup, Chromatographia, 42, 252258 (1996).
12. S. Hashimoto, S. Dayton, and J. C. Roberts, Jr., Comp. Biochem. Physiol., 20, 975
(1967).
13. F. Shahidi, J. Synowiecki, R. Amarowicz, and U. Wanasundara, in C. T. Ho, and T.G.
Hartaman, eds., Lipids in Food Flavors, ACS Symposium Series 558, American
Chemical Society, Washington, D.C., 1994, p. 233243.
14. G. K. Holmer, in R. G. Ackman, ed., Marine Biogenic Lipids, Fats, and Oils, vol. 1, CRC
15. R. F. Lee and J. S. Patton, in R. G. Ackman, ed., Marine Biogenic Lipids, Fats, and Oils,
vol. 1, CRC Press, Boca Raton, Florida, 1989.
16. E. H. Gruger, Jr., in M. E. Stansby, ed., Fish Oils, AVI Publishing, Westport, Connecticut,
1967.
17. I. A. Hansen and C. C. Cheah, Comp. Biochem. Physiol., 31, 757 (1969).
18. C. Litchfield, A. J. Greenberg, D. K. Caldwell, M. C. Caldwell, J. Sipos, and R. G.
Ackman, Conp. Biochem. Physiol. B, 50, 591 (1975).
19. C. Litchfield, A. J. Greenberg, and J. G. Mead, Cetology, 23, 10 (1976).
20. J. C. Nevenzel, Lipids, 5, 308 (1970).
21. J. R. Sargent, R. F. Lee, and J. C. Nevenzel, in P. Kolattukudy, ed., Chemistry and
Biochemistry of Natural Waxes, Elsevier, Amsterdam, 1976.
22. J. R. Sargent, Sci. Prog. Oxford, 65, 437 (1978).
23. J. H. Heller, M. S. Heller, S. Springer, and E. Clark, Nature, 179, 919920 (1957).
24. M. L. Karnovsky and W. S. Rapson, J. Soc. Chem. Ind. London, 66, 124 (1947).
25. E. Andre and H. Canal, Bull. Soc. Chim., 45, 498 (1929).
26. Y. Toyama and T. Isihawa, J. Chem. Soc. Jpn., 59, 1193 (1938).
27. T. Tsuchiya and A. Tanaka, Chem. Abstr., 50, 11035 (1956).
28. T. Tsuchiya, Chem. Abstr., 47, 12569 (1953).
29. T. Hashimoto and H. Mamuro, Chem. Abstr., 68, 51165 (1968).
30. N. R. Bottino, Lipids, 13, 18 (1978).
31. T. Tsuchiya and A. Kato, Chem. Abstr., 46, 11717 (1952).
32. M. Blumer and D. W. Thomas, Science, 148, 370 (1965).

33. R. G. Ackman, in R. G. Ackman, ed., Marine Biogenic Lipids, Fats, and Oils, vol. 1,
CRC Press, Boca Raton, Florida, 1989.
34. P. K. Christensen and R. A. Gehe, Acta Chem. Scand., 16, 1153 (1965).
35. H. O. Bang, J. Dyerberg, and N. Hjoorne, Acta Med. Scand., 200, 6973 (1976).
36. H. O. Bang, J. Dyerberg, and H. M. Sinclair, Am. J. Clin. Nutr., 33, 26572661 (1980).
37. F. Shahidi, U. N. Wanasundara, and R. Amarowicz, in O. Mills, H. Okai, A. M. Spanier,
and M. Tamura, eds. Chemistry and Novel Foods, American Chemical Society,
38. R. G. Ackman, C. A. Eaton, and P. M. Jangaard, Can. J. Biochem., 43, 15131520
(1965).
39. W. Yongmanichai and O. P. Ward, Prog. Biochem., 24, 117125 (1989).
40. E. Durnford and F. Shahidi, J. Amer. Oil Chem. Soc., 80, 405 (2003).
41. O. Grahl-Nielsen and D. Mjaavatten, Marine Biol., 110, 5964 (1991).
42. G. C. West, J. J. Burns, and M. Modafferi, Can. J. Zool., 57, 12491255 (1979).
43. F. R. Engelhardt and B. L. Walker, Comp. Biochem. Physiol. B, 47, 169 (1974).
44. G. C. West, J. J. Burns, and M. Modafferi, Can. J. Zool., 57, 189195 (1979).
45. R. G. Ackman, S. N. Hooper, and J. Hingley, Can. J. Biochem., 50, 833 (1972).
46. U. N. Wanasundara, and F. Shahidi, in F. Shahidi and K. K. Cadwallader, eds. Flavor and
Lipid Chemistry of Seafoods, American Chemical Society, Washington, D.C.,1997.
47. M. Aursand, L. Jorgensen, and H. Grasdalen, J. Amer. Oil Chem. Soc., 72, 293297
(1995).
48. T. Mag, Patent WO00/44862A1, PCT International Patent Application 2000.
49. A. Kendrick and N. Macfarlane, U.S. Patent 6623774B2, 2003.
50. H. Kimoto, Y. Endo, and K. Fujimoto, J. Amer. Oil Chem. Soc., 71, 469473 (1994).
51. K. Miyashita and T. Takagi, J. Amer. Oil Chem. Soc., 65, 1156 (1988).
52. F. Shahidi, F. Hamam, and M Ahmad, Khan, J. Amer. Oil Chem. Soc., in press.
53. F. Hamam and F. Shahidi, J. Agric. Food Chem., 52, 29002906 (2004).
54. F. Hamam and F. Shahidi, Food Chem., 91, 583591 (2005).
55. F. Hamam and F. Shahidi, J. Amer. Oil Chem. Soc., 81, 17 (2004).
56. F. Shahidi and U. N. Wanasundara, J. Food Lipids, 6, 159172 (1999).
57. R. C. Lindsay, Food Prod. Dev., 11, 9396 (1977).
58. D. B. Josephson, R. C. Lindsay, and D. A. Stuiber, J. Food Sci., 55, 59 (1985).
59. F. Shahidi and R. B. Pegg, J. Food Sci., 56, 15001504 (1991).
60. A. P. Bimbo and J. B. Crowther, Infofish Int., 6, 2025 (1991).
61. N. Haagsma, C. M. Gent, J. B. Luten, R. N. Jong, and E. Doorn, J. Amer. Oil Chem. Soc.,
59, 117118 (1982).
62. R. G. Ackman, P. J. Ke, and P. M. Jangaard, J. Amer. Oil Chem. Soc., 50, 1 (1973).
63. J. B. Brown and D. K. Kolb, Prog. Chem. Fats Lipids, 3, 5794 (1955).
64. M. I. Aveldano, M. van Rollins, and L. A. Horrocks, J. Lipid Res., 24, 83 (1983).
65. J. Halgunset, E. W. Lund, and A. Sunde, J. Chromatogr., 237, 496 (1982).
66. J. P. Roggero and S. V. Coen, J. Liq. Chromatogr., 4, 1817 (1981).
REFERENCE 277
67. R. O. Adolf and E. A. Emken, J. Amer. Oil Chem. Soc., 62, 1592 (1985).
68. W. B. Nilsson, E. J. Gauglitz, J. K. Hudson, V. F. Stout, and J. Spinelli, J. Amer. Oil
Chem. Soc., 65, 109 (1988).
69. M. A. Grompone, Fett. Wissenschaft. Technologie., 94, 388394 (1992).
71. U. N. Wanasundara and F. Shahidi, J. Amer. Oil Chem. Soc., 75, 945951 (1998).
72. H. Breivik, Modern Applications of Marine Oils, Toronto, Ontario, Canada, May 78.
1992.
73. G. G. Haraldsson, J. Amer. Oil Chem. Soc., 61, 219222 (1984).
74. H. Schlenk, J. Amer. Oil Chem. Soc., 38, 728736 (1961).
75. K. S. Markley, in K. S. Markey, ed., Fatty Acids, 2nd ed, Part 3, Interscience Publishers,
New York, 1964, pp. 19832123.
76. V. K. Mishra, F. Temelli, and B. Ooraikul, Food Res. Int., 26, 217226 (1993).
77. S. S. H. Rizvi, J. A. Daniels, A. L. Bernado, and J. A. Zollweg, Food Technol., 40, 5764
(1986).
78. T. Hamazaki, A. Hirai, T. Terano, J. Sajiki, S. Koado, T. Fujita, Y. Tamura, and
A. Kumagai, Prostaglandins, 23, 557567 (1982).
79. S. ElBoustani, C. Colette, L. Monnier, B. Descomps, C. A. Paulet, and F. Mendey,
Lipids, 22, 711714 (1987).
80. L. D. Lawson and B. G. Hughes, Biochem. Biophys. Res. Comm., 152, 328335
(1988).
81. D. D. Hong, Y. Takahashi, M. Kushiro, and T. Ide, Biochim. Biophys. Acta, 1635, 2936
(2003).
82. M. Schreiner, H. W. Hulan, E. RazzaziFazeli, J. Boehm, and C. Iben, Poultry Sci., 83,
462473 (2004).
83. S. P. J. N. Senanayake and F. Shahidi, Food Res. Int., 35, 745752 (2002).
84. P. M. Jangaard, Report No. 115, Program Development, Fisheries Branch, Halifax, Nova
Scotia, 1986.
85. R. G. Ackman, M. F. Lamothe, H. W. Hulan, and E. Proudfoot, N3 News, 111, 14
(1988).
86. M. L. Wahlqvis, Nutrition, 14, 627633 (1998).
87. P. R. C. Howe, Ann. NY Acad. Sci., 827, 339352 (1997).
88. I. Krishna Mohan and U. N. Das, Nutrition, 17, 126151 (2001).
89. T. Babcock, W. S. Helton, and N. J. Espat, Nutrition, 16, 11161118 (2000).
90. D. S. Kelly, Nutrition, 17, 669673 (2001).
91. D. P. Rose and J. M. Connolly, Pharmcol. Ther., 83, 217244 (1999).
92. T. Akihisa, H. Tokuda, M. Ogata, M. Ukiya, M. Iizuka, T. Suzuki, K. Metori, N. Shimizu,
and H. Nishino, Cancer Lett. 25, 913 (2004).
93. G. J. Anderson, W. E. Connor, and J. D. Corliss, Pediatr. Res., 27, 8997 (1990).
94. J. C. Hansen, H. Sloth Pedersen, and G. Mulvad, A Review Arctic. Med. Res., 53, 417
(1994).
95. E. C. Bonefeld Jorgensen, S. M. Moller, and J. C. Hansen, Int. J. Circumpolar. Health,
60, 2533 (2001).
96. B. Deutch, E. C. Bonefeld Jorgensen, and J. C. Hansen, Nutr. Res., 20, 10651077
(2000).
97. B. Osterud, E. Elvevoll, H. Barstad, J. Brox, H. Halvorsen, K. Lia, J. O. Olsen,
R. L, Olsen, C. Sissener, O. Rekdal, and E. Vognild, Lipids, 30, 11111118 (1995).
98. J. E. Kinsella, Food Technol., 40, 8997 (1986).
99. E. Vognild, E. Elvevoll, J. Brox, R. L. Olsen, H. Barstad, M. Aursand, and B. Osterud,
Lipids, 33, 427436 (1998).
100. H. Yoshida, I. Ikeda, M. Tomooka, M. Mawatari, K. Imaizumi, A. Seto, and H. Tsuji,
J. Nutr. Sci. Vitaminol., 47, 242247 (2001).
101. H. Yoshida, M. Mawatari, I. Ikeda, K. Imaizumi, A. Seto, and H. Tsuji, J. Nutr. Sci.
Vitaminol., 45, 411421 (1999).
102. I. Ikeda, H. Yoshida, M. Tomooka, A. Yoset, K. Imaizumi, H. Tsuji, and A. Seto, Lipids,
33, 897904 (1998).
103. J. Brox, K. Olaussen, B. Osterud, E. O. Elvevoll, E. Bjornstad, T. Brenn, G. Brattebo, and
H. Iversen, Lipids, 36, 713 (2001).
104. K. Yazawa, World Rev. Nutr. Diet., 88, 249252 (2001).
105. C. Ho, Omega 3: the Seal Connection 2003, Flanker Press Ltd., St. Johns, Newfound-
land, Canada.
106. M. Tsuji, S. Murota, and I. Morita, Prostaglandins, Leukotrienes and Essential
Fatty Acids, 68, 337342 (2003).
107. T. KanayasuToyoda, I. Morita, and S. Murota, Prostaglandins, Leukotrienes and
Essential Fatty Acids, 54, 319325 (1996).
108. S. Akiba, T. Murata, K. Kitatani, and T. Sato, Biol. Pharm. Bull., 23, 12931297 (2000).
109. J. M. Hodgson, M. L. Wahlqvist, J. A. Boxall, and N. D. Balaz, Am. J. Clin. Nutr., 58,
228234 (1993).
110. T. Rissanen, S. Voutilainen, K. Nyyssonen, T. A. Lakka, and J. T. Salonen, Circulation,
9, 72679 (2000).
111. E. AaesJorgensen, in M. E. Stansby, ed., Fish Oils, AVI Publishing, Westport,
Connecticut, 1967.
112. F. B. Padley, F. D Gunstone, J. L. Harwood, in F. D Gunstone, J. L. Harwood, and F. B.
Padley, eds., The Lipid Handbook, Chapman and Hall Ltd. New York, 1986.
113. C. Litchfield, A. J. Greenberg, C. E. Eaton, and R. G. Ackman, Lipids, 13, 860 (1978).
11
Fish Oils
R. G. Ackman
Canadian Institute of Fisheries Technology, Dalhousie University
Halifax, Nova Scotia, Canada
1. INTRODUCTION
At one time fish oils were low-cost industrial materials for the paint and lino-
leum industries. After World War II (WWII), these industries switched to chemicals
and plastics, and therefore, much information in older books became obsolete by
1960. Hydrogenation of fats to produce margarines and shortenings, starting about
1900, led to improved oil refining and better quality and included whale oils when
these animals were still plentiful. Two factors have recently impacted negatively on
large-scale and continued use of marine oils in our food supply.
Because of one of the earliest media-stimulated public health panics in the
late 1970s, that of the erucic acid (22:1) of rapeseed and mustard oils, alleged to
damage hearts, food use of partially hydrogenated fish oils in that form petered out
because of their content of both natural and artifact 22:1 isomeric fatty acids, as
described in Barlow and Stansby (1). As millions of healthy Germans and Poles
had thrived on rapeseed, and the fish-eating Scandinavians were universally healthy,
this fear was based on scanty evidence. The major result of this scare against
sources of very long-chain monoethylenic fatty acids was acceleration of develop-
ment of low-erucic-acid rapeseed oils such as canola (2). The desirable physical
properties of partially hydrogenated fish oils in some margarines and in shortening
for baking purposes continued but depended primarily on conversion of most of the
279
280 FISH OILS
natural cis-ethylenic bonds of the fatty acids to trans-ethylenic bond configurations,

rather than to saturated fatty acids (3). Recently, a media storm against trans-acids
followed initial scientific research on the health aspects of small changes in serum
cholesterol, despite long-term human exposure (4), and again partially hydroge-
nated fish oil use was condemned. Despite the adverse image, the baking industry
was conservative and this usage persisted in technically advanced countries such as
Denmark until recently (5). Dairy products and ruminant meat products continue to
include natural trans-acids (6).
In a positive turn of fate, by 1980, the observations of Dyerberg and Bang (7) on
the excellent cardiac health of the Eskimo population of Greenland, who had a high
intake of dietary omega-3 (n-3) fatty acids from seal and fish fats, were also noted
in the media. This publication in 1979 created an astounding public interest in fish
oils as nutrition supplements, usually taken in capsules. In 1985, another positive
media bombshell exploded when a long-term study in Zutphen in the Netherlands
showed that a large male population group had reduced their cardiovascular mor-
tality rate by eating fish regularly (8). Obviously, fish omega-3 (n-3) fatty acids
were beneficial.
By 1994, the United Kingdom had followed up on such reports and released an
official government medical report recommending that people eat fish at least twice
a week, one meal being of fatty fish (9). Belatedly the American Heart Association
released a Scientific Statement in 2002, with a similar recommendation (10).
Through this exciting period the dietary intake of fish in western society had
declined as other modern foods were easier to produce, store, and distribute. Fish
were also beginning to be more scarce as fisheries were overexploited and fish
populations diminished or were even wiped out.
In this review, only fish oils can be considered. The fat in edible parts of fish
ranges from about 16% down to 0.7%, the latter being almost exclusively the basic
muscle phospholipids (Figure 1). These lipids are an excellent source of the highly
unsaturated C20 and C22 omega-3 fatty acids of medical interest, and the content of
omega-3 fatty acids is stable at roughly 0.5-g/100-g muscle. However, it is the mus-
cle triacylglycerols that are variable for quantity and quality. This subject is conve-
niently reviewed for the food industry in one medium-length paper (11) in a journal
available in most major libraries, being a part of a whole issue (No. 4) on seafood.
A chapter on marine fatty acids in a reference book associating all types of fatty
acids with specific health problems could also be useful (12). As for the actual
health benefits of the omega-3 fatty acids, a special supplement issue of the
American Journal of Clinical Nutrition (13) is also highly recommended as it draws
together the thoughts and facts from several dozen distinguished authors active in
the field of omega-3 fatty acids.
It must be noted that the promotional bandwagon for fish oils and seafood
omega-3 fatty acids has been neatly hijacked by those oilseed companies offering
a vegetable oil omega-3 fatty acid, alpha-linolenic acid (18:3n-3). Only about
10% of this fatty acid in the diet survives immediate catabolism and is needed
for essential roles in the skin and elsewhere (14). Whether enough survives to
be elongated and to enrich brain and retinal tissues and thus to perform the vital
INTRODUCTION 281
Glycerol-Based Fish Lipid Classes
O
C
O
H2C O
HC O C
H2C
O
C
O
Triacyglycerols
H2
C
O
H2C O
HC O C
H2C
O
C
O
Diacyl-1-glyceryl ether
O
C
O
H2C O
HC O C
H2C
O
O
C O C CH2
H2 (+)
O ()
H3C N CH3
CH3
Phosphatidyl Choline
Figure 1. Three classes of lipids found in fish bodies and sharing in common glycerol fatty acid
linkages.
functions of the fish DHA or docosahexaenoic acid (22:6n-3) and EPA or eicosa-
pentaenoic acid (20:5n-3) is not known, but the distinction by chain lengths should
always be made. The term conditionally dispensable coined by one expert (15) is
perhaps suitable for 18:3n-3 at this time and in this context of recommending mar-
ine oils rich in long-chain n-3 fatty acids. Only human trials can be trusted to be
definitive, and modest doses of vegetable (C18) or fish (C20 C22) fatty acids for
12 weeks, although informative (16), cannot accurately predict the life expectancy
282 FISH OILS
and good health desired by the public. These trials are often linked to modifications
in diet, for example, reduction in intake of n-6 fatty acids. Critical reexamination of
published results is highly desirable and useful (17, 18).
Throughout the last 250 years, cod liver oil has steadily maintained a health role
well known to be inclusive of vitamins A and D. Our aging population is increas-
ingly faced with osteoporosis and mobilization of calcium intake to strengthen
bones, now a high priority in the adult population that requires Vitamin D, whereas
formerly the benefit target was rapidly growing children. Cod liver oil remains a
nutritional supplement staple for vitamins, but it also contains omega-3 fatty acids
(19). However, it is a specific staple outside the scope of this book except for a brief
mention on oil production and refining.
2. WHY DO WE STILL HAVE FISH OILS?
Fish oils are a byproduct of the production of fish meal. Both commodities are sub-
ject to price fluctuations that are only indirectly related. The meal has been histori-
cally used for terrestrial animal feedstuffs. The production process can be
condensed to apply to two basic functions: take fish or fish waste, cook it, and
squeeze it (Figure 2). Subsequent steps are varied for each commodity, but they
essentially are well described by Young (20), whereas the desirable properties of
fish meal are concisely described by Bimbo and Crowther (21). Similar summaries,
especially on sources and production, are provided by various chapter authors in a
book on pelagic fish (22), because most fish caught for meal and oil are pelagic
(migratory) and subject to seasonal exploitation. The principal competition to
fish meal is soybean meal and other oilseed meals (23).
Fish Oil Manufacture
Fish Grind and Cook
Press
H2O
Fish
Oil + Stickwater Press Cake Meal
Drying
Centrifuge Solids
Evaporator
Crude Fish Oil H2O
Figure 2. A simplified diagram of the fish meal manufacturing process. Shading denotes the
principle products.
WHY DO WE STILL HAVE FISH OILS 283
The gloomy chronology in Section 1 for food and health aspects of fish oils
since WWII would have ruined most basic industries, but as a byproduct of fish
meal production, that of oil has continued. The human (and pet) nutritional
supplement markets would absorb only a small amount of the total world pro-
duction, which has remained at approximately 1.21.4 million tons for the last
decade (23).
Aquaculture has provided part of the solution, which is particularly true for sal-
monids, and to trout farmed in freshwater, a very large amount of the former indus-
trial grade (unrefined) production of fish oil goes to Atlantic salmon (Salmo salar)
production in places as far away from their origin in the North Atlantic as New
Zealand and Chile. Other species are now coming into aquaculture production,
usually higher priced fish such as Atlantic halibut, grouper, sea bass, tilapia, and
so on. Fish oil is not just a cheap fat, as most of the fish will grow on diets of
fish meal plus other fats, but around 12% of the total dietary fatty acids should
be the long-chain polyunsaturated fatty acids (20:5n-3 and 22:6n-3) that are basic
to the functional membranes of new cells when combined with the protein of the
fish meal. Fish meal usually contains approximately 810% of fat, and it is often a
good source of this minimum requirement of omega-3 fatty acids. Excessive fat in
fish diets is thought to spare protein degradation for energy, which results in more
growth, and the 20% of total dietary fat for the salmon diets common a decade ago
is now frequently replaced with a total of 2628%, the so-called high-energy diets.
Peru has been a major producer of industrial grade fish oil for the aquaculture
industry, but the anchovy fished there fails to return to the coast occasionally for
Pacific Ocean and climate reasons (an EI Nino event), and reduced production
can drive the world price of fish oil to above $500 (US)/tons compared with
more normal prices of $250$400/ton (23).
The new factor in the fish oil industry is as a built-in supplement in the human
diet instead of a separate food additive. It is in the form (usually) of highly refined
fish oil added in a microencapsulated format. The oils used must be highly refined
to meet Food and Drug Administration (FDA) standards, especially in the United
States, where menhaden oil was the first oil approved for this purpose. There, the
intake of the oil must not exceed 3 g/day of the two fatty acids 20:5n-3 and 22:6n-3
in a designated list of typical foodstuffs. Although ethyl esters and concentrates of
oils and esters will eventually be approved, it is almost certain that the starting
materials will be those fish oils essentially very low in 20:1 and 22:1. As will be
discussed below, these oils typically have 20:5n-3 > 22:6n-3, i.e., EPA > DHA in
the popular terminology. The initial omega-3 publicity in the 1980s favored EPA
for superior blood vessel function, but gradually this public image has changed
to favor DHA as superior for heart muscle function and neural problems. Infant
nutrition is yet another popular field for debate on omega-3 fatty acids, and it
should be noted that the Martek Biosciences Corp., Columbia, nw, was the first
to develop algal (i.e., vegetarian) sources of 22:6n-3 and 20:4n-6 (arachidonic
acid) as preformed long-chain fatty acids representative of human breast milk
when added to enriched infant formulas. Revenues for this company were forecast
to exceed $100 million in 2003.
284 FISH OILS
3. FISH OIL FATTY ACIDS AND GAS-LIQUID

CHROMATOGRAPHY
Fish oils mostly contain triacylglycerols of fatty acids (Figure 1), recovered by
fairly simple technology (Figure 2) from the whole bodies of fish, often from spe-
cies considered inedible in contemporary Western society. In fact, many exceptions
to these generalizations exist, because in addition to the triacylglycerols, there can
be wax esters (1/2 fatty alcohols), diacylglyceryl ethers (2 fatty acids, Figure 1),
cholesterol and cholesteryl esters (Figure 3), and even the hydrocarbon squalene
(Figure 4). Also, the marine mammals are to be considered. The depot fats of
baleen whales and seals yield oils similar to the above general description of fish
oil including fatty acid composition, but the triacylglycerols differ in fatty acid
molecular arrangements and warrant a separate discussion. The depot fats of the
toothed whales can include wax esters (sperm whales), or even triacylglycerols
and wax esters incorporating short-chain fatty acids such as isovaleric (the dolphins
and similar small species). These topics are covered elsewhere (24). Marine inver-
tebrates also will have to be excluded from this discussion as there is no large-scale
or industrial use of their triacylglycerols. Krill oil produced from small Antarctic
crustacea such as Euphausia superba is now offered in small amounts for the nutri-
tional health product market, but the investment prospects are daunting (25). The
two volumes of Marine Biogenic Lipids, Fats and Oils offer lipid class and lipid
Fish Oil Miscellaneous Natural Lipids
O
HO C
Free Fatty Acid
O
HO R C O
Cholesterol and/or Ester
H3C (CH2)x C O C (CH2)x CH3

H2
O
Wax Ester
Figure 3. Lipids that may be found in marine oils. Free fatty acids are artifacts of postmortem
processes in fish bodies and may range up to 5% or even 10% of oils. Cholesterol and its esters
are usually in the 0.51.0% range, but wax esters may be major oil components in oils from
certain fish.
FISH OIL FATTY ACIDS AND GAS-LIQUID CHROMATOGRAPHY 285
Neutral Fish Oil Compound
CH3 CH3 CH3 CH3

HO
Phytol
CH3 CH3 CH3 CH3
Phytane
CH3 CH3 CH3 CH3
Pristane
CH3 CH3 CH3

CH3
CH3
CH3 CH3 CH3
Squalene
Figure 4. Phytol and phytol-related hydrocarbons, pristane and phytane, are associated and
may be found in some fish oils. Phytanic, pristanic, and 4,8,12-trimethyltridecanoic acids are
common fish oil components derived from phytol. Squalene is usually of animal origin and a
feature of some shark liver oils.
component composition information on these topics from resource and biochemical

points of view (24).
Although traditional quality factors such as iodine value, free fatty acids
(Figure 3) and unsaponifiable content (Figures 3 and 4) are still important in trading
fish oils, the fatty acid composition determined by gas-liquid chromatography
(GLC) of methyl esters is now frequently required. Figure 5 is a typical example
of what is desirable, and it may help to explain the fatty acids to be expected. This
analysis is on a special class of liquid phases based on polyglycols (e.g., SUPEL-
COWAX-10, DURA-WAX, Stabilwax, Omegawax-320, etc.), which limits even
chain length overlaps to two C24 fatty acids and 22:6n-3. These chain lengths
are marked accordingly.
Figure 6 is an extensive list of the fatty acids typical of the triacylglycerols of
fish oils and includes one baleen whale oil. These were all of interest to industrial
and food fats and oils companies two decades ago (1), but even later were of limited
interest in human nutrition (26). For practical purposes, marine oils can be defined
by 12 fatty acids that add to about 90% of the array of peaks in Figure 5 and deter-
mine the properties of a given oil. These are (in a shorthand giving the chain-length,
286 FISH OILS
Figure 5. Analysis of menhaden (fish) oil on an Omegawax-320 capillary column. Equipment:

Varian 3400 GLC, splitless injection helium carrier gas. Initial temperature 69 C for 1.4 min;
ramp to 170 C at 50 C/min; hold for 8 min at 170 C; ramp to 220 C at 3 C/min; hold 15 min;
total time 43 min. Note FAME chain lengths below baseline. Peaks identified are as follows:
1 14:0; 2 15:0;3 16:0; 4 16:1n-7; 5 16:16:2n-4; 6 16:3n-4; 7 16:4n-1; 8 18:0;
9 18:1n-9; 10 18:1n-7; 11 18:2n-6; 12 18:3n-3; (followed by 18:3n-1); 13 18:4n-3
(followed by 18:4n-1); 14 20:0; 15 20:1n-9 (followed by 20:1n-7); 16 20:4n-6; 17 20:4n-
3; 18 20:5n-3; 19 22:1 group; 20 21:5n-3; 21 22:5n-3; 22 22:6n-3 (with 24:0 pre-
ceding and 24:1n-9 following). Analysis time, 30 min. Reproduced by permission of Anal.
Chem.Acta.
number of cis-methylene-interrupted ethylenic bonds, and position of the ethylenic

bond nearest the methyl end of the chain):
14:0 16:1n-7 18:2n-6

16:0 18:1n-9 18:3n-3
18:0 20:1 18:4n-3
22:1 20:5n-3
22:6n-3
After considering the many analyses of fish oils available, the author concluded that
there was only one basic fatty acid composition of fish oils from coldwater or
from northern latitudes (27). Generally this is typified by menhaden oil, a species
that feeds exclusively by filtering phytoplankton out of the ocean water in the Gulf
of Mexico or in the Atlantic Ocean off the east coast of the United States or the
anchovy oil from Peru (Figure 6), which also feeds close to the plant base of the
food chain.
Menhaden oil has a fatty acid composition that provides a good example of the
basic marine fish oil fatty acid system. For example, it is characterized (Figure 6)
by low values of 20:1 and particularly 22:1. The origin of these two fatty acids has
been discussed in reviews of freshwater lipids (28) as well as of marine lipids (29,
30). In principle the extension by one acetate unit of the plentiful 18:1n-9 and
18:1n-7 will give 20:1n-9 and 20:1n-7, and a second step leads to some 22:1n-9
and 22:1n-7. Generally the process stops there, and only small amounts of these
are found relative to 22:1n-13 and especially 22:1n-11. In addition to a little
24:0, a more obvious peak for 24:1 is found, which accounts for that in Figure 6.
The marine 24:1n-9 includes nervonic acid, functional in various organs of animals,
and may accumulate from food fatty acids. Other isomers are possible. The 22:1
peak is actually resolved by efficient open-tubular GLC into four distinct peaks
(Figure 7), of which the major peak is primarily 22:1n-11 but also includes some
22:1n-13, followed by 22:1n-9, 22:1n-7, and 22:1n-5. Similarly the 20:1 array
(Figure 7) includes a frontal shoulder of 20:1n-11, sometimes difficult to see on
the dominant 20:1n-9 peak, which is followed by 20:1n-7 and 20:1n-5 and some
other peaks of non-methylene-interrupted dienoic (NMID) acids that might occur
as discussed below.
The origin of the unexpected 22:1n-13 and 22:1n-11 fatty acid isomers (Table 1)
is now not a mystery (29, 30). The pioneering work of many scientists some dec-
ades ago on marine lipids led to those of a group of tiny marine crustaceans called
copepods. They found that one major lipid class in several species known to be
important as food for fish in the North Atlantic was wax esters (31). In fact, lipids
of a copepod sample examined in Halifax contained 61.2% wax esters and only
31.6% triacylglycerols (29). The distribution of ethylenic bond positions in the
copepod lipid wax ester fatty alcohols is compared with the alcohols recovered
from the body depot fats of several regional fish species known to feed directly
on copepods in Table 1. It is believed that the copepod fatty alcohols are converted
directly by the fish to the corresponding fatty acids, which accounts for the very
high proportion of 22:1 usually observed in fish feeding at this trophic level
(Table 1). However, it is fair to point out that the copepod fatty acids of both the
triacylglycerols and the wax esters had modest contents of the same isomers (29).
The dominance of the unusual 22:1 wax ester isomer n-11 in the copepod is not
readily explained, but it may be based on physical properties such as melting point,
specific gravity, and so on.
It will be noticed that in most tables of fish oil, fatty acid compositions 20:1
and 22:1 are simply listed as such and do not include isomer details. They are
perfectly acceptable in diets for aquaculture fish (32). As reviewed earlier, their
presence in fish oils did contribute to false alarms about 22:1 fatty acids and alleged
heart damage in animals consuming rapeseed oil or fish oils, the latter usually
being in partially hydrogenated form for margarines and shortening. This was inde-
pendent of the much more recent possibilities of trans-fatty acids of any origin
leading to cholesterol and atheroma problems in the human population (33, 34).
288 FISH OILS
Fatty acid composition of six samples of fish oil and one whale oil in commercial trade. Unpublished data
reproduced by courtesy of W. Schokker and H. Boerma, Unilever Research, Vlaardingen.
Herring Anchovy Whale Pilchard Sardine Menhaden Pilchard

North Peru Antarctic South Portugal U.S.A ?
Sea Africa
12 0.10 0.10 0.20 0.20 0.10 0.15 0.10
14 6.10 7.45 7.45 7.75 6.70 7.30 7.30
14:1 0.15 0.75 0.15
16:Branched 0.40 0.40 0.40 0.40 0.50 0.45 0.55
15:0 0.40 0.60 0.65 0.40 0.75 0.65 0.60
16:0 10.75 17.45 13.40 15.65 17.80 19.00 15.60
16:1 7.30 9.00 10.50 8.50 6.00 9.05 9.00
16:2 7 0.20 0.20 0.20 0.55 0.40 0.50 0.40
16:2 4 0.40 1.00 0.65 1.45 0.65 1.25 1.55
16:3 4 6.70 2.05 0.10 2.00 0.40 1.45 1.70
16:3 3 0.20 0.20 0.20 0.15
16:4 4 0.10 0.10 0.15 0.20
16:4 1 1.20 2.45 0.95 3.20 1.60 2.30 2.60
17:Br 0.30 0.35 0.25 0.25 0.20 0.20 0.15
17:0 0.35 0.55 0.95 0.80 0.80 0.90 0.85
17:1 0.30 0.25 0.30
18:Br 0.80 0.70 1.10 0.60 0.60 0.45 1.00
18:0 1.40 4.00 2.70 3.65 3.60 4.20 3.45
18:1 10.30 11.55 27.60 9.25 13.00 13.20 10.40
18:2 9 tr. 0.10 0.10 0.15 0.15 0.30 0.20
18:2 6 0.95 1.20 1.90 0.80 1.20 1.30 1.30
18:2 4 0.10 0.60 0.20 0.50 0.30 0.40 0.50
18:3 6 0.05 0.30 0.20 0.35 0.20 0.25 0.30
18:3 tr. 0.20 0.10 0.30 0.10 0.30 0.20
18:3 3 2.00 0.75 0.85 0.45 1.00 1.30 0.65
18:4 3 3.15 3.05 1.05 2.05 3.15 2.75 2.65
18:4 0.15 0.20 0.20 0.15 0.10 0.15 0.20
19:Br 0.10 0.20 0.20
19:0 0.20 0.10 0.60 0.20 0.40 0.40 0.10
19:1 0.10 0.40
20:0 0.10 0.30 0.20 0.60 0.40 0.35 0.30
20:1 13.40 1.55 6.75 2.50 4.30 2.00 1.45
20:2 9 0.30 0.10 0.40 0.15 0.45 0.15
20:2 6 0.15 0.35 0.15 0.25 0.20 0.35 0.30
20:3 6 0.10 0.10 0.20 0.30 0.10 0.15 0.20
20:3 3 0.30 1.10 0.60 1.35 0.85 0.80 1.00
20:4 6 tr. 0.10 0.25 0.10 0.10 0.15 0.15
20:4 3 0.75 0.70 1.30 0.70 1.05 1.35 0.80
20:5 3 7.45 17.00 4.70 19.30 11.00 11.00 18.30
21:0 0.10 tr. 0.05 0.15 0.10 0.05 0.10
21:5 2 0.25 0.70 0.15 0.90 0.50 0.60 0.90
22:0 0.05 0.05 0.10 0.20 0.20 0.20 0.15
22:1 21.25 1.15 2.40 3.10 3.80 0.55 1.55
22:2 0.20 0.10 0.05 0.05 0.10 0.20 0.10
22:3 3 0.15 0.25 0.15 0.15 0.15 0.20
22:4 3 0.25 0.55 0.20 0.40 0.70 0.50 0.60
22:5 3 0.75 1.60 2.40 2.35 1.30 1.90 1.80
22:6 3 6.75 8.75 5.70 6.45 13.00 9.10 9.60
23:0 0.10 0.05 0.05 0.10 0.10 0.10 0.15
24:0 0.15 0.05 tr. 0.15 0.10 0.15 0.10
24:1 0.75 0.50 0.30 0.50 0.60 0.35 0.70
Wijs I.V. 135.7 181.0 121.7 182.0 169.7 162.1 189.4
GLC- I.V. 138.0 181.5 122.1 183.9 170.0 161.5 190.7
Figure 6. Reproduction of a personal communication to R. G. Ackman from scientists of

Unilever Research, Vlaardingen, the Netherlands, as published in (1). Reproduced by
permission of Academic Press.
Figure 7. Part of gas-liquid chromatographic analysis of methyl esters of cod liver oil on an
Omegawax-320 column, 0:32 30 0:25. Temperature 160 C for 8 min, 3 C/min to 220 C,
hold. Peaks identified are as follows: 1 18:4n-3; 2 18:4n-1, unmarked probably 18:5n-3;
3 20:0; 4 20:1n-9 with frontal shoulder of 20:1n-11 after peak 3; 5 20:1n-7; 6 20:1n-5;
unknown; 7 20:2n-6; unknown; 8 20:3n-6; 9 20:4n-6; 10 20:3n-3; 11 20:4n-3;
12 20:5n-3; small peak for 22:0; 13 - 22:1n-13 22:1n 11; 14 22:1n-9; 15 22:1n-7;
16 22:1n-5 (Ackman, unpublished).
As of the time of writing (mid-2003), the FDA has announced that trans-acid
contents of foods will soon be required on food composition labels. Fish oils,
now omega-3 nutritional supplements in some foods, are essentially excluded
from such considerations because they contain almost exclusively cis-ethylenic
bonds.
290 FISH OILS
TABLE 1. Principal n-3 Fatty Acids, Saturated, and Monoethylenic Fatty Acid Isomers
(w/w%) in Triacylglycerols and Wax Esters of Copepods and Commercial Oils of Pelagic
Species of North Atlantic Fish Likely to be Consuming Copepods.
Commercial Oils
Copepod Capelin Mackerel Herring
Triacylglycerol Wax Esters Total Acids Wax Esters Total Acids Total Acids
14:0 19.84 38.42 7.85 5.23 7.81 8.77
16:0 28.98 11.15 8.81 8.36 15.93 14.84
18:0 1.04 0.35 0.72 1.03 1.73 0.97
16:1n-9 0.05 0.35 0.03 1.58 0.29 ND
16:1n-7 8.89 12.33 15.42 18.16 8.20 7.22
16:1n-5 0.73 1.04 0.73 0.12 0.54 0.52
18:1n-9 3.14 3.46 4.40 5.95 8.61 12.27
18:1n-7 0.91 0.56 3.43 1.69 3.78 3.66
18:1n-5 0.41 0.15 0.62 0.65 0.54 0.64
18:2n-6 0.97 0.83 0.78 0.86 1.28 0.78
18:3n-3 1.08 1.00 0.20 0.36 0.99 0.39
18:4n-3 3.23 5.63 1.36 1.87 2.47 0.93
20:1n-11 0.33 0.36 1.20 0.46 0.24 0.50
20:1n-9 4.12 4.37 14.53 9.34 10.59 14.37
20:1n-7 0.34 0.55 1.84 0.92 1.13 0.94
20:1n-5 0.02 0.01 0.23 0.10 0.09 0.19
20:4n-6 0.29 0.19 0.29 0.60 0.36 0.24
20:5n-3 8.38 5.81 9.35 22.38 7.84 2.85
22:1n-11(13) 5.16 4.59 17.45 5.90 12.74 20.92
22:1n-9 0.34 0.65 1.70 0.92 1.00 1.36
22:1n-7 0.11 0.11 0.42 0.19 0.19 0.33
22:5n-3 0.60 0.23 0.60 0.74 0.57 0.37
22:6n-3 4.90 0.64 2.70 5.32 7.66 2.70
24:1 0.49 0.24 0.59 0.13 0.69 0.52
ND not detected.
From Ratnayake and Ackman (30).
4. SATURATED, ISOMERIC MONOENOIC, AND UNUSUAL

FATTY ACIDS
The basic fish oil (27) also included a generous amount of saturated fatty acids.
As can be seen from Figure 5 and Table 1, the saturated fatty acids are dominated
by the 16:0 (palmitic acid), usually accompanied by about half as much or less of
14:0 (myristic acid) and much less of 18:0 (stearic acid). Usually the saturated fatty
acid totals are at least 20%, especially as the odd chain (15:0, 17:0) and methyl-
branched (iso, anteiso, pristanic, phytanic) fatty acids (compare Figure 4) are satu-
rated and will total around 23%. An unsaturated peak that is often observed is
17:1n-8, which is roughly equal to 17:0. The details of these peaks are discussed
in other publications, but those researchers attempting modern open-tubular gas
chromatography analyses should be aware of their presence and influence on
peak identification and quantitation. As can be seen from Figure 6, there is an
SATURATED, ISOMERIC MONOENOIC, AND UNUSUAL FATTY ACIDS 291
inverse relation among classes of fatty acids, so oils rich in 20:1 and 22:1 generally
have lower levels of saturated fatty acids.
The peaks for C14 monounsaturated fatty acids follow 14:0 and are usually a
jumble of peaks for 14:1 isomers, mixed up with those for iso and anteiso 15:0
and the isoprenoid 4,8,12-trimethyltridecanoic acid, followed reasonably clearly
by that for 15:0 (Table 2). Little interest exists in these details, and the only obvious
next peak before 16:0 should be iso-16:0, and sometimes another isoprenoid acid,
pristanic or 2,6,10,14-tetramethylpentadecanoic, is found just ahead of 16:0.
TABLE 2. Fatty Acid Composition (w/w%), with Relative Retention Times, for Japanese
Sardine Oil (36), Compared with That of Triacylglycerols of Cultured Cells of the Marine
Diatom Phaeodactylum tricornutum (39).
Peak No. Fatty Acid RRT Sardine P. tricornutum
1 10:0 0.057 0.02

2 12:0 0.119 0.11 0.4
3 iso-13:0 0.144
4 13:0 0.166 0.02 1.1
5 iso-14:0 0.208
6 14:0 0.245 7.25 6.9
7 14:1 (n-9) 0.260 0.25
8 14:1 (n-7) 0.275 0.06
9 14:1 (n-5) 0.287 0.07
10 iso-15:0 0.296 0.15
11 anteiso-15:0 0.313 0.08
12 15:0 0.348 0.31 2.0
13 iso-16:0 0.421
14 anteiso-16:0 0.437 0.08
15 16:0 0.502 19.42 21.2
16 16:1 (n-11) 0.529 0.84
17 16:1 (n-9) 0.541 4.4
18 16:1 (n-7) 0.564 8.64 23.0
19 iso-17:0 0.600 0.07
20 anteiso-17:0 0.631 0.32
21 16:2 (n-7) 0.643 0.05
22 16:2 (n-4) 0.686 1.23 3.4
23 17:0 0.704 0.33
24 17:1 (n-10) 0.732 0.52
25 16:3 (n-4) 0.762 1.57 3.3
26 17:1 (n-8) 0.785 0.13 0.4
27 17:1 (n-6) 0.830 0.05
28 iso-18:0 0.844 0.09
29 anteiso-18:0 0.875 0.03
30 16:4 (n-1) 0.892 2.55 0.2
31 18:0 1.000 2.62 2.3
32 18:1 (n-13) 1.041 0.11
33 18:1 (n-9) 1.110 6.38 6.5
34 18:1 (n-7) 1.131 3.04 0.8
35 18:1 (n-5) 1.169 0.23 1.2
(Continued)
292 FISH OILS
Peak No. Fatty Acid RRT Sardine P. tricornutum
36 iso-19:0 1.222
37 18:2 (n-6) 1.291 0.85 2.5
38 18:3 (n-6) 1.371 0.47 0.3
39 19:0 1.415 0.14
40 19:1 (n-8) 1.515 0.11
41 18:3 (n-3) 1.575 0.40 1.3
42 18:4 (n-3) 1.730 2.06 0.2
43 18:4 (n-1) 1.768 0.22
44 20:0 1.995 0.11 0.1
45 20:1 (n-11) 2.142 2.37
46 20:1 (n-9) 2.176 1.33
47 20:1 (n-7) 2.235
48 20:2,5,11 2.403
49 20:2 (n-6) 2.783 0.4
50 20:4 (n-6) 2.928 1.26 0.3
51 20:4 (n-3) 3.396 0.74 0.4
52 20:5 (n-3) 3.608 17.01 8.7
53 22:1 (n-11 13) 4.205 2.09
54 22:1 (n-9) 4.288 0.27
55 22:2,7,11 4.902
56 22:2,7,13 4.985
57 21:5 (n-3) 5.158 0.60
58 22:5 (n-3) 7.150 2.17 0.9
59 22:6 (n-3) 7.510 10.09 2.3
60 24:1 (n-9) 8.408 1.09
The C16 polyunsaturated fatty acids are often confusing because two series can
coexist and overlap in GLC analyses, basically the familiar n-3 and n-6 series with
16:2n-6, 16:3n-3, and 16:4n-3. Superimposed on these acids are members of an n-1,
n-4, and n-7 series. The latter are well documented (35), and they may be prominent
in fish lipids where algal fatty acids are deposited more or less directly, for example,
in Japanese sardine oil (Table 2). All should be included in extensive tabulations of
marine fatty acids, and their methyl ester peak relative retention times for the low-
polarity GLC phase SILAR-5CP (36) are included in Table 2, which will clarify
their positions after 16:1n-7 for similar GLC columns, but on higher polarity
GLC columns they will overlap with C18 fatty acids. In default of GC-MS, older
techniques of plotting or separation factors may help in identifications (37, 38),
although these require an isothermal GLC analysis. To illustrate how the n-1, n-
4, and n-7 fatty acids can be transferred to fish oils, the fatty acids in the triacylgly-
cerols of a well-known unicellular alga, Phaeodactylum tricornutum, are included
in Table 2, because it is a basic diatom in that geographic location (39). It may be
noted that 16:4n-1 is a minor fatty acid presumably because very little 18:4n-3
exists. The 16:4n-1 would be generated in algae by the enzymatic process that pro-
duces 18:4n-3, but acting on a C16 chain length instead of a C18 chain length.
POLYUNSATURATED FATTY ACIDS 293
Other sources must provide the amount obvious in the sardine oil, and similar or
lesser contents are common in other fish oils.
Curiously, in fish oils, 18:4n-1 is frequently clearly visible after the peak for
18:4n-3 (No. 12) in Figure 5 and is usually about one-quarter of its size. Another
unusual fatty acid of algal origin is 18:5n-3 (40). It is known from GLC analyses of
both phytoplankton and zooplankton lipids. In Figure 7, it is probably the small
sharp peak between peak 2 (18:4n-1) and peak 3 (20:0), approximately as for the
moderately polar SILAR-5CP liquid phase of (40). All unusual fatty acids should
be neatly handled by the mammalian body degradation process and even by those
in fish, and they are seldom reported in fish oils. Incidentally, peak 3 is wider than
expected, a characteristic of methyl esters of saturated fatty acids in such analy-
ses. Of the other unusual fatty acids in fish oils, the 21:5n-3 (20 in Figure 5) is
common and well documented (41). It is of interest if 23:0 (tricosanoic acid methyl
ester) is used as an internal standard in GLC (42). The latter may coincide with
21:5n-3, and the Omegawax-320 was a slight modification of SUPELCOWAX-10
to avoid this problem. On the other hand, the NMID referred to earlier (both C20
and C22) appear in mollusc lipids, possibly in physical imitation of the polyunsa-
turated fatty acids of membranes (43), and they are not apt to be observed in indus-
trial pelagic fish oils. On tropical reefs or among fish-consuming molluscks, they
should be considered as likely if the normal resolution among the later-eluting
C20 and C22 monoethylenic isomers (n-7, n-5) is obscured on polyglycol capillary
columns.
5. POLYUNSATURATED FATTY ACIDS
The C18 polyunsaturated fatty acids include the familiar terrestrial fatty acids
18:2n-6 and 18:3n-3, along with some 18:3n-6, but in most marine fish oils, these
are 2% of each (Figure 6, Table 1). The amount of 18:4n-3 is in the 24% range.
Among the C20 polyunsaturates, there will be found very small proportions of
20:2n-6, 20:3n-6, 20:3n-3, and 20:4n-3. The 20:4n-6 (arachidonic acid) peak is gen-
erally the same size as these except in tropical fish lipids, where it can be more
important. The latter are, however, not usually commercial fish oil sources.
Although of nutritional importance in mammals, 20:4n-6 is usually grossly over-
shadowed by the C20 and C22 omega-3 polyunsaturated fatty acids in marine
lipids. One reason for avoiding higher polarity gas-liquid chromatographic columns
is that with the chain length overlap, this usually puts 22:1 and 20:4n-6 in juxtapo-
sition or coincidence.
The polyunsaturated fatty acids were intimately associated with early (ca. 1960)
attempts to lower serum cholesterol with varying degrees of success (44). In that
particular study, with ethyl ester concentrates given to provide up to 4 g/day n-3
PUFA, a side effect of an increased feeling of well-being coupled with improved
cerebration was reported only for the group receiving omega-3 fatty acids. About
the same time, up to 26 mL/day of seal oil with a content of about 5 g of omega-3
fatty acids was fed for cholesterol-lowering effect without ill effects (45). Today
294 FISH OILS
this simplistic dietary approach to one cardiac risk factor, cholesterol, is no longer
acceptable in some circles because of the multiplicity of risk factors for heart dis-
ease (46). Essentially this recent review rules out arachidonic acid (20:4n-6) as
having little effect, and it considers EPA (20:5n-3) as being more effective than
DHA (22:6n-3) in the lowering of serum triacylglycerols, another independent
risk factor (47). Another study also suggested that EPA was more effective than
DHA in lowering blood pressure (48).
Tables 1 and 2 and Figures 5 and 6 show a third long-chain omega-3 fatty acid,
7,10,13,16,19-docosapentaenoic acid (or 22:5n-3), one of two DPA isomers. The
other is the n-6 isomer, 22:5n-6, eluting just before 22:5n-3 on GLC. In Figure 5,
the 22:5n-3 is peak 21 and a 22:4n-3 may be present after peak 20. In 1996, two
papers (49,50) gave this 22:5n-3 fatty acid considerable potency in respect to
endothelial cells. Considering that the related 20:5n-3 was originally considered
to keep the blood vessel walls elastic through providing a prostaglandin (51),
this could well be a beneficial role for EPA and even DPA in circulation separate
from the heart, but research on the relative roles of the three omega-3 fatty acids in
the aortic endothelium seems to have fallen into abeyance, perhaps because they are
often said to be freely interconverted (52). This statement is not necessarily true,
but the conversion of EPA to n-3 DPA and the reverse process, by one acetate
unit, is generally accepted. The critical step, the conversion of 22:5n-3 to 22:6n-
3, is not so easy, involving elongation to 24:6n-3 and peroxisomal shortening to
22:6n-3, and some biochemists now think that exogenous supplies of DHA are
the preferable route to increasing the amount available in the body, especially in
late pregnancy and lactation. This digression into the biochemistry of the logical
beneficiaries of fish oils, humans, cannot cover the maternal/infant problems dis-
cussed at length by various authors in Ref. 13, a topic not without controversy
(53, 54).
More emphasis has been placed on the loss of heart function taking place
through arryhthmia (55), and it is possible that DHA is the more functional ome-
ga-3 fatty acid in the heart muscle. Unfortunately, the available fish oils divide into
the menhaden/anchovy type, richer in EPA, an omnivorous group such as cod with
EPA DHA, and the tuna oils, both body and orbital, with 2025% DHA. In fact,
the first research on the beneficial aspects of omega-3 fatty acids in arrhythmia was
conducted in Australia (56), conveniently near an Asian source of tuna oil rich in
DHA, which lead to misunderstandings over a role for this omega-3 fatty acid in
particular for preventing arrhythmia.
The health and welfare interest in long-chain omega-3 fatty acids inevitably
raises the question of where do they come from and are they safe. The latter
question applies to oils and will be addressed in a production and quality section,
but in reality, most fatty acids are of plant origin and perfectly safe.
The phytoplankton produce toxins dangerous even to fish, usually observed as
red tides (57). These are fish kills in the oceans and near the shores, and fish
oils are not made from fish found dead. A different issue is shellfish toxins where
the digestive assimilation of unstable algal toxins does not kill the host (58), but fish
oils are not made from such filter-feeding animals.
POLYUNSATURATED FATTY ACIDS 295
Origin of Marine Fatty Acids
EPA > DHA Phytoplankton 18:2n-6 18:3n-3
Shellfish Shellfish
Crustacea Crustacea
Fish
Energy
TG PL
EPA > DHA DHA > EPA
Human Diet Seals Seabirds
circulatory
DHA EPA Prostaglandin Group 3
system
Neurological systems? TXA3 PGI3

Retina?
Cell membrane?
Figure 8. Origin of unsaturated fatty acids in phytoplankton, followed by discrimination
by invertebrates leading to accumulation of EPA and DHA in oils and eventually in higher
animals.
Each new discovery of unusual fatty acids in marine organisms leads to concerns
that eventually dissipate. An example is the furanoid fatty acids that were once well
documented as common in most oils (26). More recent papers and books on omega-
3 fatty acids simply ignore the subject.
Figure 8 is a simplification of much work by many scientists over decades. The
phytoplankton in the ocean produce all fatty acids necessary for fish oil and with
somewhat similar compositions in places as remote as Australia and Scotland. Spe-
cific differences exist among them, but broadly the fatty acid patterns are related to
colors familiar to people who see the macrophytes growing on the edges of the sea:
red, green, and brown. However, even the accumulation of the amount of triacyl-
glycerols is controlled by the available nutrients, especially nitrogen, and light
intensity (59). The 22:6n-3 is not universal in phytoplankton, but it is found all
over the world in benthic algae (60) or phytoplankton (61,62). For unknown rea-
sons, fatty acid phytoplanton biosynthesis often stops at 20:5n-3, which was pre-
sumably the reason for the original (ca. 1980) ratio of 18:12 for 20:5n-3 and
22:6n-3 in oil from the filter-feeding menhaden fish (Figure 5), repeated on hun-
dreds of labels for bottles of fish oil capsules, and repeated again in most concen-
trates prepared from it or from anchovy oil (Fig. 6). Actually the production of
18:2n-6 and 18:3n-3 is limited in a total lipid context for brown and red macrophyte
296 FISH OILS
algae, but it is more common in the green, Ulva pertusa being a familiar and much
studied example (60).
As already remarked, various invertebrates feed on the phytoplankton, and smal-
ler carnivorous fish feed on those vegetarian species as well as on carnivorous inter-
mediate invertebrates. Available 22:6n-3, with its neural and visual implications, is
probably conserved at these lower levels and is vital in fish muscle phospholipids
(3040% of fatty acids). In oils, 18:4n-3 and 18:4n-1 are preserved as shown in Fig-
ures 5 and 6. Recent progress in the lipid biochemistry of fish shows that the rain-
bow trout can perform biosynthesis of 22:6n-3 (DHA) from 18:3n-3 (alpha-
linolenic acid). Only a small part of that provided is converted to DHA (63), and
surprisingly this was substantially converted by the pyloric cecum as well as by the
liver (64), which raises an interesting point about adaption of fish biochemistry to
circumstances. The marine fish ingest preformed EPA and DHA, but there was a
curious change in diet for North American freshwater fish such as the rainbow trout.
The recent glaciations should have wiped out the resident invertebrates, and after
salmonids returned from the ocean, recolonization of food species for freshwater
fish would have been based on insect life introduced from Central America. Thus,
deprived of an excess of marine long-chain omega-3 fatty acids, adaption to elongate
the available insect C18 fatty acids would have been necessary once the returning
salmonids penetrated waters remote from the ocean. A similar situation for insect
lipids and fatty acids is known to come from Britain (65). These sources could pro-
vide EPA but not DHA for freshwater fish. Freshwater fish, although beneficial in
most respects among our sources of both n-3 and n-6 fatty acids (12,66,67), are not
apt to produce large volumes of fish oils of distinctive character. An exception is the
U.S. farmed catfish industry, which is subject to an excess of n-6 dietary fatty acids
from local aquaculture diets. The production of the visceral oil has been described,
in 2003 in JAOCS but with confusion in published fatty acids. The northern lakes of
Canada support fisheries with more potential for accumulating longer-chain omega-
3 fatty acids in the oils (67) or muscle. In four oils, the longer chain omega-3 fatty
acids (including 18:4n-3) totaled 9.6%, 13.3%, 17:0%, and 18:6% of total fatty
acids. The corresponding totals for the longer chain n-6 fatty acids arachidonic
acid (20:4n-6), 22:4n-6, and 22:5n-6 were 3.3%, 2.5%, 3.7%, and 4.2%, mostly
of 20:4n-6. Thus, a generally higher level of n-6 fatty acids over marine oils
is found, but overall a favorable balance exists between preformed C20 and C22
n-3 and n-6 longer chain PUFA contents in the oil. The fatty acids of the edible
muscle, widely eaten locally and also exported, are similarly skewed in favor of
omega-3 fatty acids, but with the additional n-6 fatty acids already mentioned
(R.G. Ackman, unpublished).
6. FISH OIL PRODUCTION AND QUALITY
A paramount concern in maintaining oil (and meal) quality is speedy processing

after catch. As 100 tons or more could be involved, chilling is not always practical,
but fish pumps can transfer the catch quickly from boats to the processing plant.
FISH OIL PRODUCTION AND QUALITY 297
Even there delays must be avoided. Preferably no more than 2430 hours should
elapse, depending on temperature before fish reduction. The fish enzymes, both
of muscle tissue and of digestive tissue, and those of gut bacteria, combine to break
down protein. The oil degradation is basically from lipolytic enzymes, but some of
the oil-soluble free fatty acids may come from partially digested food and phospho-
lipids and not necessarily from triacylglycerols. The free fatty acids, abbreviated to
FFA, are one of the oldest and simplest guides to fish oil quality. In one detailed
review (68), it is pointed out that the oil stored in fat cells (adipocytes) illustrated
for salmon by Zhou et al. (69) will be set free by 50 C, although cooking with
steam usually reaches 95 C. Separation of the oil is achieved with presses and
with very expensive and very efficient centrifuges. Once cooled, the oils are stable,
provided no protein particulates are carried over from the first separations of oil. A
second polishing centrifuge can handle this matter. The oil should be cooled
before storage in clean, dry tanks. These tanks should be filled as full as possible
and provided with provision for drainage of any sediments and water (foots).
Complex flow diagrams are provided by various authors (20,23,26,70). They are
complex because the thermodynamics, mainly for water removal, dictate costs.
References 20, 23, 26, and 70 provide additional detailed diagrams for those inter-
ested. Subsequent technology leading to consumer products is summarized in
Figure 9.
The basic crude fish oils are exemplified by the quality details of Table 3. Ranges
are given because these are specifications for crude oils, produced at the level of
over a million tons per year. The first six properties are traditional wet chemistry
assays and the American Oil Chemists Society (AOCS) Official and Tentative
Crude Fish Oil
Neutralization
Deodorizer
Bleaching
Ester
Fatty Acid Molecular
Concentrate Winterization Ditillation
Purified Fatty Acid
Purified Ester
Mono-Diglyceride Vacuum Distilled
Fish Oil
Production Finished Oil
Product
Packaging, Microencapsulation,
Soft Gelatin Capsules, Bottles
Figure 9. Production of pharmaceutical-grade fish oil or nutritional supplements. Reproduced

from (70) by permission of the American Oil Chemists Society.
298 FISH OILS
TABLE 3. Crude Fish Oil Quality Guidelines and Physical

Characteristics.
Quality Guidelines
Moisture and impurities, % usual basis 0.5 up to 1% maximum
Free fatty acids, % oleic range 17% but usually 25%
Peroxide value, meq/kg 320
Anisidine number 460
Totox value 1060
Iodine value
Capelin 95160
Herring 115160
Menhaden 120200
Sardine 160200
Anchovy 180200
Jack mackerel 160190
Sand Eel 150190
Color, Gardner scale up to 14
Iron, ppm 0.57.0
Copper, ppm less than 0.3
Phosphorus, ppm5-100
Physical characteristics
Specific heat, cal/g 0.50.55
Heat of fusion, cal/g about 54
Caloric value, cal/g about 9,500
Slip melting point, C 1015
Flash point, C
As triglycerides about 360
As fatty acid about 220
Boiling point, C greater than 250
Specific gravity
At 15 C about 0.92
At 30 C about 0.91
At 45 C about 0.90
Viscosity, cp
At 20 C 6090
At 50 C 2030
At 90 C about 10
From A. P. Bimbo (70).
Methods, Champaign, IL, or the Association of Official Analytical Chemists

(AOAC), Gaithersburg, MD, are the usual sources for North America to access
recognized, standardized, and detailed analytical methods. Traditionally, even dif-
ferent staff members in one laboratory can get somewhat different results. Techni-
cal spectral methods are now becoming useful, but the search for some exact and
rapid replacement fish oil technology goes on. Some problems are discussed in
Appendix 1, courtesy of Codex Alimentarius. A problem peculiar to the peroxide
value, the anisidine number, and hence the totox value is that it is a moving target.
Figure 10 shows that oxidation of highly unsaturated fatty acids proceeds, but at the
NUMERICAL VALUES
PEROXIDES ACIDS
ALDEHYDES
POLYMERS
TIME
Figure 10. Time relationship among peroxides and their degradation products after oxidation of
marine oils. No quantitative relationship is implied. Described in text as a moving target for
analytical methods.
same time degradation of the peroxides can also proceed, with degradation to alde-
hydes producing the familiar fishy flavor of both oils and fish muscle lipids
undergoing development of rancidity. Many candidate molecules are offered for
consideration (26), some being unstable themselves. The acids produced can be
volatile, and one ending to a peroxide free radicals career can be to lead to poly-
merization, either within a triacylglycerol or between triacylglycerols. The splitting
of an oxidized fatty acid chain can take place anywhere, but one-half of the pro-
duct(s) will still be attached to the glycerol molecule. Thus, removal of free volatile
aldehydes, for example, reduces the aroma from rancidity, but after their removal,
refining can leave the anisidine value for the remaining glycerol-bound aldehydes
as a real number of 5 or more. Addition of antioxidants depends on the value of the
raw material and of the final product, so it is not likely to be added to crude fish oils.
These already include the natural antioxidant benefit of natural alpha-tocopherol.
Sometimes cheaper vegetable oil deodorization mixtures of tocopherols may be
added after refining because the alpha-tocopherol may be lost in that step. The pro-
ducts of oxidation of oils protected by mixed tocopherols may then differ some-
what, but they have recently been studied in detail (71).
As already mentioned, the iodine value has been largely replaced by exact fatty
acid composition from gas-liquid chromatography of the methyl esters of the single
fatty acids. In addition to the AOCS Ce 1b-89 and AOAC 991.39 methods,
European Pharmacopeia 4 method 01/02/1352 includes their standards for ome-
ga-3 acidorum triglycerida and the GLC analysis for EPA and DHA in triacylgly-
cerols and an associated method for ethyl ester products (see below), as does the
Voluntary Monograph (October 2002) on Omega-3, DHA, EPA and DHA EPA
of the Council for Responsible Nutrition of Washington. This body also gives
300 FISH OILS
TABLE 4. Council for Responsible Nutrition Quality Standards for Nutraceutical Grade
Fish Oils in the United States.
CRN Quality Standards for Nutraceutical Grade Fish Oils
Measures of Oxidation
Peroxide Value (PV), meq/kg 5 Max
Anisidine Value (AV) 20 Max
TOTOX ((2 PV) AV) 26 Max
Purity
Dioxins (PCDDs, PCDFs) 2 pg/g WHO-TEQ Max
PCBs <0.09 mg/kg (ppm)
Lead <0.10 mg/kg (ppm)
Cadmium <0.10 mg/kg (ppm)
Mercury <0.10 mg/kg (ppm)
Arsenic <0.10 mg/kg (ppm)
Omega 3 Fatty Acids Expressed on a weight/weight basis (mg/g)
Acid Value 3 mg KOH/g Max
recommended maxima for heavy metals, dioxins and PCBs, and wet chemistry
value maxima (Table 4).
New regulations require new technology, which is especially true for food use.
Bleaching with activated clays has been a long-established practice to remove
chlorophyll green or a brownish tint acquired from heating oils in the presence
of other materials, and the objective of a clear yellow oil is usually possible.
Recently, the purification target has been extended to removal of organochlorine
materials, polyaromatic hydrocarbons, and most recently dioxins in particular. Acti-
vated carbon is recommended for dioxins, typically at 0.51.5% carbon for 2045
minutes at 80100 C. Some operators add this to the bleaching earth to reduce
handling steps. Various associated matters, such as disposal of the oil-coated
clay, develop that should be considered, and all are dealt with in a recent conference
report (72).
Occasionally, a highly purified fish oil is required for research purposes. Such
oils can be prepared by large-scale chromatography (73) for use in studying oxida-
tion products. The recent paper based on such technology is instructive in illustrat-
ing the variety of products produced from oxidation of even purified fish oil (71).
Modern technology is beginning to investigate such materials in situ (74).
For crude fish oil, deodorization has changed considerably, with new technology
designed to reduce temperature and/or exposure times. In 1990, a review chapter in
a book (75) described classic batch tray technology with gravity cascade transfer of
the oil and steam sparging to carry away volatiles. This process was satisfactory for
hydrogenated fish oils, but thermal damage to one of the highly unsaturated fatty
acids of vegetable oil had been recognized nearly 20 years earlier (76). A critical
temperature of about 185 C was observed, and at 230 C, severe isomerization of
ethylenic bonds (cis to trans) was observed in 18:3n-3 with prolonged heating.
These isomers were also found in retail oil samples and so were produced by
standard deodorizers of the type mentioned and illustrated in a chapter in a book on

fish oils (75). A thin-film design, the Campro unit is designed for a very short resi-
dence time, and with oil transfer as a thin film by gravity and by high-velocity
sparge steam, and it is described as suitable for fish oil refining at 210 C. In
fact, this is satisfactory for fish oils based on recent experience with this unit for
seal oil (private communication). Designs for throughputs of up to 300 tons/day
are available. The damage done to fish oil fatty acids at 220 C was investigated
in depth with open-tubular gas-liquid chromatography (77), and it illustrated the
problems of prolonged heat exposure (Figure 11). This case is a time-temperature
relationship, so operating flexibility is desirable.
A true molecular still ejects a low-molecular-weight molecule from a less vola-
tile liquid surface. If a very good vacuum exists ( 103 mm), this molecule may
bounce around with a few residual air molecules but eventually will either fall back
or attach to a cold condenser. Sometimes this is described as short-path distillation,
which is not usually an efficient separation process as shown by a comparison of
removal of DDT from cod liver oil and concurrent losses of Vitamin A (78). The
term molecular still has now become attached to units of a different design applied
to fish oils, for example, from the Pope Company of Menomonee Falls, WI, or the
Pfaudler company of Rochester, NY. The Pope model of molecular still illustrated
(75, 79) shows this wiped-wall concept. There is an outer cylindrical shell with heat
applied on the outside. An interior centered post is actually a cold condenser, and
wiper blades rotate continuously against the inner walls of the outer shell as the oil
is fed in at the top, spreading the oil as a film of <1 mm thickness. The constant
agitation of the film moves fresh volatile oil materials to the surface, and they pass
into the evacuated space. The desired vacuum is sufficient to cause the volatiles,
especially organochlorine materials (MW 358), but even cholesterol (MW
386), to pass over to the condenser at moderate (ca. 200 C) temperatures. Also
removed are squalene (MW 410) and most of the volatile fatty acid oxidation pro-
ducts such as 2,4-decadienal (MW 152) and other lower molecular weight mole-
cules contributing to objectionable flavors (73,74). Unfortunately, the natural
alpha-tocopherol present (MW 430) may suffer the same fate. The antioxidant
value of it and of squalene (80) may be lost. That these materials are effectively
removed from triacylglycerol oils with molecular weights in the 8001000 range
is a result of the repeated turnover and mixing of the oil during its descent of several
meters in very large units. The low viscosity at high temperatures helps refining by
such units, and very large surface areas permit flowthroughs of tons of fish oil per
hour.
As desirable products have similar molecular weights (ethyl docosahexaenoate
is 356, and ethyl eicosapentaenoate is 330), it is clear that this type of fish oil pro-
duct cleanup is best done at the oil stage to avoid losses at the same time as con-
taminant removal, which is carried out as described above. However, dimers and/or
other polymeric and/or colored materials may be left behind if wiped wall equip-
ment is used in its capacity of a simple and inefficient short-path still for the ethyl
esters only. This process will produce a water-white distillate product. Regrettably,
it is not an efficient way to separate ethyl-EPA from ethyl-DHA.
302 FISH OILS
Figure 11. The GLC C20 region of a menhaden omega-3 PUFA concentrate (ethyl ester): (a)
before and (b) after heat treatment at 220 C, and (c) the 20:5 region of an artifact concentrate
isolated by AgNO3 column chromatography. Peaks AE refer to artifacts formed after heat
treatment. Analysis on a SUPELCOWAX-10 fused-silica capillary column operated isothermally
at 195 C. Note that components BE fall into the region where several 22:1 isomers may be
found (cf. Figs. 5 and 7). From (77).
CONCENTRATES OF FISH OIL OMEGA-3 PRODUCTS 303
7. CONCENTRATES OF FISH OIL OMEGA-3 PRODUCTS
Table 5 shows the marketing and label strategies for some current marine omega-3
products sold in Halifax, Canada. Obviously samples No. 1 and No. 6 are natural
oils. Perhaps No. 5 is simply winterized oil, a process demonstrated in Table 6 for
menhaden oil, which follows from the tendency for DHA to be in the 2-position of
fish oil triacylglycerols (81). EPA is reputed to be somewhat less specific. In the
absence of 20:1 and 22:1, the outer 1- and 3-positions may, in some molecules, pre-
sent two saturated fatty acids from 14:0, 16:0, and 18:0 in one triacylglycerol mole-
cule, which leads to the stearine composition of Table 6.
Simple biochemical rules are made more complex by other factors, and in fish
oil triacylglycerols, the 20:1 and 22:1 fatty acids of a high melting point confuse the
issue (82). The traditional enzymatic approach to fatty acid distribution will soon
be replaced by nondestructive instrumental methods, particularly nuclear magnetic
resonance (NMR). It can distinguish the proportion of DHA between the 1,3- and
2-positions (74) and otherwise provide the details shown in Table 7 for a few oils,
which shows verification of the method through an international exchange (83). A
TABLE 5. Some Recent Omega-3 Product Retail Labels in Canada.
Label Content in mg Capsule size or liquid intake
No Description EPA DHA (1000 mg basis)
1 Wild Sockeye (Total 90) 1000

2 Wild Harvested Pacific Salmon 180 111 1000
3 O3mega 400 200 1065
4 Natural Sea* 775 500 1500
(517) (333) (1000)
5 Holista Premium Fish Oil** 180 120 1000
6 Omega Gold (liquid) 900 (180) 600 (120) 5000 (1000)
*Other (undefined) 225 mg.

**Each capsule is said to be equivalent in omega-3 content to 2.5 oz (70 g) portion of cooked salmon.
TABLE 6. Fatty Acid Composition (w/w%) of Menhaden Oil with

Olein and Stearine Fractions.
Menhaden Oil Olein Fraction Stearine Fraction
14:0 9 8 11
16:0 21 18 31
16:1 11 12 9
18:0 3 3 5
18:1 12 12 10
20:1 2 2 2
20:5 14 15 11
22:5 2 2 1
22:6 10 11 7
From A. P. Bimbo (26).

304 FISH OILS
TABLE 7. Comparison of DHA Content from the Interlaboratory 1H NMR Analysis

Between Japan and Norway Together with GC Data. Ethylene Glycol Dimethyl Ether was
Used as Internal Standard (1H NMR Analysis with 30s Pulse Repetition Time).
DHA Proportion Proportion (mol%)

DHA Content (mg/g) (mol%) n-3 Fatty Acids
Sample Oil Data Norway Japan GC Norway Japan GC Norway Japan GC
No. 1 Average 248.17 276.1 267.6 27.03 27.5 24.5 32.05 32.1 33.6
Bonito CV 0.64 1.33 1.67 0.78 0.46 2.49 0.08 0 2.03
No. 2 Average 123.79 122.6 123.4 11.93 12 11 22.56 22.2 21.6
Tuna CV 4.01 1.38 0.73 2.98 0.43 1.71 0.85 0.4 1.35
No. 3 Average 214.51 208.3 215.7 20.65 20.7 20.2 29.08 28.8 29.9
Tuna CV 2.72 0.91 0.2 0.66 2.02 1.43 0.05 0.73 0.89
No. 4 Average 217.97 206.7 215.4 21.18 20.8 19.7 32.22 31.8 32.1
Tuna CV 3.84 0.07 0.72 1.59 0.47 1.35 0.23 0.34 1.07
No. 5 Average 111.07 111.1 106 10.65 10.8 10.2 28.03 28.1 28.6
Salmon CV 3.27 0.95 1.47 0.88 0.32 1.15 0.32 0.33 0.74
From S. Wada (74).
recent enzymatic examination of tuna oil (EPA 6.69%, DHA 26.4%) showed 8.7%
EPA and 56.3% DHA in the 2-position (84).
Retail product No. 1 of Table 5 is likely to be simply salmon waste oil, the fish
name conferring an elite status. Our research (Ackman, unpublished) suggests that
many salmon oil encapsulated oils are unrelated to any salmon oil in fatty acid
composition. In 1989, our analysis showed many products of this type to be exag-
gerated as to omega-3 fatty acid content (85), and a more recent European survey in
1998 gave comparable results and reported on quality (86).
In the menhaden oil winterization of Table 6, the increase in polyunsaturated
fatty acids is modest in the olein fraction, and the commercial objective may
have been the stearine fraction, 50% richer in saturated acids of commercial inter-
est. As already remarked for polyunsaturated fatty acids, 18/12 (in implied percents) or
180/120 (in mg/g) were obtained as triacylglycerols from menhaden and/or anchovy oils
with minimal trouble and technology. Retail product No. 2 is also a product like No. 5.
Winterization also would prevent either capsules or oil turning cloudy if refrigerated.
Strangely, hardly anybody challenges label claims as to chemical nature, although the
word oil may be carelessly used because concentrates are almost always ethyl esters, but
numbers should always be expressed in the free acid form.
Among the various laboratory procedures used for studying fatty acids, concen-
trations by chromatography on silver nitrate impregnated silica gel, or equivalent
(87), are too expensive to scale up; although effective for fish oils (88), they and
mercuric adducts (89) would not be acceptable for health and safety reasons.
Before 1980, there was a now forgotten industrial technology for concentrating
fish and vegetable oils called the Solexol process. It can be described as the first
large-scale use of supercritical gases, in this case, propane. A report with much con-
venient detail based on iodine values was published in 1949, but as a historical
record (90), because several large plants were closed during the war and the prime
markets such as paints, linoleum, and oil cloth disappeared after the war. Two dec-
ades later supercritical carbon dioxide was to be the panacea in this field, but early
promises, including fish oil fractionation, were seldom realized (91, 92).
Much research was conducted to prepare concentrates when the U.S. National
Marine Fisheries Service provided funds for exploring most existing technologies
for concentrating fish oil fatty acids. This exploration would provide omega-3 con-
centrates for the medical research sponsored by the National Institutes of Health
and included supercritical fractionation of esters. Unlike commercial research,
results were available to all interested parties. A book chapter by authors located
in the Charleston, Seattle, and Gloucester NMFS laboratories, and in the Hormel
Institute, Austin, Minnesota (93), provides detail on all methods considered for frac-
tionating methyl or ethyl esters, because menhaden oil was clearly intractable. For
supercritical CO2, the differences in factors such as the influence of chain length
(i.e., molecular weight) and unsaturation, including explanations of results achieved
by others, are helpful in explaining why this fractionation process has not been
commercially developed. The final total enrichment process is described below.
Urea complexing was demonstrated for fractionation of fatty acids of a marine oil
(as methyl esters) as early as 1963 (94), and laboratory-scale tests in Halifax, Canada
(95) were followed by further tests on a scale of 40 kg of crude oil. At that time, 50%
omega-3 ethyl esters was considered a good possible result and doubling the omega-3
content made the product more acceptable on a retail basis by 1989 (79, 85).
A flow chart, courtesy of H. Breivik of Norway, is provided as Figure 12, the
work of Norsk Hydro, Porsgrunn, Norway, and dating to 1990. The molecular
weight of ethyl myristate (14:0, 256), palmitate (16:0, 284), palmitoleate (16:1,
282), and even oleate (18:1, 310) are sufficiently different from those of the ethyl
esters of C20 and C22 polyunsaturated acids (330 and 356) to allow removal of
most of the shorter chain fatty acids by short-path distillation. From Figures 5
and 6, it can be seen that 3050% of the ethyl esters in question are available to
be distilled off, with 18:4n-3 as the omega-3 fatty acid of interest that may be lost.
By eliminating these fatty acids as a first step with the simple operation of short-path
distillation, the subsequent urea complexing step is much more economically carried
out, which led to the very successful Provnova Biocare product of 85% EPA DHA
as ethyl esters. Under the trade name EPAX, a variety of triacylglycerol and ethyl
ester products are now offered by this firm, with different proportions of EPA and
DHA, which is evidence of further process development since 1990.
At about the same time, the Charleston Laboratory of the U.S. National Marine
Fisheries Service prepared menhaden oil omega-3 fatty acid ethyl ester concen-
trates on a large scale for participants in projects funded by the U.S. National Insti-
tutes of Health. Their flow chart. Figure 13, combined development work also
carried out at the Seattle and Gloucester Laboratories, much of it recorded in the
book by M.E. Stansby (26). The final process adopted includes urea complexing of
ethyl esters in a total of seven stages, viz:
I. Vacuum deodorization
II. Transesterification
306 FISH OILS
MASS BALANCE
PART OF EARLY SCALE-UP EXPERIMENT
PRIOR TO LATER MODIFICATIONS
Fish Oil
30.800 kg
Ethyl ester
29800 kg
Precut: 16510 kg
2 step molecular
distillation
Residue: 3245 kg
Ethyl ester (EPA + DHA = 50%)

9575 kg
Urea Urea precipitation Isolated adduct

16150 kg 18890 kg
Ethyl ester (EPA + DHA = 84%)

3140 kg
Further purification
Figure 12. Development stage of Norsk Hydro Research Center Porsgrun scheme for
production of ethyl ester concentrates of marine oil fatty acids, ca. 1990. Courtesy of H. Breivik.
III. Urea adduction

IV. Film evaporation
V. Short-path distillation
VI. Supercritical fluid fractionation
VII. High-performance liquid chromatography
Although subsequently closed down only because of the cessation of the joint pro-
ject, the demonstration of what could be done, even if uneconomic, created wide
REFINED MENHADEN OIL
OMEGA - 3 CONCENTRATES
N2 N2
N2 U R
D
EA
C g
II
III
P1 P1
P1 N2
N2
D
V
N2 C
O IV
P2
P2 P2
PURE EPA AND DHA
VI VII
Figure 13. Production of biomedical test materials in the Charleston Laboratory of the U.S.
National Marine Fisheries Service for a joint NMFS-NIH project. Code numbers for steps are
explained in text. Small print identities are N2 nitrogen atmosphere, A alkali, E ethanol,
D distill, U urea, R reflux, C cold water, and S steam. From (96).
interest in employing concentrates in clinical trials of omega-3 fatty acids. The pro-
duct standards that were set were, for the time, remarkably high (Table 8).
Investigations carried out for the NMFS Charleston process indicated that chain
length (i.e., molecular size) was a dominant factor observed in research on super-
critical CO2 separations of ethyl esters of marine oil fatty acids (93). The initial
308 FISH OILS
TABLE 8. Quality Specifications for Fish Oil Derived n-3 Ethyl Esters
to be Shipped from Charleston Laboratory.
Test Material
Analysis Type n-3 Conc EPA DHA
Esters, % >90 >95 >95

EPA, mg/g >400 >900 <50
DHA, mg/g >200 <50 >900
Total n-3, mg/g >700 >950 >950
Free Fatty Acids, % <0.2 <0.2 <0.2
Trans Acids, % <5 <5 <5
Cholesterol, mg/g <5.0 <0.1 <0.1
Peroxide Value, meq/kg <10.0 <5.0 <5.0
Iodine Value, g I/100g >320 * *
Anisidine Value <80 * *
Antioxidant Content
a-tocopherol, mg/g 0.55.0 ** **
g-tocopherol, mg/g 0.55.0 ** **
TBHQ, mg/g 0.10.2 ** **
Moisture, ug/g <500 <500 <500
Residual urea, ug/g <20 <20 <20
PCB, ug/g <0.5 <0.5 <0.5
Total DDT, ug/g <0.5 <0.5 <0.5
Trace Metals, ug/g
Arsenic <1.0 <1.0 <1.0
Cadmium <1.0 <1.0 <1.0
Lead <1.0 <1.0 <1.0
Mercury <1.0 <1.0 <1.0
Selenium <1.0 <1.0 <1.0
Sensory Attributes:
Odor (TIO) <6.0 * *
Flavor (TIF) <6.0 * *
Other:
Specific Gravity 0.89 ** **
Solidification Range *** *** ***
*not applicable.
**not enough material to conduct these analyses routinely.
***Esters are a liquid at 5 C or higher.
Reproduced from (96).
promise of supercritical fluid extraction (SFE) for actual recovery of lipids from
natural samples is a separate issue from oil fractionation, but a few references estab-
lish the difficulties faced with use of cosolvents, water removal, and other factors
not applicable to oil or ester fractionation (9799). Thanks to many pioneers and
recent commercial stimulus, supercritical fluid fractionation has the potential for
concentrating ethyl esters of fish oil fatty acids.
Gradually, health benefits have attracted financial interest and a market role for
omega-3 concentrates was seen in U.S. foods. A leader in this field was Roche Vita-
mins, Inc., of Parsippany, NJ, soon to be acquired by the DSM company of the
THE OTHER OILS 309
TABLE 9. Product Data for ROPUFA 75 n-3 EE.
Product identification for refined ethyl esters of fish oil, containing eicosapentaenoic acid ethyl
ester and docosahexaenoic acid ethyl ester
Specifications EPA content (gas chromatography): min 42%

Appearance: yellowish liquid EPA content (weight as ethyl ester): min 380
Acid value: max. 3 KOH/g mg/g
Peroxide value: max. 10.0 mEq/kg DHA content (gas chromatography): min 22%
Anisidine value: max. 20 DHA content (weight as ethyl ester): min 200
Oligomers: max. 2% mg/g
Conjugated dienes: max. 1.5% Total content of n-3 PUFAs (gas
Iron: max. 1 ppm chromatography): min 75%
Copper: max. 0.1 ppm Total content of n-3 PUFAs (weight as ethyl ester):
min. 720 mg/g
Arsenic: max. 0.1 ppm
Courtesy of Roche Vitamins Inc., Parsippany, NJ.
Netherlands. The basic product was ROPUFA 75 n-3 EE with specifications as

set forth in Table 9. Information on this ethyl ester product clearly shows a GLC
profile with traces of C16 fatty acids, traces of 18:0 and 18:1, a little of each of
20:4n-6 and of 20:1 and 22:1, and the ubiquitous 21:5n-3 and 22:5n-3.
Many students have received degrees in recent years for exploring enrichment of
marine oils (or other oils) by selective hydrolysis of triacylglycerols, or ester inter-
changes between esters and natural oils by enzymes. These explorations tend to be
somewhat theoretical (100), but they can be effective, although impractical, for
example, a 100-hour reaction time (101). Having a starting material rich in the
desired product fatty acid (DHA) helped in one case (102), but the complexity of
these proposed processes requires a separate article.
There is no real reason for concentrations of mixed omega-3 fatty acid ethyl esters
to be provided when the estimated need in oral supplement form is approximately 1 g
per day for adults. This supplement can be provided by several capsules of any sui-
table oil taken with meals. However, the food industry requirement is for oils that are
microencapsulated powders. These powders have to be provided with stable, yet
digestible, shells and in a suitable powder format to be unnoticed in the food and
yet carry a significant proportion of mass as the marine oil or omega-3 concentrates.
A capacity of about 5050% oil is acceptable, and a target of 7080% content is
being sought. Eventually, concentrates will boost delivery of EPA and DHA.
8. THE OTHER OILS
Aside from cod liver oil, no mention has been made of other fish liver oils, although
at one time, there was production of vitamins A and D from the liver oil of a Pacific
dogfish Squalus acanthius. This industry collapsed when synthetic vitamins were
introduced. The Atlantic spiny dogfish is essentially the same fish, and the liver
oils from both coasts have been compared (103). The distinguishing feature of
310 FISH OILS
TABLE 10. Liver Oil and Squalene Analysis of Dogfish from the Continental Slope
of New Zealand.
Oil Yield Squalene Yield

Total No. (g/100g Liver) (g/100g Oil)
No. of of Dogfish
Species Sex Samples Sampled Mean Range Mean Range
Shovelnose Dogfish Male 4 40 87.7 8689 65.7 6469

Shovelnose Dogfish Female 11 110 86.4 8390 52.7 4559
Shovelnose Dogfish Juvenile 3 30 89.2 8690 58.4 5760
Baxters Dogfish All 2 20 77.7 48.9 4454
Seal Shark All 2 15 84.5 8188 72.6 7273
Leafscale Gulper Shark All 3 22 81.5 7486 62.9 6164
Plunkets Shark All 1 2 84.2 1.4
Owstons Dogfish Male 2 20 80.6 67.7 6670
Owstons Dogfish Female 2 16 44.2 6670
Portugese Dogfish All 1 5 81.4 44 4147
From G. Summers et al. (105).
this shark liver oil is the content of diacyl glyceryl ethers (DAGE, Figure 1). The
limited survey showed that the Pacific oil was the richer in DAGE (41% vs. 18% in
Atlantic liver oil), but the C20 and C22 omega-3 fatty acids quantitatively were
similar (103). In a further examination of the liver oil from Atlantic dogfish, the
ability of Atlantic salmon to digest was excellent (104). The oil, therefore, has
value for aquaculture feeds.
Squalene (Figure 4) is also sold in capsules in health supplement stores. Shark (and
related elasmobranch) livers do not necessarily have this hydrocarbon in any more
traces than those found in other species of fish. It depends on the exact species. A
survey of New Zealand shark resources (Table 10) shows what may be expected
(105). This report also explores liver oil recovery and processing. Unfortunately, local
fishing in underdeveloped countries, often for shark fins only, has destroyed a large
part of the resource. Squalene can exaggerate the iodine value of such oils (106),
but it is easily measured by GLC after hydrogenation of methyl esters carefully pre-
pared from the whole liver oil (107). Curiously, a small anadromous fish, ooligan or
eulachon, spawning in the Fraser and other Pacific coast rivers has considerable
(19%) squalene in its body fats (108, 109). The eulachon Thaleichthys pacificus
was long recognized by local aborigines as a source of inedible but useful oil, but
although the fish is fatty, it is edible and is fished to some extent for that reason. As
olive oil is a potential source of squalene if it is needed, the slaughter of sharks for this
hydrocarbon for any purported health benefits is to be deplored.
Wax esters are another useful marine lipid class, which are now historical when
derived from the heads of sperm whales. Although various marine invertebrates
contain wax esters (110), there is an unexploited resource in relatively small fish
called myctophids. These fish can be caught by modern fishery technology as
was shown in South Africa some decades ago, but the use of any oil and meal pro-
duced would have to be carefully considered. The biosynthesis of their wax esters
has recently been resolved (111) and reviews most questions on that topic that were
CODEX ALIMENTARIUS EXPLANATORY MATERIALS RELATING TO FISH 311
left unanswered. Analysis of the whole lipids (18%) of a North Atlantic fish, the
barracudina Paralepis notolepis rissoi kryeri, showed this to contain 85% wax
esters (112). The fatty acids of the wax esters differed from those of sperm body
or head wax esters, but the alcohols from barracudina showed a remarkable simi-
larity to those of sperm head wax.
Algae in theory and in practice can produce oils that are triacylglycerols
(39, 113). The DHA-rich oils commercially produced by the Martek Biosciences
Corp. are approved for use in infant formulas. They have also been shown to pos-
sess the favorable clinical attributes of fish oil DHA in healthy adults, whether
alone (114) or combined with arachidonic acid (20:4n-6) of fungal origin (115).
Products of these two sources of refined fatty acids show no particular resistance
to oxidation, compared with fish oils, when finally purified of natural components
that might not be allowed in products for human consumption (116). A competing
source of DHA (Nutrinova Inc., Somerset, NJ) advertises its products as from
vegetarian source. As noted earlier, algae can also be good sources of EPA
(117), which reopens the question posed earlier. Are the EPA and DHA of fish
oils all or mostly from the fatty acids originally supplied by phytoplankton?
9. CONCLUSION
In a decade, the advanced lipid analytical technology that defines DHA as a nat-
ural fatty acid of marine oils from a deep-sea shark described around 1994 (118)
has been surpassed by nondestructive NMR measurement of DHA in situ (Table 7).
Thus, advanced analytical technology is supporting with new developments the
benefits from marine oil omega-3 fatty acids in our daily lives.
APPENDIX 1. CODEX ALIMENTARIUS EXPLANATORY

MATERIALS RELATING TO FISH OIL QUALITY TERMINOLOGY
Quality Guidelines and Potential Problem Areas or Disadvantages of Various Parameters.
Codex Specif. CAC/RS

Quality Unit Disadvantage or Potential Problem Area 19-1981 Rev. 1 1989.
Color Dark-colored oils may be crude and contain No Standard

contaminants normally removed by refining
or the color might indicate overheating during
refining.
Iodine Value IV varies with the species of fish. In general, IV No Standard
is a measure of the unsaturation in oils. High IV
oils are generally more susceptable to oxidation.
Acid Value1 High acid value crude oils might indicate that 0.6 mg KOH/g fat max
poor-quality fish were processed or the oil refined oils
deteriorated in storage.
4 mg KOH/g fat max
virgin oils
4 mg KOH/g fat max cold
pressed oils
312 FISH OILS

Peroxide Value Peroxide value is the primary measure of 10 meq/kg fat max virgin
rancidity (Oxidation) in an oil or fat. It reflects and cold pressed oils
recent oxidation. 5 meq/kg fat max other
oils
p Anisidine The Anisidine Number also measures products No Standard

Number of oxidation; however, it reflects oxidation that
has taken place in the past.
Totox Value A relationship between peroxide value and No Standard
anisidine number that is used to measure the
rancidity level of fats and oils. It is defined as
(2 PV) AN. It reflects total oxidation to date.
Moisture Considered an impurity. High levels of moisture 0.20% max
in an oil can lead to deterioration in storage.
Soap Soap can be formed when moisture is present 0.005% max
in the crude oil and reacts with the free fatty
acids and a catalyst (alkali ion), or it can result
from incomplete removal of soap from washed
refined oil.
Insoluble These are substances including traces of 0.05% max
Impurities protein, dirt, rust, and other materials that tend
to precipitate out of the oil during storage.
Depending on the substance, they can reduce
the stability of the oil.
Unsaponif. Matter These are composed of sterols, hydrocarbons, No Standard
glyceryl ethers, and fatty alcohols. There may
also be traces of pigments, vitamins, and
oxidized oil. Unsaponifiables vary with the
species of fish.
Organochlorine, Numerous compounds come under this group. No Standard
Organophos- Generally, the pesticide content of the oil
phorous reflects the environmental conditions in the
Pesticides and area where the fish are caught. The level of
other Chlorinated these compounds in the oil must be within the
Hydrocarbons regulatory limits of the locality involved.
Total Cholesterol Cholesterol is a major part of the unsaponifiable No Standard
fraction of fish oils. Generally it is not removed
except by vacuum stripping of the oil.
Iron Iron is considered a pro-oxidant in fish oil and 1.5 mg/kg max refined oil
is removed by degumming and refining. 5 mg/kg max virgin oil
5 mg/kg max cold
pressed oil
Copper Copper is considered a pro-oxidant in fish oil 0.1 mg/kg max refined oil
and is removed by degumming and refining. 0.4 mg/kg max virgin oil
0.4 mg/kg max cold
pressed oil
REFERENCES 313

Arsenic A heavy metal, naturally ocurring in sea 0.1 mg/kg mx

water. It is removed by the refining process.
Lead A heavy metal removed by the refining process. 0.1 mg/kg max
Mercury A heavy metal removed by the refining process. No Standard
Selenium A heavy metal removed by the refining process. No Standard
Cadmium A heavy metal removed by the refining process. No Standard
Hygiene Microbiological contamination by enterobacteria, CAC/RCP 1-1969, Rev. 2
salmonella, coliforms, or E. coli would be an 1985 limits.
indication of the sanitary conditions under
which the oil was manufactured.
Oil Soluble Normally part of the unsaponifiable fraction of No Standard
Vitamins the oil. High Vitamin A and/or D would indicate
that the oil is a liver oil rather than a body oil.
1
Acid value is defined as two times the free fatty acid content of the oil. CAC defines edible fats and oils as
foodstuffs composed of glycerides of fatty acids. They are of vegetable, animal, or marine origin. They may
contain small amounts of other lipids such as phosphatides, unsaponifiables, or free fatty acids naturally
present in the fat or oil. CAC defines virgin fats and oils as edible fats and oils obtained without altering the oil,
by mechanical procedures, and the application of heat only. They may be purified by washing with water,
settling, filtering, and centrifuging only. CAC defines cold-pressed fats and oils as edible vegetable fats and oils
obtained, without altering the oil, by mechanical procedures without the application of heat. They may have
been purified by washing with water, settling, filtering, and centrifuging only.
REFERENCES
1. S. M. Barlow and M. E. Stansby, eds., Nutritional Evaluation of Long-Chain Fatty Acids

in Fish Oil, Academic Press, London, 1982.
2. R. G. Ackman, Fatty Acids in Newer Fats and Oils, in Y. H. Hui, ed., Baileys Industrial
Oil and Fat Products, 5th ed., Vol. 1, Edible Oil and Fat Products: General Applications,
3. J-L. Sebedio, M. F. Langman, C. A. Eaton, and R. G. Ackman, J. Amer. Oil Chem. Soc.,
58, 41 (1981).
4. J-L. Sebedio and W. W. Christie, eds., Trans Fatty Acids in Human Nutrition, The Oily
Press, Dundee, (1998).
5. T. Leth, A. Bysted, K. Hansen, and L. Ovesen, J. Amer. Oil Chem. Soc., 80, 475 (2003).
6. R. G. Ackman, Natural Trans Fatty Acids, in Ref. 4, pp. 303311.
7. J. Dyerberg and H. O. Bang, Lancet, (2), 433 (1979).
8. D. Kromhaut, E. B. Bosschieter, and C. De Lezenne Coulander, New Eng. J. Med., 312,
1205 (1985).
9. Committee on Medical Aspects of Food Policy, Department of Health, Nutritional
Aspects of Cardiovascular Disease. Report of the Cardiovascular Review Group, No.
46, HMSO, London, (1994).
314 FISH OILS
10. P. M. Kris-Etherton, W. S. Harris, and L. J. Appel, for the Nutrition Committee, AHA
Scientific Statement, Circulation, 106, 2747 (2002).
11. R. G. Ackman, Food Revs. Intern., 6, 617 (1990).
12. R. G. Ackman, Fatty Acids in Fish and Shellfish, in C. K. Chow, ed., Fatty Acids in Foods
and their Health Implications, 2nd ed., Marcel Dekker, New York, 2000, pp. 169184.
13. W. E. Connor and A. Bendich, eds., Amer. J. Clin. Nutr., 71(suppl) (1), (2000).
14. A. J. Sinclair, N. M. Attar-Bashi, and D. Li, Lipids, 37 1113, (2002).
15. S. C. Cunnane, Can. J. Physiol. Pharmacol., 74, 629 (1996).
16. F. Thies, E. A. Miles, G. Nebe-von-Caron, J. R. Powell, T. L. Hurst, E. A. Newsholme,
and P. C. Calder, Lipids, 36, 1183 (2001).
17. R. Marchioli, F. Barzi, E. Bomba, et al., Circulation, 105, 1897 (2002).
18. A. Leaf, Circulation, 105, 18741875 (2002).
19. S. A. Reed, J. Sci. Food Agric., 15, 399 (1964).
20. F. V. K. Young, The Production and Use of Fish Oils, in Ref. 1, pp. 123.
21. A. P. Bimbo and J. B. Crowther, J. Amer. Oil Chem. Soc., 69, 221 (1992).
22. J. R. Burt, R. Hardy, and K. J. Whittle, eds., Pelagic Fish. The Resource and its
Exploitation. Fishing News Books, Blackwell Sci. Pubs., Oxford, United Kingdom,
1992.
23. A. P. Bimbo, Fish Meal and Oil, in R. E. Martin, E. P. Carter, G. J. Flick, Jr., and L. M.
Davis, eds., Marine and Freshwater Products Handbook, Technomic Pub. Co. Inc.,
Lancaster, Pennsylvania, 2000, pp. 541581.
24. R. C. Ackman, ed., Marine Biogenic Lipids, Vols. 1 and 2, CRC Press, Boca Raton,
Florida, 1989.
25. J. K. McElvoy, A techno-economic model for designing large factory vessels and its
application to the Antarctic krill-finfish fishery, in Ref. 22, pp. 94114.
26. M. E. Stansby, ed., Fish Oils in Nutrition, An avi Book, Van Nostrand Reinhold,
New York, 1990.
27. R. G. Ackman, W. M. N. Ratnayake, and B. Olsson, J. Amer. Oil. Chem. Soc., 65, 136
(1988).
28. R. G. Ackman, J-L. Sebedio, and M. I. P. Kovacs, Mar. Chem., 9, 157 (1980).
29. W. N. Ratnayake and R. G. Ackman, Lipids, 14, 795 (1979).
30. W. N. Ratnayake and R. G. Ackman, Lipids, 14, 804 (1979).
31. R. F. Lee and J. S. Patton, Alcohol and Waxes, in Ref. 24, Vol. 1, pp. 73102.
32. D. Menoyo, C. J. Lopez-Bote, J. M. Bautista, et al., Aquat. Liv. Resour, 15, 217 (2002).
33. R. N. Lemaitre, I. B. King, and T. E. Raghunathan, et al., Circulation, 105, 697 (2002).
34. A. M. Katz, Circulation, 105, 669 (2002).
35. R. G. Ackman, Fatty Acids, in Ref. 24, Vol. 1, pp. 103137.
36. Y. Ando, T. Ota, and T. Takagi, Yukagaku, Proc. Vol. 2, Sess. Lect. Sci. Pres. ISF-JOCS
World Congress, 1989, pp. 12631268.
37. R. G. Ackman, Anal. Chim. Acta, 465, 175 (2002).
38. R. G. Ackman, INFORM, 5, 1119 (1994).
39. Y. Itabashi, N. Harada, T. Ota, and K. Matsunaga., Yukagaku, 45, 271 (1996).
40. P. Mayzaud, C. A. Eaton, and R. G. Ackman, Lipids, 11, 858 (1976).
REFERENCES 315
41. P. Mayzaud and R. G. Ackman, Lipids, 13, 24 (1978).

42. J. D. Joseph and R. G. Ackman, J. Assoc. Off. Anal. Chem. Int., 75, 488 (1992).
43. R. G. Ackman and S. N. Hooper, Comp. Biochem. Physiol., 46B, 153 (1973).
44. H. E. Worne and L. W. Smith, Amer. J. Med. Sci., 710721 (1959).
45. C. M. Harlow, L. Steeves, and A. Myrden, Nova Scotia, Med. Bull., 37, 127 (1958).
46. E. J. Schaefer, Amer. J. Clin. Nutr., 75, 191212 (2002).
47. P. Cullen, Amer. J. Cardiol, 86, 943 (2000).
48. K. H. Bnaa, K. S. Bjerve, B. Straume, I. T. Gram, and D. Thelle, New Eng. J. Med., 322,
795 (1990).
49. T. Kanayasu-Toyoda, I. Morita, and S. Murota, Prostaglandins Leukotri. Essent. Fatty
Acids, 54, 319 (1996).
50. C. Benistant, F. Achard, S. Ben Slama, and M. Lagarde, Prostaglandins Leukotri. Essent.
Fatty Acids, 55, 287 (1996).
51. W. S. Harris, J. Lipid Res., 30, 785 (1989).
52. F. Achard, C. Benistant, and M. Lagarde, Biochim. Biophys. Acta., 1255, 260 (1995).
53. J. D. Carver, Amer. J. Clin. Nutr., 77(suppl.), 1550S (2003).
54. S. M. Innis and S. L. Elias, Amer. J. Clin. Nutr., 77, 473 (2003).
55. A. Leaf, J. X. Kang, Y-F. Xiao, and G. E. Billman, Circulation, 107, 2646 (2003).
56. J. S. Charnock, Prog. Lip. Res., 33, 355 (1994).
57. T. Helm, Dangerous Sea Creatures, Funk and Wagnall, New York, 1976.
58. W. M. Indrasena and T. A. Gill. J. Food Sci., 65, 948 (2000).
59. P. Pohl and H. Wagner, Z. Naturforsch., 27, 53 (1972).
60. K. Hayashi, S. Kida, K. Kato, and M. Yamada, Bull. Jap. Soc. Sci. Fish., 40, 609 (1974).
61. R. G. Ackman, C. S. Tocher, and J. McLachlan, J. Fish. Res. Bd. Can, 25, 1603 (1968).
62. S. M. Budge, C. C. Parrish, and C. H. McKenzie, Mar. Chem., 76, 285 (2001).
63. M. V. Bell, J. R. Dick, and A. E. A. Porter, Lipids, 36, 1153 (2001).
64. M. V. Bell, J. R. Dick, and A. E. A. Porter, Lipids, 38, 39 (2003).
65. C. Ghioni, J. G. Ben, and J. R. Sargent, Comp. Biochem. Physiol., 1143, 161 (1996).
66. T. Ota and M. Yamada, Bull. Fac. Fish. Hokkaido University, 26, 277 (1975).
67. R. G. Ackman, C. A. Eaton, E. G. Bligh, and A. W. Lantz., J. Fish Res. Bd. Can., 24,
1219 (1967).
68. W. Schmidtsdorff, Fish meal and fish oilnot only by-products, in A. Ruiter, ed., Fish
and Fishery Products, CAB International, Wallingford, 1995, pp. 346376.
69. S. Zhou, R. G. Ackman, and C. Morrison, Can. J. Fish. Aquat. Sci., 53, 326 (1996).
70. A. P. Bimbo. INFORM, 9, 473 (1998).
71. E. Kulas, E. Olsen, and R. G. Ackman, Eur. J. Lipid. Sci. Technol., 104, 520 (2002).
72. K. Berger, INFORM, 14, 307 (2003).
73. E. Kulas, R. G. Ackman, J. Amer. Oil Chem. Soc., 78, 361 (2001).
74. S. Wada, J. Oleo Sci., 50, 329 (2001).
75. A. P. Bimbo, Production of Fish Oil, in Ref. 26, pp. 181225.
76. R. G. Ackman, S. N. Hooper, and D. L. Hooper, J. Amer. Oil. Chem. Soc., 51, 42
(1974).
316 FISH OILS
77. R. C. Wijesundera, W. M. N. Ratnayake, and R. G. Ackman, J. Amer. Oil Chem. Soc., 66,
1822 (1989).
78. K. Julshamn, L. Karlsen, and O. R. Braekkan, Fiskeridirektoratets Skrifter, Ser. Tek.
Underskelser, Vol. 5, No. 15, Directorate of Fisheries Bergen, 1973.
79. R. G. Ackman, Chemistry and Industry, 7, 139 (1988).
80. M. A. Dessi, M. Deiana, B. W. Day, A. Rosa, S. Banni, and F. P. Corongiu., Eur. J. Lipid.
Sci. Technol., 104, 506 (2002).
81. G. Hlmer, Triglycerides, in Ref. 24, Vol. 1, pp. 139173.
82. R. Ackman, V. R. Orozco, and W. M. N. Ratnayake, Fat. Sci. Technol., 93, 447 (1991).
83. T. Igarishi, M. Ausand, Y. Hirata, I. S. Gribbestad, S. Wada, and M. Nonaka, J. Amer. Oil
Chem. Soc., 77, 737 (2000).
84. M. Cansell, F. Nacka, and N. Combe, Lipids, 38, 554 (2002).
85. R. G. Ackman, W. M. N. Ratnayake, and E. J. Macpherson, J. Amer. Oil. Chem. Soc., 66,
1162 (1989).
86. V. K. S. Shukla and E. G. Perkins, INFORM, 9, 955 (1998).
87. W. W. Christie, Silver-ion high-performance liquid chromatography, in J. L. Sebedio and
E. G. Perkins, eds., New Trends in Lipid and Lipoprotein Analyses, AOCS Press,
Campaign, Illinois, 1995, pp. 5980.
88. J. L. Guil-Guerrero and E-H. Belarbi, J. Amer. Oil Chem. Soc., 78, 477 (2001).
89. J-L. Sebedio and R. G. Ackman, Lipids, 16, 461 (1981).
90. H. J. Passino, Ind. Eng. Chem., 41, 280 (1949).
91. V. J. Krukonis, Processing with supercritical fluids, overview and applications, in B. A.
Charpentier and M. R. Sevenants, eds., Supercritical Fluid Extraction and Chromato-
graphy, Techniques and Applications, Amer. Chem. Soc., Washington, D. C., 1988, pp.
2643.
92. S. S. H. Rizvi, R. R. Chao, and Y. J. Liaw, Concentration of omega-3 fatty acids from fish
oil using supercritical carbon dioxide, in B. A. Charpentier and M. R. Sevenants, eds.,
Supercritical Flyid Extractioned, 1988, pp. 89108.
93. V. F. Stout, W. B. Nilsson, J. Krzynowek, and H. Schlenk, Fractionation of fish oils and
their fatty acids, in Ref. 26, pp. 73119.
94. R. G. Ackman, R. D. Burgher, and P. M. Jangaard, Can. J. Biochem. Physiol., 41, 1627
(1963).
95. W. M. N. Ratnayake, B. Olsson, D. Matthews, and R. G. Ackman, Fat. Sci. Technol., 90,
381 (1988).
96. Jeanne D. Joseph, Ed. 1989. Biomedical Test Materials Program: Production Methods
and Safety Manual. NOAA Technical Memorandum NMFS-SEFC-234, 120 p.
97. J. Via, L. T. Taylor, and F. K. Schweighardt, Anal. Chem., 66, 1459 (1994).
98. S. L. Taylor, F. J. Eller, and J. W. King, Food Res. Intern., 30, 365 (1997).
99. T. H. Walker, H. D. Cochran, and G. J. Hulbert, J. Amer. Oil. Chem. Soc., 76, 595 (1999).
100. M. Linder, J. Fanni, and M. Parmentier, OCL, 8, 73 (2001).
101. G. P. McNeill, R. G. Ackman, and S. R. Moore, J. Amer. Oil Chem. Soc., 73, 1403 (1996).
102. Y. Shimada, K. Maruyama, A. Sugihara, S. Moriyama, and Y. Tominaga, J. Amer. Oil.
Chem. Soc., 74, 1441 (1997).
REFERENCES 317
103. S-J. Kang, M. C. A. Timmins, and R. G. Ackman, J. Amer. Oil. Chem. Soc., 75, 1667
(1998).
104. S-J. Kang, S. P. Lall, and R. G. Ackman, Lipids, 32, 19 (1997).
105. G. Summers, R. Wong, and L. Eyres, Handling and processing shark livers for the
recovery of squalene and diacyl glyceryl ethers, DSIR Crop Research Report No. 1,
Christchurch, New Zealand, 1990.
106. R. G. Ackman, E. Macpherson, and A. Timmins. J. Amer. Oil. Chem. Soc., 75, 1223
(1998).
107. R. G. Ackman, E. J. Macpherson, and A. Timmins, J. Amer. Oil. Chem. Soc., 77, 831
(2000).
108. R. G. Ackman, R. F. Addison, and C. A. Eaton, Nature, 220, 1033 (1968).
109. H. V. Kuhnlein, F. Yeboah, M. Sedgemore, S. Sedgemore, and M. Chan, J. Food Compos.
Anal., 9, 18 (1996).
110. J. Joseph, Distribution and composition of lipids in marine invertebrates, in Ref. 24, Vol.
2, pp. 49143.
111. H-S. Seo, Y. Endo, M. Moku, K. Kawaguchi, and K. Fujimoto, Lipids, 36, 389 (2001).
112. R. Ackman, S. N. Hooper, S. Epstein, and M. Kelleher, J. Amer. Oil Chem. Soc., 49, 378
(1972).
113. P. W. Behrens and D. J. Kyle, J. Food Lipids, 3, 259 (1996).
114. J. A. Conquer and B. J. Holub, J. Nutr., 126, 3032 (1996).
115. S. M. Innis and J. W. Hansen, Amer. J. Clin. Nutr., 64, 159 (1996).
116. E. N. Frankel, T. Satue-Gracia, A. S. Meyer, and J. B. German, J. Agric. Food Chem., 50,
2094 (2002).
117. L. Poisson, M. Devos, G. Pancreach, and F. Ergan, OCL, 9, 92 (2002).
118. C. G. Bordier, N. Sellier, A. P. Faucault, and F. Le Goffic, Chromatographia, 39, 329
(1994).
12
Minor Components
of Fats and Oils
Afaf Kamal-Eldin
SLU
Uppsala, Sweden
1. INTRODUCTION
Lipids from natural sources consist mainly of fatty acids esterified to glycerol, pre-
dominantly in the form of triacylglycerols (9598%). Soluble in these lipids are two
types of minor compounds: glycerolipids and nonglycerolipids (Table 1). Upon
reaction with alkali, the unsaponifiable fractions of vegetable oils (i.e., nongly-
cerolipids) do not form soap and can be extracted from the saponified mixtures
with lipophilic solvents such as diethyl ether, hexane, or cyclohexane. Over the
past decades, these unsaponifiable materials have interested researchers for differ-
ent reasons. In the first instance, the interest in the unsaponifiable fraction stemed
from the observation that many components, especially tocopherols but also pheno-
lic and other compounds, have antioxidant properties of utmost importance for the
oxidative stability of these oils. Constituents of the unsaponifiable fraction were
also found to provide good markers for the authentication of some oils and fats.
Moreover, much research is currently focusing on the biological and physiological
activities of various unsaponifiable constituents and their possible contribution to
improved human health. Apart from the carotenoids of palm oil, vitamin E-active
compounds and phytosterols are perhaps the most important nutraceuticals present
319
320 MINOR COMPONENTS OF FATS AND OILS
TABLE 1. Important Minor Constituents of Vegetable Oils.
Glycerolipids Non-Glycerolipids
Diacylglycerols Sterols, triterpene alcohols, and their esters

Monoacylglycerols Tocopherols/tocotrienols
Phospholipids Hydrocarbons
Galactolipids Waxes
Sulfolipids Free fatty acids
Lipid-soluble vitamins
Pigments
Phenolic compounds
Metals and Metalloproteins
in vegetable oils. Tocopherols/tocotrienols are purified from the deodorizer distil-

lates to be used as vitamin E supplements in capsules or as food additives, and ster-
ols are recovered to be used as cholesterol-lowering agents in innovative foods,
such as commercialized margarines and other products (1, 2). Phospholipids, gly-
colipids, and other complex lipids are generally present as minor lipid components
in crude oils but are mostly removed by degumming during refining. These lipids
have interesting properties as emulsifiers, as discussed in more detail in other
chapters.
As shown in Table (2), vegetable oils contain variable levels of unsaponifiable
matter (3) of variable composition and characteristics (4). The refining process,
which is necessary to remove undesirable pigments and fatty acid oxidation pro-
ducts such as peroxides and their degradation products from vegetable oils to
make them suitable for consumption, unfortunately, brings about loss of valuable
nutrients and natural antioxidants (57). The technical and nutritional importance
TABLE 2. The Content and Special Composition Charcteristics of the Unsaponifiable

Fraction of Selected Vegetable Oils.
Unsaponifiable Matter (%)

Oil Schwartz (3) Special Composition Characteristics
Coconut 0.2 Not reported

Cottonseed 0.7 Not reported
Linseed 0.9 Plastochromanol-8
Olive 0.7 Squalene, special phenolic compounds
Palm 0.1 Red carotenoids
Peanut 0.4 Not reported
Rapeseed 1.0 Brassicasterol, plastochromanol-8
Sesame 1.4 Sesamin, sesamolin and derivatives
Soybean 0.6 d-tocopherol
Sunflower 0.7 Not reported
Maize 1.5 Not reported
Wheat germ 4.8 Not reported
Rice bran 4.25.2 g-oryzanol (cinnamic acid esters of sterols)
THE CHEMISTRY OF MINOR LIPID COMPONENTS 321
of many compounds in the unsaponifiable fraction is increasingly recognized, and

extraction and refining technologies are being developed to produce vegetable oils
with maximized functional and nutritional properties. This chapter reviews the
chemistry and the importance of a wide range of natural minor components occur-
ring in vegetable oils.
2. THE CHEMISTRY OF MINOR LIPID COMPONENTS
2.1. Sterols
The sterols (Table 3) are generally the major components of the unsaponifiable frac-
tions of vegetable oils. Their structures are based on a steroidal alcohol framework
comparable with that of cholesterol (Figure 1). The molecules are planar and are
based on a tetracyclic cyclopentaphenanthrene system with four fused rings (A,
B, C, and D). The hydroxyl group at C-3, side chain at C-17, and two methyl groups
at C-18 and C-19 are all angular to the ring structure and have b-stereochemistry
(i.e., above plane configuration) (8).
Plant sterols, or phytosterols, are generally the predominant compounds in the
unsaponifiable fractions of vegetable oils that generally account for about 1% of
the oils. The main sterols belong to the 4-desmethylsterols family, but 4-methyl-
sterols and 4,4-dimethylsterols (also called triterpene alcohols) are present as minor
components in most oils (Figure 2). Apart from some exceptions, the desmethyl-
sterol, b-sitosterol, is generally the most abundant and is usually accompanied by
variable levels of campesterol, stigmasterol, 5-avenasterol, and other sterols
(Table 4). Some sterols are characteristic for certain oils; e.g., brassicasterol is char-
acteristic for rapeseed/canola oil and can be used to detect the presence of this oil
in foods. Recently, it was reported that camolina oil is special because of its high
content of cholesterol (188 ppm) besides brassicasterol (133 ppm), campesterol
(893 ppm), stigmasterol (103 ppm), sitosterol (1884 ppm), and 5-avenasterol
(393 ppm) (12). Black cumin (Nigella sativa L.) is charcterized by high levels of
b-sitosterol (11351182 ppm), 5-avenasterol (9251025 ppm), and 7-avenaster-
ol (615809 ppm), and small amounts of stigmasterol and campesterol (13).
Sterols occur in vegetable oils in the free and esterified forms in relative levels
that are dependent on the type of oil. In the sterol esters of vegetable oils, the hydrogen
21 22 24
26
18 20
12 23 25
11 17
19 13 27
1 9
C D 16
2 14
10 8 15
A B
7
HO 4 6
Figure 1. Structure of cholesterol (CAS # 57-88-5) and numbering of the sterol skeleton and
side chain.
TABLE 3. Structures, Trivial and IUPAC Numbers of Common Sterols and Triterpene Alcohols in Vegetable Oils.
Trivial Name IUPAC Name CAS Numbera Structureb
4-desmethylsterols
Brassicasterol 24b-methyl cholest-5,22E-dien-3b-ol 474-67-9 Ac
Campesterol 24a-methyl cholest-5-en-3b-ol 474-62-4 Ab
Stigmasterol 24a-ethyl cholest-5,22E-dien-3b-ol 83-48-7 Ag
Sitosterol 24a-ethyl cholest-5-en-3b-ol 83-46-5 Af
5-Avenasterol 24E-ethylidene cholest-5-en-3b-ol 18472-36-1 Ah
Stigmastenol 24a-ethyl cholest-7-en-3b-ol 481-18-4 Bf
7-Avenasterol 24E-ethylidene cholest-7-en-3b-ol 23290-26-8 Bh
4-methylsterols
24-ethyl lophenol 4a-methyl-24a-ethyl cholest-7-en-3b-ol 36735-29-2 Cf
Gramisterol 4a-methyl-24-methylene cholest-7-en-3b-ol 1176-52-9 Cd
Citrostadienol 4a-methyl-24E-ethylidene cholest-7-en-3b-ol 474-40-8 Ch
Obtusiofoliol 4a,14a-dimethyl-24-methylene-cholest-8-en-3b-ol Dd
Cycloeucalenol 9,19-cyclo-4a,14a-dimethyl-24-methylene-cholest-8-en-3b-ol 469-39-6 Ed
4,4-dimethylsterols
Cycloartenol 9,19-cyclo-4,4,14a-trimethyl-cholest-24-en-3b-ol 469-38-5 Fa
24-Methylenecycloartanol 9,19-cyclo-4,4,14a-trimethyl-24-methylene-cholestan-3b-ol 1449-09-8 Fd
Cyclobranol 9,19-cyclo-4,4,14a,24-tetramethyl-cholest-24-en-3b-ol Fe
a-Amyrin 5a-urs-12-en-3b-ol 638-95-9
b-Amyrin 5a-olean-12-en-3b-ol 559-70-6
a
CAS numbers from LIPIDAT, Lipid Molecular Structure Database of Ohio State University (http://www.lipidat.chemistry.ohio-state.edu/cis888/rehner/searchLMSD_2_4.asp).
a
See Figure 2.
Sterol skeletons
Desmethyl Sterols
Sc Sc
HO HO
(A) (B)
4-Methyl Sterols
Sc Sc Sc
HO HO HO
(C) (D) (E)
4,4-Dimethyl Sterols and Triterpene Alcohols
Sc
HO HO HO
( F) -Amyrin -amyrin
Side Chains (Sc)
(a) (b) (c) (d)
(e) (f) ( g) (h)
Figure 2. Sterol skeletons and side chains of the major sterols in vegetable oils (for names and
structures, see Table 1).
TABLE 4. Sterol Content (mg/100g) of Selected Vegetable Oils (911).
Cotton Palm Rape Rice Saf- Shea Soy Sun Wheat

Sterol Coconut Corn Seed Linseed Olive Pam Kernel Peanut Seed Bran flower Sesame Butter Bean flower Germ
Desmethylsterols
Cholesterol 2 4 3 4
Campesterol 2 269 17 122 2 36 12 36 153 505 45 117 74 31 570
Stigmasterol 30 70 4 38 1 20 14 216 271 31 62 72 31
b-Sitosterol 132 772 396 193 131 189 92 154 355 885 181 382 191 235 1734
5-Avenasterol 32 47 8 55 3 5 8 19 12 90 3 43 11 16 155
7-Stigmastenol 14 12 8 6 2 1 7 31 18 70 12 94 11 59 78
7-Avenasterol 2 36 10 28 4 16 52
Others 61a 7 13 4
4-Monomethylsterols
Obtusifoliol 2 4 11 22 2 6 5 4 8 29 10 81 22 5 34 7
Gramisterol 6 4 8 108 4 24 8 4 6 25 5 142 4 16 22 47
Citrostadienol 4 4 17 6 6 3 13 4 2 3 126 35 35 53
Othersb 4 5 1 2 2 10 2 12 114 6 57 7 8 20 8
4,4-Dimethylsterols
Cycloartanol 1 tr. 1 1 1 2 106 1 4 7
b-Amyrin tr. 2 12 22 1 3 2 7 4 306 13 5 40
Butyrospermol 5 10 10 6 994 14
Cycloartenol 50 106 13 29 12 32 482 34 62 17 29 38
a-Amyrin 19 21 12 1759 18
24-Methylene- 19 30 9 612 23 12
cycloartenol
Cyclobranol 12 7 16 34 47 3 16 16 494 7 107 8 16 99
Cycloaudenol 14 10 3 2 38 6 4
Othersb 4 1 115
Total 286 1194 514 710 293 321 230 482 690 3055 428 1199 3992 502 571 2910
a
Brassicasterol.
b
Unknown compounds.
TABLE 5. Content of Free and Acyl Sterol Esters and Relative Distribution of Sterols in Selected Crude Vegetable Oils.a
Free Sterols Acyl Sterol Esters

Total Sitosterol Campesterol Sigmasterol 5-Avenasterol Total Sitosterol Campesterol Sigmasterol 5-Avenasterol
Oil (mg/100 g) (%) (%) (%) (%) (mg/100 g) (%) (%) (%) (%)
Coconut 26 71.9 11.0 17.1 41 70.4 11.1 18.5

Palm 16 59.5 22.8 11.0 6.7 49 66.0 30.0 14.0
Olive 30 65.0 4.6 30.4 151 76.9 23.1
Rapeseed 475 54.0 40.6 5.4 336 50.7 29.0 20.3
Corn 423 71.6 18.0 7.9 2.5 485 70.2 23.2 6.6
a
Modified from Verleyen et al. (15).
of the hydroxyl group at C-3 is substituted with a fatty acyl or with ferulic acid as in
g-oryzanols. Oils of corn, cottonseed, and rapeseed are especially rich in sterol
esters (14). According to this report, there are no differences in the composition
of the free and acylated sterol fractions, which suggests nonspecific esterification
(Table 5). Rice bran oil is currently recognized for its high content (about 25%)
of g-oryzanol, i.e., ferulic acid esters of mainly cycloartenol [CAS # 21238-33-5]
and 24-methylene cycloartanol, but also cyclobranol, cycloeucalenol, sitosterol, and
campesterol (16). Two-thirds of the total sterols of wheat germ oil, 20004000 mg/
100 g, are present in ester form (17).
Refining affects the sterols of vegetable oils in various ways and is responsible
for sterol losses in the range of 1070% (18). Sterols are partially washed with the
soap stock after chemical neutralization (5, 19). The use of acid clay bleaching
agents and elevated temperatures catalyzes different isomerization, dehydration,
and esterification reactions. Tisconia and Bertini (20) observed a very remarkable
change in the content of 5-avenasterol in olive oil during bleaching depending on
the level of bleaching earth used, and Touche et al. (21) made a similar observation
for bleached coconut oil. Kesselmeier et al. (22) later suggested that 5- and 7-
avenasterols present in oat lipids are destroyed by acid hydrolysis. Kamal-Eldin
et al. (23) explained the mechanism of this transformation as involving secondary
and tertiary carbonium ions (Figure 3). Bleaching effects on phytosterols are
generally minor and mainly limited to the formation of some nonpolar dehydration
products (18, 24) and partial hydrolysis of sterol esters (25). Steradienes and
H+
Fucosterol
+H+
H+
+H+
+
5-Avenasterol
H+
5,23-Stigmastadienols a & b
H+
5,24(25)-Stigmastadienol
Figure 3. Acid-catalyzed isomerization of 5-avenasterol [modified from Kamal-Eldin et al.

(23)].
R R
, H +
HO H2O
sterol H H
H2O , H + H2O H +
R R R
O
disteryl ether 3,5-steradiene
Figure 4. Transformation of sterols to steradienes and disteryl esters catalyzed by heat and acid
in anhydrous conditions.
disteryl ether dehydration products (Figure 4) are formed during bleaching as well
as deoderization (14). Dehydration products of triterpene alcohols were also iso-
lated from stearin fractions of refined shea butter (26). Deodorization, on the other
hand, causes a significant reduction in the total sterol because of distillation (5, 19,
27) and esterification of free sterols (14, 25). Whenever applied, hydrogenation has
a tremendous effect on sterol structures, including hydrogenation of double bonds,
opening of cyclopropane rings, and positional isomerization of side chain unsatura-
tion (2833).
Phytosterols are industrially isolated from the distillates, resulting from the deo-
dorization of vegetable oils (1, 3436). Phytosterols are sometimes hydrogenated to
produce phytostanols (37). As the solubility of sterols and stanols is very low (<1%
at 25 C), it limits their application in food products. Esterification of sterol and sta-
nols is, therefore, performed to make them fat-soluble and easy to incorporate in
food products (37, 38). Two margarines containing 89% sterols (Becel Proactiv
of Unilever) or stanols (Benecol of Raisio), in the form of esters, are now available
in the markets in Europe and the United States.

The tocopherols and tocotrienols are generally not the major components of
vegetable oil, but their presence is vital for stabilizing the unsaturated fatty acids
of these oils against oxidative deterioration (39). Their structures are based on a
chroman head with two rings, one phenolic and one heterocyclic, and a phytyl
tails isoprenoid side chain at C-2. The phytyl tail is saturated in the case of
Isoprenoid Side Chains (Sc)
Chromanol skeleton CH3 CH3 CH3

R1
4
4 8 CH3
5 1
HO 6
3 Tocopherols
Sc
R2 8 O
1 CH3
CH3 CH3 CH3 CH3
CH3
1 n
Tocotrienols and Plastochromanol-8
Tocopherol Tocotrienol R1 R2
-tocopherol -tocotrienol CH3 CH3
-tocopherol -tocotrienol CH3 H
-tocopherol -tocotrienol H CH3
-tocopherol -tocotrienol H H
Figure 5. Chromanol ring and isoprenoid side chains in tocopherols, tocotrienols (n 3),
and plastochromanol-8 (n 8, R1
H, R2
CH3 ) in vegetable oils (for systematic names, see
Table 6).
TABLE 6. Structures, Trivial, and IUPAC Numbers of Tocopherols, Tocotrienols,

and Related Compounds Present in Vegetable Oils.
Trivial Name IUPAC Name CAS Number
Tocopherols
a-tocopherol 3,4-Dihydro-2,5,7,8-tetramethyl-2-(40 ,80 ,120 - 59-02-9
trimethyltridecyl)-benzopyran-6-ol
b-tocopherol 3,4-Dihydro-2,5,8-trimethyl-2-(40 ,80 ,120 -trimethyltridecyl)- 16698-35-4
benzopyran-6-ol
g-tocopherol 3,4-Dihydro-2,7,8-trimethyl-2-(40 ,80 ,120 -trimethyltridecyl)- 54-28-4
benzopyran-6-ol
d-tocopherol 3,4-Dihydro-2,8-dimethyl-2-(40 ,80 ,120 -trimethyltridecyl)- 119-13-1
benzopyran-6-ol
Tocotrienols
a-tocotrienol 3,4-Dihydro-2,5,7,8-tetramethyl-2-(40 ,80 ,120 -trimethyl- 1721-51-3
30 ,70 ,110 -tridecatrienyl)-2H-1-benzopyran-6-ol
b-tocotrienol 3,4-Dihydro-2,5,8-trimethyl-2-(40 ,80 ,120 -trimethyl-30 ,70 ,110 - 490-23-3
tridecatrienyl)-2H-1-benzopyran-6-ol
g-tocotrienol 3,4-Dihydro-2,7,8-trimethyl-2-(40 ,80 ,120 -trimethyl-30 ,70 ,110 - 91-86-1
d-tocotrienol 3,4-Dihydro-2,8-dimethyl-2-(40 ,80 ,120 -trimethyl-30 ,70 ,110 -
Plastochro- 3,4-Dihydro-2,7,8-trimethyl-2-(40 ,80 ,120 ,160 ,200 ,240 ,280 ,320 -
manol-8 octamethyl-30 ,70 ,110 ,150 ,190 ,230 ,270 ,310 -octadecaoctenyl)-
2H-1-benzopyran-6-ol
tocopherols and unsaturated in the case of tocotrienols and other derivatives such as
plastochromanol (Figure 5, Table 6). The four members of each subfamily, i.e., a-,
b-, g-, and d-, differ from each other in the number and position of methyl groups
on the chromanol ring. Naturally, tocopherols occur only as free alcohols, but toco-
trienols were mentioned to also occur in esterified forms (40). The tocopherol mole-
cule has three chiral centers in its phytyl tail, giving the possibility of eight total
stereioisomeric forms. Naturally occurring tocopherols have the same molecular
configurations (2R, 40 R, 80 R, RRR, or d-) in their phytyl groups, and tocopherols
obtained by synthesis (all-rac-tocopherols) represent a mixture of approximately
equal amounts of the eight possible stereoisomers [2D,40 D,80 D (RRR),
2L,40 D,80 D (SRR), 2D,40 D,80 L (RRS), 2L,40 D,80 L (SRS), 2D,40 L,80 D (RSR),
2L,40 L,80 D (SSR), 2D,40 L,80 L (RSS), and 2L,40 L,80 L (SSS)] (41). The tocotrienols
only have one chiral center at position 2, so they are either 2D or 2L stereoisomers.
The presence of the double bonds at positions 30 and 70 of the phytyl tails of toco-
trienols allows for the existence of four cis/trans geometrical isomers, i.e., a total of
eight isomers [2D,30 cis,70 cis (R,cis-cis); 2D,30 cis,70 trans (R,cis-trans);
2D,30 trans,70 cis (R,trans-cis); 2D,30 trans,70 trans (R,trans-trans); 2L,30 cis,70 cis
(S,cis-cis); 2L,30 cis,70 trans (S,cis-trans); 2L,30 trans,70 cis (S,trans-cis); 2L,30 trans,
70 trans (S,trans-trans)] per tocotrienol (42). It is, therefore, easy to detect the
presence of synthetic tocopherols/tocotrienols in lipids.
The tocopherol and tocotrienol contents of selected vegetable oils are shown in
Table (7). Seed oils are mostly dominated by g- or a-tocopherol, but high levels of
d-tocopherol especially characterize soybean oil, which is the richest and commer-
cially the most used source of tocopherols (46). Rice bran and palm oils, on the
other hand, represent the important sources of tocotrienols (45, 47). Rapeseed
TABLE 7. Levels (mg/100 g) of Tocopherol, Tocotrienols, and Related Compounds

in Selected Vegetable Oils.
Oil a-T b-T g-T d-T a-T3 b-T3 g-T3 d-T3 Total
Corn 26 1 75 3 1 2 108
Cottonseed 40 38 tr. 1 79
Linseeda tr. 57 1 58a
Olive 12 1 13
Palm 38 5 13 56
Peanut 14 2 13 1 30
Rapeseedb 19 49 1 69b
Safflower 45 1 3 1 50
Sesame 58 1 59
Soybean 9 1 69 24 103
Sunflower 62 2 3 67
Maize 22 57 2 5 1 6 93
Rice branc 59 2 6 tr. 45 44 3 159
Wheat germ 151 31 53 4 2 241
a
Linseed oil also contains ca 20 mg/100 g of plastochromanol-8 (43).
b
Rapeseed oil may also contain 5580 mg/100 g of plastochromanol-8 (44).
c
Data from (45).
22
14
18
6
10
Figure 6. Structure of squalene (all-trans-2,6,10,15,19,23-hexamethyl-2,6,10,14,18,22-tetraco-

sahexaene, CAS # 111-02-4), a major hydrocarbon in olive oil.
and linseed oils contain the special, tocotrienol-like, plastochromanol-8 (4648).

Cereal lipids, namely, wheat germ, maize, and rice bran oils, generally contain
very high levels of tocopherols and tocotrienols. Of nut oils, almonds and hazlenut
ois are rich in a-tocopherol and pecans and walnuts are rich in g-tocopherol (49).
2.3. Squalene, Carotenoids, and Other Hydrocarbons

Squalene (2,6,10,15,19,23-hexamethyl-2,6,10,14,18,20-tetracosahexaene), also known
as spinacene, is a naturally occurring 30-carbon chain triterpenoid hydrocarbon
(Figure 6). Shark liver oil is the richest source of squalene, about 50% by weight
(50, 51). Squalene is present at levels of 0.11.2% in olive oils together with low
levels of other hydrocarbons (52). Squalene is found in smaller quantities, <30 mg/
100 g, in other vegetable oils (53). Squalene is used in health foods, but it is hydro-
genated to squalane for use, as a moisturizing agent, in cosmetics because of its
high susceptibility to oxidation (54, 55). The double bonds of squalene undergo
acid-catalyzed isomerization during vegetable oil refining to produce a large num-
ber of positional isomers (56). Other hydrocarbons representing a homologous ser-
ies of saturated C15C33 chains are also present in vegetable oils. For example,
olive oil was reported to contain odd-numbered n-alkanes (C23C29, 30180
ppm), a series of C22-C27 n-9-alkenes, 8-heptadecene 6,10-dimethyl-1-undecene
(0.52 ppm), as well as about 31 sesquiterpenes (237 ppm), the most abundant of
which were a-farnesene, a-copaene, eremophyllene, and a-muurolene (57).
Carotenoids are the yellow-red pigments present in most crude oils, although
their presence is often masked by the green color of chlorophyll. They are com-
posed of a long-chain, 40-carbon skeleton, of eight isoprenoid units joined head-
to-tail to give a completely conjugated system with alternated double bonds
(Figure 7). Biosynthetically, carotenoids are derived from this acyclic long-chain
conjugated C40H56 structure of lycopene by hydrogenation, dehydrogenation, cycli-
zation, and oxidation, or a combination of these processes (58). The carotenoids can
be classified into (1) nonpolar unsaturated carotenes with the basic structure
of lycopene, and (2) more polar carotenoids, or xanthophylls, with oxygen function
at one or both end groups. The rules for semisystemstic nomenclature of carote-
noids have been published by the IUPAC-IUB Commission on Biochemical
Nomenclature (59).
-carotene
-carotene
Lycopene
OH
-cryptoxanthin
OH
HO lutien
OH
HO zeaxanthin
Figure 7. Structure of some major carotenoids present in vegetable oils, mainly red palm oil and
olive oil.
Red palm oil is the richest source of readily available carotenoids and is, therefore,
very useful as a pro-Vitamin A supplement. Crude palm oil contains 500800 ppm
of carotenoids, of which b-carotene and a-carotene account for about 90%
(approximately 2:1 w/w), and lycopene, phytoene, and zeacarotenes are inter alias
the remaining carotenoids (60). Other vegetable oils contain much lower levels
of carotenoids (<100 ppm), but these are removed during the bleaching step in
refining. Olive oil contains variable, but low, levels of carotenoids mainly as
b-carotene (617% of total pigments) and lutein (1838% of total pigments), but
xanthophylls such as neoxanthin, violaxanthin, luteoxanthin, antheraxanthin, muta-
toxanthin, and b-cryptoxanthin also occur at low levels (6163). Marine oils, inter
alias cupelin and salmon oils, also contain large amounts of carotenoids, mainly
xanthophylls (64).
OH OH
R1 R2 R1 R2
COOH
Hydroxybenzoic acid derivatives COOH
p-hydroxy benzoic acid (R1 & R2 = H)

Hydroxycinnamic acid derivatives
procatechuic acid (R1 = OH3 , R2 = H)
vanilic acid (R1 = OCH3 , R2 = H) p-coumaric acid (R1 & R2 = H)
gallic acid (R1 = OH3 , R2 = OH) caffeicic acid (R1 = OH , R2 = H)
syringic acid (R1 = OCH3 , R2 = OCH3) ferulic acid (R1 = OCH3 , R2 = H)
HO
O O O O
HO COOCH3 HO COOCH3
O O
OH O OH
Ligstroside Oleuropein isomer
O O
HO COOCH3
R
OH
HO O
O
p-hydroxyphenyl ethanol (p-HPEA, R = H) p-Hydroxyphenyl ethanol-
3,4-dihydroxyphenyl ethanol (3,4-DHPEA, Eleanoldialdehyde (p-HPEA-EDA,
R = OH) R = H)
3,4-Dihydroxyphenyl ethanol-
Eleanoldialdehyde (3,4-DHPEA-EDA,
R = OH)
Figure 8. Chemical structures of secoiridoid derivatives and phenolic alcohols in virgin olive oils.
2.4. Phenolic Compounds

Low levels of a wide range of phenolic compounds have been reported to be present
in all vegetable oils, which is very important for the oxidative stability of the poly-
unsaturated fatty acids of these oils. The mostly studied phenolic constituents are
perhaps those of olive oil (Figure 8). The level of phenolic compounds in olive oil
O O O
O O O
O O O
H H+ H H+ H
H H H
O O O
O O O
O O O
Sesamin Episesamin Diasesamin
O O O
O O O
O O O
H H+ H H+
OH H OH
H H H
O O O O
O O
O O O
Sesamolin O Sesaminol 6-Episesaminol
H+ H+ H+
O O
O O
OH
O O
H+
H OH H OH
H H
O
O O O
O O
O O
Sesamol 6-Episesaminol Diasesaminol
Figure 9. Sesame seed lignans (sesamin and sesamolin) and their acid-catalyzed transforma-
tion products generated during bleaching.
Figure 10. Changes in sesame oil lignans induced by refining [data from Fukuda et al. (75)].
was reported to vary between 50 and 1000 ppm depending on agronomic and envir-
onmental factors, ripeness, processing, and storage conditions of the oil (65), but
they mostly lie in the range of 200500 ppm (66). The phenolic compounds of olive
oil include the secoiridoid aglycons of oleuropein and ligstroside and the lignans
1-acetoxypinoresinol and pinoresinol (6772). Tyrosol, hydroxytyrosol, and their
esters are also major phenols of olive oils (63) and their levels were found to
increase during storage of virgin olive oil (73) because of acid-catalyzed hydrolysis
of secoiridoid aglycons (69).
Crude sesame oil is characterized by the presence of two unique lignans:
sesamin [2,6-(3,4-methylenedioxyphenyl)-cis-2,7-dioxabicyclo-[3.3.0]-octane] and
sesamolin [2-(3,4-methylene-dioxyphenoxy)-6-(3,4-methylenedioxyphenyl)-cis-2,7-
dioxabicyclo-[3.3.0]-octane] with interesting physiological effects (Figure 9).
Sesamin and sesamolin are present in the range of 0.10.9% and tr.0.7%, respec-
tively (74). Fukuda et al. (75) showed that when sesame oil is refined, sesamin
is partly isomerized into episesamin and sesamolin is hydrolyzed to small amounts
of sesamol but is mainly lost and isomerized into three sesaminol isomers (Fig-
ure 10). These transformations are catalyzed by the acidic residues in the bleaching
earth, and they can be used inter alias to differentiate crude from refined sesame
oils.
2.5. Other Minor Compounds

High aliphatic alcohols and wax esters in which aliphatic alcohols or sterols are
esterified to fatty or phenolic acids are also present in crude vegetable oils at low
levels and are partially removed in the winterization process during oil refining.
Waxes, mainly esters of long-chain saturated fatty acids and a monounsaturated
alcohol, especially eicosenoic alcohol, are found in crude vegetable oils such as
olive, sunflower, soybean or peanut but are absent from corn or rice bran oils
(76, 77). Diterpenic, phytyl, and geranylgeranyl esters were reported in olive and
sunflower oils, but benzyl esters were only detected in olive oils (78). The wax
esters of sunflower oil were found to have carbon atom numbers between 36 and
48, with a high concentration in the C-40C-42 fraction (79). Waxes in dewaxed
and refined sunflower oils mainly contained <42 carbon atoms indicating selective
retention after refining.
Vegetable oils, especially in crude forms, contain variable levels of chlorophyll
pigments. The maturity of the oilseed and the method of oil extraction determine
the content of chlorophyll pigments in the oil. For example, unripe rapeseed con-
tains chlorophyll A, chlorophyll B, pheophytin A, and pheophytin B as major pig-
ments and pheophorbide A, methylpheophorbide A, and pyropheophytin A as
minor pigments, whereas ripe seeds are characterized by chlorophylls A and B
(80). Industrial extraction and refining of chlorophyll pigments shows selectivity
in its effects on chlorophylls and pheophytins so that latter prevail in solvent
extracted and bleached oils (80, 81). Pheophytin A is the major pigment in olive
oil (4458% of total pigments) that co-occur with small amounts of pheophytin
B and sometimes chlorophylls A and B (61, 62).
Free fatty acids and partial glycerides (mono- and di-acylglycerols) occur in dif-
ferent levels in crude oils but are removed by refining from most oils, except brands
of olive, sesame, and nut oils. Phospholipids, or phosphatides, are important minor
lipid components because of their profound surfactant effects with wide applica-
tions in, inter alia, the baking industry. Lecithin (phosphatidyl choline) and other
phospholipids (phosphatidyl ethanolamine, phosphatidyl inositol, phosphatidyl
serine, phosphatidic acid, etc.) are important byproducts of vegetable oil refining,
especially soybean oil. The presence of phospholipids and other polar lipids
is, however, not desirable in refined vegetable oils. Currently, methods are being
developed for removal of phospholipids from seed oil by physical refining
(82), membrane filtration (83, 84), or enzymatic degumming by phospholipase
(85, 86).
Cereal lipids are special because of their content of unusually high amounts of
waxes, free fatty acids, unsaponifiable matter, sterol glucosides, diacylglycerols,
phospholipids, and glycolipids (87, 88), which cause difficulties in the refining
process. The main problem seems to be because of the high content of glycolipids
and phosphorus-containing glycolipids (up to 5%), which cause high losses during
alkali refining because of their high surface activity. A sample of crude wheat germ
oil contained 1428-ppm phosphorus, 15.7% free fatty acids, and 2682-ppm toco-
pherols. The oil was refined by conventional degumming, neutralization, bleaching,
and continuous tray deodorization, and the neutralization step was found to signif-
icantly remove the free fatty acids and most of the phospholipids (89). A special
physical refining procedure based on simultaneous dewaxing and degumming
was recently developed for rice bran oil (90).
Gormet oils (e.g., olive oil, sesame oil, avocado oil, and nut oils) that are
consumed as virgin oils usually have characteristic flavors that distinguish them
from other edible vegetal oils. In olive oil, these compounds included, among a
wide range of compounds, hexane, heptane, octane, ethanol, isobutanol, pentenol,
hexenol, octanol, pentanal, hexanal, 2-propenal, (E)-2-pentenal, (E)-2-hexenal, (E)-

2-heptenal, octanal, (E)-2-octenal, nonanal, (E)-2-nonenal, isovaleraldehyde,
decanal, (E)-2-hexen-1-ol, 2-propanone, 3-pentanone, acetic acid, formic acid, hex-
anoic acid, heptanoic acid, and limonene (91). Hazelnut oil is characterized by the
strong flavor of 5-methyl-(E)-2-hepten-4-one (92). More than 30 volatile com-
pounds, including pyrazines, thiazoles, pyridines, furfurals, and oxazoles, were
identified as flavor compounds in sesame oils obtained from roasted seeds (93
96). Pyrazines are the major compounds and are responsible for the roasted flavor,
and furfurals are responsible for the sweet candy-like flavor (97).
3. SIGNIFICANCE OF MINOR LIPID COMPONENTS
3.1. Technical Quality of Vegetable Oils

The unsaponifiable matter (e.g., sterols, tocopherols, lipophilic pigments, and
hydrocarbons) generally can make up to 2.5% of vegetable oils. When free fatty
acids are present at critically high concentrations, they may contribute an undesir-
able taste to oils that are used in crude forms, e.g., olive oil, sesame oil, and nut oils.
This effect markedly results if free short-chain acids (C6C10) are present in coc-
nut, palm-kernel, and butter oils. Free fatty acids also negatively affect the foaming
properties and stability attributes of these oils (98). As another example, olive oil
polyphenols have been associated with the bitter taste of the oil but longer shelf life
(99). On the other hand, minor lipids (such as diacylglycerols, monoacylglycerols,
cholesterol, and phospholipids) are known to significantly affect the interfacial
properties and crystallization of fats even at the low concentrations at which
they are naturally present in most oils (100, 101). They were found to positively
influence crystallization of cocoa butter by affording spherical crystals of regular
shape (102).
Vegetable oils are variably susceptible to free radical autoxidation depending on
the degree of unsaturation of their fatty acids and their contents of prooxidant and
antioxidant species. The unsaponifiable fractions of wheat germ, corn, or olive oil
are known to retard oxidation in vegetable oils (103). Trace amounts of metals and
metalloproteins occurring in vegetable oils catalyze the oxidaton by enhancing
initiation and re-initiation by decomposing lipid hydroperoxides. The antioxidant
activity of tocopherols is well documented, albeit considerable controversy in con-
nection with the relative antioxidant potency of different tocopherols exists (39).
The commencement of exponential oxidation of vegetable oils is associated with
critical minimal concentrations of antioxidants (104). For example, a significant
increase in the rate of oxidation was observed when the concentration of a-toco-
pherol in olive oil reaches a threshold value of 6070 mg/kg (105). The main
reaction responsible for the antioxidant action of tocopherols is hydrogen atom
donation (39). The antioxidant (AH) tocopherols and tocotrienols scavenge lipid
peroxyl radical (LOO ) to form a relatively stable lipid hydroperoxide (LOOH).
Then the tocopheroxyl or tocotrienoxyl antioxidant radicals (A ), formed by this
SIGNIFICANCE OF MINOR LIPID COMPONENTS 337
reaction, scavenge another peroxyl radical through radicalradical coupling to form

a variety of stable secondary oxidation products.
LOO AH ! ROOH AO
LOO AO ! stable products
Besides scavenging peroxyl radicals, tocopherols and tocotrienols are also excellent
quenchers for singlet oxygen and peroxynitrile (39). It has been reported that
a-tocopherol shows a pro-oxidant activity when present at high concentrations
(106108). It was later found that tocopherols do not show a pro-oxidant effect
if they are compared with controls devoid of them and other lipid-related antioxi-
dants (109, 110). However, tocopherols lose their antioxidant efficacy at high con-
centrations because they participate in a number of reactions that lead to production
of alkoxyl radicals that consume tocopherols and initiate new reaction chains
because of their low selectivity (111, 112). Using a kinetic approach, a number
of reactions were found to contribute to the loss of tocopherol antioxidant activity
that is greater for a-tocopherol than for g-tocopherol (113).
AH LOOH ! A LO H2 O
AOOL ! AO LO
A LOOH ! AH LOO
A LH ! AH L
AH O2 ! A HO2
A O2 ! AOO
g-Tocopherol is generally a better antioxidant for vegetable oils than a-tocopherol,

both at low-temperature storage and under thermo-oxidation conditions (114). It
was found that g-tocopherol degrades at a much slower rate and, thereby, is able
to protect the oil longer compared with a-tocopherol (110, 111).
Photo-oxidation plays a significant role in the oxidation of oils rich in chloro-
phyll and other photosensitizers. However, tocopherols, carotenoids, and polyphe-
nols present in these oils provide considerable protection against light-induced
photo-oxidation. These different antioxidants work synergistically and have differ-
ent protection abilities. The best example for this situation is olive oil (115). Pheo-
phytin A and b-carotene were found to be sensitive to oxygen, light, and
temperature, and a-tocopherol and polyphenols were more stable toward high tem-
peratures. The chlorophyll pigments function as photosensitizers and b-carotene as
an efficient quencher both for the excited chlorophylls and singlet oxygen (116).
However, a-tocopherol and polyphenols are needed to protect b-carotene against
destruction caused by heat and excessive oxygen (117119). During photo-oxida-
tion, squalene was found to protect a-tocopherol (120).
Paradoxically, chlorophyll was found to synergize the antioxidant effect of
a-tocopherol in the dark (121), but chlorophyll degradation was found to contribute
negatively to the stability of vegetable oils (122127). High chlorophyll levels in

rapeseed oil were also found to decrease oil stability (126) and to increase the
rate of tocopherol decomposition and formation of polymers when the oil was
heated at 180 C (114). Chlorophyll pigments can be removed from vegetable oil
inter alias by bleaching using montmorillonite (128, 129) or membrane processing
(130). Besides affecting the oxidative staility of vegetable oils, chlorophyll pig-
ments also affect their bleachability during refining and add to refining costs (81).
b-Carotene and other carotenoids are known to possess antioxidant activity
under low temperatures and oxygen pressure conditions (131). The main mechan-
ism by which carotenoids exert their antioxidant effects is by quenching reactive
singlet oxygen and excited photosensitizers (132143). The second mechanism is
by scavenging peroxy and other reactive free radicals (131, 144150). Exposure of
b-carotene to free radical initaitors caused the chain reactions to differ from the
zero-order reaction kinetics charcterizing the induction period of autoxidation reac-
tions (151154). Although the mechanisms for singlet oxygen quenching by the
carotenoids are fairly well understood and defined, mechanisms for their reactions
with free radicals are not yet clarified. These reactions seem to be influenced in a
very complex manner by a number of factors such as oxygen tension, temperature,
and presence of enzymes, among others (153157). High concentrations of
b-carotene were reported to lack antioxidant activity and to cause an increase in
peroxide levels in rat plasma and liver (158) and to increase the thiobarbituric
acid reactive substances (TBARS) and 15-lipoxygenase activity in rat testis (159).
b-Carotene was also found to act as a prooxidant promoting lipid peroxidation in
some systems under high oxygen partial pressure (131, 147, 155). The high reac-
tivity of carotenoids with oxygen, especially at high temperatures, poses a great
limitation on their use as antioxidants. However, when tocopherols are present or
scavenge radical species formed from the reaction of carotenoids with oxygen, a
significant synergistic antioxidant effect can be achieved (143, 160162).
Several lipid minor components are known to synergize the protective effect of
tocopherols and protect the unsaturated fatty acids from being rancid. These include
phospholipids (163) and unsaponifiable components such as sterols, carotenoids,
and squalene (114, 164166). The phenolic compounds of olive oil have parti-
cularly received much attention with respect to their effects on the oxidative stabi-
lity of the oil (63, 72, 117, 167171). A study based on virgin olive oil, from
Cornicabra variety obtained from three successive crop seasons (94/95 to 96/97),
showed a clear influence of total polyphenols on the oil stability and a much lower
contribution of a-tocopherol (172). Univariate analysis revealed a significant corre-
lation between Rancimat1 stability and the following compositional parameters:
phenols (R 0.87), ortho-diphenols (R 0.77), oleic/linoleic ratio (R 0.71),
total tocopherols (R 0.65), chlorophylls (R 0.68), and carotenoids (R 0.59)
(165). When a stepwise linear regression analysis was employed, a high multicoli-
nearity was found for total phenols, oleic/linoleic ratio, and tocopherols as having
together the maximum correlation with stability. The regression equation for
adjusted-R2 0.91 was 18.52 0.13 total phenols 3.03 oleic/linoleic ratio
0.07 total tocopherols (165).
3.2. Nutritional Value of Minor Substances in Vegetable Oils
During the early 1950s, it was reported that phytosterols lower serum cholesterol
(173175). This effect was appreciated as a possible protection strategy against car-
divascular disease risk after the results of several convincing animal and human stu-
dies (176184). Studies have shown that a daily intake of 2-g phytosterol or
phytostanol causes 4050% reduction in the dietary cholesterol absorption, 6
10% reduction in total serum cholesterol, and 814% reduction in the serum
low-density lipoprotein cholesterol (37, 185187).
There is some evidence that g-oryzanol present in rice bran oil lowers serum
total- and LDL cholesterol as well (16, 185, 188). Butter from the Shea tree, Butyro-
spermum parkii, contains a very high level of 4,4-dimethylsterols (about 8%) mostly
as esters of cinnamic acid (oryzanols). Apparent absorption of these sterols, as
estimated by their disappearance from feces, was found to be 2752% in Wistar
rat and 13 49% in humans (189). It was found that the cinnamic acid esters of
the 4-desmethylsterols of rice bran oils, but not those of the 4,40 -dimethylsterols
of shea nut oil possess hypolipidemic activity (190).
The mechanism for the inhibition of cholesterol absorption is thought to involve
competitive transfer to the micellar phase during absorption from the intestinal
lumen. Phytosterols in the micellar phase may also act as emulsifying agents that
selectively inhibit the transfer of cholesterol and other lipids (e.g., carotenoids and
vitamins) and, thereby, limit their absorption. The exact kinetics governing the ster-
ol competition for transfer are not known, but dietary sterols are absorbed differ-
ently in the digestive tract: 4050% for cholesterol, 1216% for campesterol,
45% sitosterol, and <0.5% for phytostanols (37). Before absorption, esterified
sterols are hydrolysed effectively in the upper intestine (191). Absorbed phytosterols
are excreted by the liver into the bile but are hardly converted to bile acids (192).
Numerous studies in animals and humans approved the safety of phytosterols
and phytostanols (37).
Tocopherols are well known for vitamin E activity since the early work of Evans
and Bishop (193). Vitamin E compounds (tocopherols and tocotrienols) are well
known for their strong inhibitory effects against lipid oxidation in foods and biolo-
gical systems (41, 194205). As Vitamin E is only synthetized by plants, humans
and animals have to satisfy their nutritional needs by eating plant sources rich in
this vitamin (206). The richest sources of tocoperols are by far vegetable oils fol-
lowed by other sources such as nuts, cereal grains, green vegetables, and milk,
among others (207209).
A number of epidemiological studies suggest that Vitamin E has a moderate pro-
tective effect against the progression of cardiovascualr diseases (210215). The
antioxidant hypothesis suggests that the inhibition of low-density lipoprotein oxida-
tion is the main mechanism by which Vitamin E exerts this protective effect (187,
216) by scavenging the chain-carrying lipid peroxy radicals (217220). Other bio-
chemical mechanisms, distinctive of Vitamin Es antioxidant properties, have also
been proposed to explain the inhibitory effects on cardiovascular disease. One
hypothesis stated that larger dosages of Vitamin E (>400 IU/day) inhibit platelet
adherence to proteins such as fibrinogen, fibronectin, and collagen (221). Another

hypothesis states that Vitamin E controls smooth muscle cell proliferation by acting
as a sensor and information transducer of the cells oxidation state (222). At opti-
mum physiological concentrations, a-tocopherol is thought to bind to a receptor
protein resulting in the activation of transcription factor AP-1, which leads to the
expression of a protein phosphatase that dephosphorylates protein kinase C, forcing
its own inhibition and resulting in the inhibition of smooth muscle cell prolifera-
tion. In a highly oxidative environments, a-tocopherol diminishes because of its
action as a radical scavenger, and as a result of this diminution, less a-tocopherol
binds to its receptor protein resulting in a greater activity of protein kinase C and,
thereby, a higher amount of smooth muscle cell proliferation (222). a-Tocopherol
was also reported to inhibit the secretion of interleukin-1b and the adhesion mono-
cyte to endothelium (223).
Currently, much attention is paid to g-tocopherol, the major tocopherol supplied
by most vegetable oils. The amount of g-tocopherol in the diet of most humans is
approximately twice that of a-tocopherol (224), but its plasma levels are only about
1020% of those of a-tocopherol (225). This difference was explained by biodis-
crimination by a specific tocopherol-binding protein with special preference for the
a-homologue (226). Upon feeding equal amounts of the two vitamers to experi-
mental rats, it was found that a-tocopherol is preferentially bound to the cellular
membranes of the liver and to the transporting proteins while g-tocopherol is
excreted through the bile because of very low incorporation into the transporting
protein (227). g-Tocopherol may have some specific biological activities that
may complement those of a-tocopherol (228). For example, g-tocopherol was
found to inhibit nitrogen dioxide (NO2)-mediated nitrosation of morpholine
much more significantly than a-tocopherol (229231). Nitrogen dioxide, one major
constituent of cigarette smoke, is a mutagenic substance reactive toward lipids,
DNA, and other biological molecules (232). Relevant to this perhaps is the finding
that smoking, which is a recognized risk factor for cancer and cardiovascular dis-
ease, causes a decrease in g-tocopherol levels in LDL and HDL particles (233).
Recently, 5-nitro-g-tocopherol was found to increase in Alzheimer brain, which
suggests a major role of g-tocopherol as a scavenger of nitrogen reactive species
(234). A study from Loma Linda University showed that a human endogenous
natriuretic factor, LLU-a, possibly a product from the in vivo oxidative metabolism
of g-tocopherol, inhibited the 70-ps K channel in the apical membrane of the thick
ascending limb of the kidney (235). The effect of this natriuretic factor on the con-
trol of the body pool of extracellular fluids, and its implications on hypertension,
congestive heart failure, and cirrhosis remains to be investigated. Swedish patients
suffering from coronary heart disease were found to have significantly lower serum
g-tocopherol concentrations compared with healthy controls, and a-tocopherol
levels were not significantly different (236). Similar findings were reported by other
authors, which suggests a possible role of g-tocopherol as a marker of atheroscle-
rosis (237239).
Many studies suggest that a- and g-tocopherols influence each others post-
absorption transport and metabolism, and some studies suggested that g-tocopherol
might transform into a-tocopherol in vivo. For example, supplementation with all
racemic-a-tocopheryl acetate was found to cause a significant decrease in g-toco-
pherol levels and a significant increase in the a-/g-tocopherol ratio in human tissues
and adipose tissue. This effect was found to disappear when the supplementation
was stopped (240, 241). On the contrary, g-tocopherol supplements were shown
to induce a marked increase in a-tocopherol concentrations in serum and tissues
of Vitamin E-deficient rats (242). A national health survey of 18,000 Germans,
age 2569 years, showed that g-tocopherol levels were greatly reduced in persons
taking supplements containing >50-mg a-tocopherol/day (243). Based on the
results, these authors pointed out that mixed tocopherols should be considered
when tocopherol supplemental medication is considered necessary. When g-toco-
pherol was supplied continuously in the diet, it accumulated significantly in rat tis-
sues but to a much lesser extent than a-tocopherol (244). Furthermore, g-tocopherol
was found to accumulate in the plasma in relatively very high concentrations in
cases of a-tocopherol deficiency (245, 246). The studies by Elmadfa et al. (247)
show that rats fed high doses of g-tocopherol were able to survive for four genera-
tions because of an in vivo conversion of g-tocopherol to a-tocopherol. Yamashita
et al. (248, 249) showed that a combination of g-tocopherol and sesamin is able to
produce Vitamin E activity equivalent to that of a-tocopherol in rats.
Together with g-oryzanol, tocotrienols are responsible for the cholesterol-lower-
ing effect of rice bran oil (250). Tocotrienol concentrates have been shown to have a
hypocholesterolemic effect in animals and humans (251257) possibly because of
inhibition of hepatic 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMG-CoA
reductase) (251, 258, 259). The presence of a-tocopherol at concentrations
>20% in tocotrienol concentrates was, however, found to attenuate the inhibitory
effect of tocotrienols on HMG-COA reductase, thereby weakening their cholesterol-
lowering activities (254). Tocotrienols were especially found to synergize the cho-
lesterol-lowering effect of lovastatin (254). In addition, tocotrienols have been
shown to influence certain hemostatic parameters and to reduce the occurrence
of chemical-induced tumors in the rat (253).
Carotenoids have long been known as food colorants and as precursors of the
Vitamin A retinoids in animals (260, 261). Red palm oil is the richest source of
biologically available carotenes, which are efficient in restoring Vitamin A activity
in malnurished preschool children (262). The importance of carotenoids and reti-
noids for vision is well documented (263, 264). The carotenoids are now receiving
considerable interest because they are believed to have potential in delaying or pre-
venting degenerative diseases such as atherosclerosis (146, 265, 266) and some
types of cancer (266271), in modulating immune response (272274) and in pro-
viding other beneficial effects such as ulcer inhibition and life extention (275277).
b-Carotene and other hydrocarbon carotenoids are poorly-to-moderately
absorbed from the gastrointestinal tract (550%), but highly polar carotenoids
are poorly absorbed (278). Ingested pro-Vitamin A carotenoids (b-carotene and
other carotenoids containing b- or 3-retinylidene end groups) are partly cleaved
in the intestinal mucosa to retinoids through the oxidative cleavage of the C-15,
C-150 double bond, mainly under the action of the b-carotene15, 150 -oxygenase
15
15
-carotene-15,15-oxygenase
-carotene-15,15-peroxide
CHO
Retinal
Retinal reductase
CH2OH
Retinol
Figure 11. Conversion of b-carotene to Vitamin A retinoids.
enzyme (279, 280). The process of carotenoids cleavage to retinol (Figure 11)
is relatively inefficient and self-limiting because the conversion rate decreases
with increased body Vitamin A status. Originally only about 6% of all carotenoids,
those with b-retinylidene end groups, were considered as Vitamin A precursors
(provitamin A).
The metabolism and pharmacokinetics of the carotenoids are not well under-
stood, and their bioavailability is associated with much interindividual and intra-
individual variation (281). b-Carotene, a-carotene, cryptoxanthin, lycopene, and
lutein are the major carotenoids in human serum, and lycopene and b-carotene
are the major carotenoids in other human tissues (282). Small amounts of zeax-
anthin, phytofluene, and phytoene are also found in various organs. Various carote-
noids tend to be present in similar ratios in human plasma and tissues (282, 283).
The carotenoids are safe and even long-term intake of 180 mg of b-carotene per day
did not lead to hypervitaminosis. When large amounts of carotenoids are stored in
the adipose and other lipid-rich tissues, they may cause reversible yellowing of the
skin (cartenemia). Carotenemia is a benign condition, and Vitamin A poisoning

does not occur despite massive doses of carotene because the conversion of caro-
tene to Vitamin A is slow (284).
Recently, olive oil polyphenols have gained much more attention because of
their potential beneficial health effects (65, 285287). For example, olive oil was
found to improve lipid metabolism and antioxidant protection in rats fed cholesterol-
rich diets (288, 289). Hydroxytyrosol was especially effective in lowering the levels
of hydroperoxides, DNA damage, and mRNA levels of the antioxidant enzyme,
glutathione peroxidase (290).
The lignans of sesame oil, exemplified in sesamin, are especially interesting
from a physiological point of view. Yamashita et al. (291) published a note suggest-
ing that some antioxidants, other than tocopherols, in sesame exert some suppres-
sive effect on the manifestation of senescence in mice. Sesamin was then found to
lower cholesterol absorption as well as to inhibit its synthesis by downregulating
the activity of 3-hydroxy-3-methyl-glutaryl-CoA reductase, the key enzyme in
the biosynthetic pathway (292). Yamashita et al. (248) showed that sesamin caused
a significant increase in g-tocopherol levels in the plasma and liver, an effect that
was later confirmed (293, 294). It was recently found that sesamin competitively
inhibits certain cytochrome P450 3A isozymes that are involved in g-tocopherol
metabolism (295297). In addition, sesamin was found to inhibit 5-desaturation
of n-6 PUFA in rats resulting in accumulation of dihomo-g-linolenic acid at the
expense of arachidonic acid (298, 299), consequently decreasing the formation of
proinflammatory 2-series prostaglandins (300). This effect might be associated with
effects on sterol-binding proteins involved in fatty acid biosynthesis (301). The
kinetics of sesamin and episesamin metabolism has shown that they exhibit a
peak concentration value in plasma and most tissues by 6 hours after adminstration
and that they are cleared from the body by 24 hours (302). It is worthwhile studying
whether the reported physiological effects of sesame lignans can be translated into
measurable health benefits in humans.
3.3. Authentication of Vegetable Oils

Expensive oils, such as olive oil and cocoa butter, are sometimes mixed with other
cheaper oils to provide oils of reasonable price (303307). Methods for authentica-
tion of oil blends and characterization of constituent oils are, therefore, important
from an economical point of view (308). The few fatty acyl moieties involved as
major lipid components of most oils provide limited structural identity for most
vegetable oils especially when mixed together. Thus, lipid scientists were primarily
interested to use the knowledge about the minor lipid components as a way to
authenticate vegetable oils, to detect adulteration, and to identify the effects of
refining on their composition (309, 310). This mission can ideally be achieved
by general and specific authentication protocols taking into consideration the com-
bination of characteristics with respect to fatty acid composition, triacylglycerol
types, and unsaponifiable constituents. However, these theoretically possible proto-
cols are complex and costly in comparison with alternative protocols based on the
selection of a minimum number of markers. However, caution should always be

exercised as the minor markers of vegetable oils can still be removed by refining
and might not be detectable in the blends.
The analysis of sterols and sterol esters has been proposed as one way to identify
oils in blends (311, 312). Johansson and Croon (313) discussed the use of 4-des-
methyl-, 4-monomethyl-, and 4,4-dimethylsterols in characterizing different vege-
table oils, and the results are summarized in Table (8). The levels of total sterols and
sterol classes as well as the relative distribution of the individual sterol members
vary between oils. The presence of steradienes can also be used as a marker for
the presence of refined oils (314, 315). High oleic acid oils can easily be used to
adulterate olive oil. The presence of rapeseed oil in other oils can be detected by the
analysis of brassicasterol and its dehydration product, campestatriene. The presence
TABLE 8. The Use of Sterol Composition for the Authentication of Vegetable Oils
(% in Respective Fraction).
Vegetable Oil Desmethyl Sterols 4-Methyl Sterols 4,4-Dimethyl Sterols
Coconut Total 102 mg/100 g Total 7 mg/100 g Total 20 mg/100 g

campesterol (8%) obtusifoliol (9%) a-amyrin (7%)
stigmasterol (13%) cycloeucalenol (36%) b-amyrin (5%)
sitosterol (47%) citrostadienol (33%) cycloartenol (55%)
D5-avenasterol (26%) 24-methylenecycloartanol (22%)
Cottonseed Total 510 mg/100 g Total 12 mg/100 g Total 17 mg/100 g
campesterol (7%) obtusifoliol (8%) b-amyrin (7%)
sitosterol (86%) gramisterol (11%) cycloartenol (12%)
cycloeucalenol (5%) 24-methylenecycloartanol (21%)
citrostadienol (42%)
Linseed Total 471 mg/100 g Total 39 mg/100 g Total 246 mg/100 g
campesterol (27%) obtusifoliol (35%) cycloartenol (66%)
stigmasterol (8%) gramisterol (16%) 24-methylenecycloartanol (27%)
sitosterol (42%) cycloeucalenol (10%)
D5-avenasterol (13%) citrostadienol (11%)
Olive Total 150 mg/100 g Total 68 mg/100 g Total 292 mg/100 g
sitosterol (82%) cycloeucalenol (14%) cycloartenol (18%)
citrostadienol (22%) 24-methylenecycloartanol (31%)
cyclobranol (10%)
Palm kernel Total 140 mg/100 g Total 3 mg/100 g Total 22 mg/100 g

campesterol (10%) obtusifoliol (9%) cycloartenol (80%)
stigmasterol (13%) cycloeucalenol (33%) 24-methylenecycloartanol (18%)
sitosterol (69%) citrostadienol (19%)
5-avenasterol (7%)
D7-avenasterol absent
Peanut Total 321 mg/100 g Total 18 mg/100 g Total 17 mg/100 g
campesterol (10%) obtusifoliol (19%) b-amyrin (7%)
stigmasterol (7%) gramisterol (16%) b-amyrin (7%)
5-avenasterol (5%) citrostadienol (23%) 24-methylenecycloartanol (35%)
TABLE 8 Continued
Vegetable Oil Desmethyl Sterols 4-Methyl Sterols 4,4-Dimethyl Sterols
Rapeseed Total 954 mg/100 g Total 7 mg/100 g Total 18 mg/100 g

brassicasterol (10%) obtusifoliol (26%) cycloartenol (49%)
campesterol (33%) gramisterol (21%) 24-methylenecycloartanol (37%)
sitosterol (48%) cycloeucalenol (17%)
citrostadienol (16%)
Sesame Total 331 mg/100 g Total 47 mg/100 g Total 20 mg/100 g

stigmasterol (6%) gramisterol (15%) cycloartenol (51%)
sitosterol (57%) cycloeucalenol (12%) 24-methylenecycloartanol (16%)
5-avenasterol (6%) citrostadienol (15%) cyclobranol (5%)
Soybean Total 394 mg/100 g Total 25 mg/100 g Total 40 mg/100 g

Sunflower Total 494 mg/100 g Total 78 mg/100 g Total 33 mg/100 g
sitosterol (59%) citrostadienol (38%) cycloartenol (19%)
5-avenasterol (8%) 24-methylene-
cycloartanol (48%)
7-stigmastenol (6%)
7-avenasterol (5%)
Maize Total 1441 mg/100 g Total 62 mg/100 g Total 54 mg/100 g

5-avenasterol (10%) citrostadienol (29%) 24-methylenecycloartanol (40%)
Wheat germ Total 1425 mg/100 g Total 59 mg/100 g Total 59 mg/100 g

sitosterol (60%) gramisterol (25%) b-amyrin (12%)
5-avenasterol (7%) cycloeucalenol (6%) cycloartenol (25%)
Rice bran Total 1055 mg/Kg 3 main unknown cycloartenol and 24-methylene-
campesterol (24%) compounds in cycloartanol present as c-oryzanol
stigmasterol (11%) addition to obtusifoliol (see text).
sitosterol (52%) and gramisterol.
5-avenasterol (8%)
of high oleic sunflower oil can be detected by analyzing 8(14) and 14-sterols,
which result from the isomerization of 7-stigmasterol. Hazelnut oil in olive oil can
be detected by analyzing its biomarkers: filbertone (E-5-methylhept-2-en-4-one)
and d-tocopherol (306, 316, 317). Sesame oil can be detected in blends by the Bau-
duin test of its characteristic phenol, sesamol (318).
Squalene was suggested as a biomarker for olive oil, but its use as such was dif-
ficult because of the high variability for its occurrence in the oil. Moreover, squa-
lene isolated from rich oil sources can be added to oils to make them more olive-
like. For this reason, it was a mandate to add anthranilic acid as a marker for iso-
lated squalene (306). A combination of carotenoid and chlorophyll composition
may also be used for the authentication of virgin olive oil. The presence of
carotenoid and chlorophyll degradation products can be used as an index for
virgin olive oil quality. Virgin oil is characterized by a ratio of chlorophyll/
carotenoid 1 and a ratio of minor carotenoids/lutein 0.5 (319). The fact that
markers might also be present in the adulterant, although in a different ratio, neces-
sitate the use of sophisticated mathematics, e.g., pattern recognition analysis (320).
4. ANALYSIS OF MINOR LIPID COMPONENTS
The principle for the saponification of fats and oils is based on the reaction of the
fatty acid moieties of the oil by boiling under reflux with an ethanolic potassium
hydroxide solution. Saponification transforms the glyceridic compounds into polar
soaps making possible the extraction of the unsaponifiable matter with hexane or
diethyl ether. The solvent is evaporated, and then the residue is dried and weighed.
Schwartz (3) reported that dry homogenization of oils and alkali in a mortar pro-
vides an improved method for obtaining and quantifying the unsaponifiable matter
of fats and oils. Saponification is not applicable for glyceridic lipids, such as wax
esters, sterol esters, minor glyceridic compounds, and phospholipids, and it is not
suitable for the analysis of alkali-sensitive phenolic compounds and pigments. His-
torically, the content of sterol in vegetable oils was determined gravimetrically after
precipitation of its complexes with digitonin (321).
Gas chromatographic analyses for total aliphatic alcohols, sterols, and tocopher-
ols in vegetable oils are usually performed, with or without silylation, using
columns with low-polarity or polar stationary phases (322) and flame ionization
detectors. Silylation is often preferred because it provides sharp peaks. Analyses of
compound classes are sometimes performed separately after fractionation of the
respective classes by thin-layer or column chromatography (305). This fractionation
is, however, often not necessary, and many unsaponifiable constituents can be quanti-
fied directly by gas chromatography (GC) of the unsaponifiable fraction after sily-
lation (323, 324). Thus, a gas chromatogram of the unsaponifiable fractions from
different oils can present a fingerprint for oil identity. Besides gas chromatography
and gas chromatographymass spectrometry, methods such as Fourier Transform
Raman spectra can be used for the interpretation of these differences (325).
When sterol analysis is performed after saponification of ester bonds, a total
value is obtained for the level of each sterol by GC analysis. Free sterols can be
separated from sterol esters before saponification by means of thin-layer chromato-
graphy (326) or solid phase extraction (327330) using hexane to elute the sterol
esters and diethyl ether or a mixture of diethyl ether and hexane to elute the sterols.
Once separated, sterol esters need be saponified to release sterols before gas
REFERENCES 347
chromatography. Sterols and other compounds containing hydroxy or phenolic

functional groups are generally converted to silyl, or sometimes acetyl, derivatives
to increase their voltility to obtain sharp chromatographic peaks.
Tocopherols and tocotrienols can be analyzed by GC as well as by high-perfor-
mance liquid chromatography (331). For GC analysis, the elimination of interfering
substances, mainly acyl lipids, by saponification is necessary. This preparation step
is not needed in HPLC analysis where diluted oil samples can be injected directly
into normal- or reversed-phase columns. Normal-column HPLC is preferred because it
is able to separate the b- and g-tocopherol and tocotrienol isomeric pairs, not
separable by reversed-phase HPLC, and because it operates with organic solvents
and tolerates high loads of lipids that are easily washed-out by nonpolar solvents.
Recent developments in the manufacture of normal-phase columns enables repre-
ducible separations (331). Tocopherol/tocotrienol peaks are detected with ultravio-
let (290 nm) and evaporative light scattering detectors, but fluorescence and
electrochemical detectors provide higher sensitivity and specificity (331, 332).
Carotenoids can be coanalyzed with tocopherols and other fat-soluble vitamins
by HPLC using normal-phase (333), or reversed-phase C18 (334) or C30 columns
(335) using multiple-chanel Ultraviolet or diode array detectors to monitor carote-
noids and tocopherols at 295 nm and 450 nm, respectively. Normal-phase chroma-
tography provides the advantage of eliminating the need for saponification to
remove nonpolar lipids and ensures long column life. Reversed-phase chromato-
graphy columns are affected by the lipid load but have the advantage of providing
excellent separation for different carotenoids and their isomers. The C30 reversed-
phase columns are currently several times more expensive than C18 and normal-
phase columns. The usual methods for determination of chlorophylls in vegetable
oils are absorption spectrophotometry, fluorometry, and liquid chromatography
using fluorescence detectors (336). Chlorophyll pigmants can also be separated
and analyzed by HPLC using Ultraviolet-visible detection (337, 338).
Pre-extraction using normal solid-phase extraction columns to remove the neu-
tral lipids by nonpolar solvents followed by extraction with polar solvents (metha-
nol, acetonitrile, etc.) is suitable for the isolation of phenols, sterols, polycyclic
aromatic hydrocarbons, and chlorophylls from vegetable oils (52, 339341). The
extracts can then be analyzed by an HPLC-diode array detector to separate indivi-
dual compounds (342). Using this methodology, the main phenolic compounds in
virgin olive oil were identified as a dialdehydic form of elenolic acid linked to tyr-
osol and dialdehydic form of elenolic acid linked to hydroxytyrosol as well as
oleuropein aglycone. Detectors other than diode-array, e.g., electrochemical coular-
ray, can also be used (70, 343).
REFERENCES
1. S. Ghosh and D. K. Bhattacharyya, J. Amer. Oil Chem. Soc., 73, 12711274 (1996).
2. M. E. M. Almeida, E. C. Gusman, P. R. N. Carvalho, and O. Rusig, Arq. Biol. Technol.,
37, 10031011 (1994).
3. D. P. Schwartz, J. Amer. Oil Chem. Soc., 62, 246251 (1988).

4. B. S. Narasinga-Rao, Food Nutr. Bull., 22, 8186 (2001).
5. T. Gutfinger and A. Letan, J. Sci. Food. Agric., 25, 11431147 (1974).
6. A. Johansson, J. Amer. Oil Chem. Soc., 56, 886889 (1979).
7. A. Lanzon, T. Albi, A. Cert, and J. Gracian, J. Amer. Oil Chem. Soc., 71, 285291 (1994).
8. IUPAC-IUB Joint Commission on Biochemical Nomenclature (JCBN), Eur. J. Biochem.,
186, 429458 (1989). Available: http://www.chem.qmw.ac.uk/iupac/steroid.
11. T. M. Jeong, T. Itoh, T. Tamura, and T. Matsumoto, Lipids, 9, 921927 (1974).
12. V. K. S. Shukla, P. C. Dutta, and W. E. Artz, J. Amer. Oil Chem. Soc., 79, 965969 (2002).
13. M. F. Ramadan and J. T. Moersel, Eur. Food Res. Technol., 214, 202206 (2002).
14. T. Verleyen, Doctoral dissertation, University of Gent, Belgium, 2002.
15. T. Verleyen, M. Focardes, R. Verhe, K. Dewettink, A. Huyghebaert, and W. de Greyt,
16. E. J. Rogers, S. M. Rice, R. J. Nicolosi, D. R. Carpenter, C. A. McClelland, and L. J.
Romanczyk, J. Amer. Oil Chem. Soc., 70, 301307 (1993).
17. B. Kiosseoglou and D. Boskou, Oleagineux, 42, 169170 (1987).
18. S. P. Kochhar, Prog. Lipid Res., 22, 161188 (1983).
19. A. Serani and D. Piacenti, Riv. Ital. Sost. Gras., 69, 311315 (1992).
20. E. Tisconia and G. C. Bertini, Riv. Ital. Sost. Gras., 52, 101108 (1975).
21. J. Touche, M. Derbesy, J. Castang, M. Olle, and J. Estienne, Ann. Falsif. Expert. Chim.,
70, 263270 (1977).
22. J. Kesselmier, W. Eichenberger, and B. Urban, Plant Cell Physiol., 26, 463471 (1985).
23. A. Kamal-Eldin, K. Maatta, J. Toivo, A. M. Lampi, and V. Piironen, Lipids, 33, 1073
1077 (1998).
24. R. A. Ferrari, E. Schulte, W. Esteves, L. Bruhl, and K. D. Mukherjee, J. Amer. Oil Chem.
Soc., 73, 587592 (1996).
25. E. Homberg and B. Bielefeld, Fette Seifen Anstrichm., 84, 141146 (1982).
26. C. Crews, R. Calvet-Sarret, and P. Brereton, J. Chromatogr. A, 847, 179185 (1999).
27. I. M. Jawad, S. P. Kochhar, and B. J. F. Hudson, Lebensm. Wiss. u. Technol., 17, 155159
(1984).
28. A. Strocchi and G. Guerzoni, Riv. Ital. Dellla Sci. Aliment., 14, 125133 (1985).
29. A. Strocchi, Riv. Ital. Sost. Gras., 64, 401409 (1987).
30. A. Strocchi, Rev. Fr. Corps Gras., 4, 163168 (1988).
31. A. Strocchi, La Riv. Ital. Sci. Aliment., 17, 4150 (1988).
32. A. Strocchi and G. Marascio, Riv. Ital. Della Sost. Gras., 70 79 (1993).
33. A. Strocchi and G. Marascio, Fat Sci. Technol., 95, 293299 (1993).
34. S. Ramamurthi and A. R. Mccurdy, J. Amer. Oil Chem. Soc., 70, 287295 (1993).
35. I. Walsh, L. Winters, and R. Gonzalez, Inform, 9, 7883 (1998).
36. C. Watkins, Inform, 11, 580588 (2000).
37. I. Wester, Eur. J. Lipid Sci. Technol., 102, 37 44 (2000).
REFERENCES 349
38. V. Piironen, D. G. Lindsay, T. A. Miettinen, J. Toivo, and A. M. Lampi, J. Sci. Food.

Agric. 80, 939966 (2000).
39. A. Kamal-Eldin and L. A. Appelqvist, Lipids, 31, 671701 (1996).
40. G. F. Combs, Jr., The Vitamins: Fundamental Aspects in Nutrition & Health, Academic
Press Inc., San Diego, California, 1992, pp. 179203.
41. G. W. Burton and M. G. Traber, Ann. Rev. Nutr., 10, 357382 (1990).
42. A. T. Diplock, in A. T. Diplock, ed., Fat Soluble Vitamins, Technomic Pub., Lancaster,
Pennsylvania, 1985, pp. 154224.
43. D. Olejnik, M. Gogolwski, and M. Nogala-Kalucka, Nahrung, 41, 101104 (1997).
44. F. D. Goffman and C. Mollers, J. Agric. Food Chem., 48, 16051609 (2000).
45. S.-N. Ko, C.-J. Kim, C.-T. Kim, H. Kim, S.-H. Chung, S.-M. Lee, H.-H. Yoon, and I.-H.
Kim, Eur. J. Lipid Sci. Technol., 105, 225228 (2003).
46. A. Lampi, A., Kamal-Eldin, and V. Piironen, in M. Le Maquer and J. Shi, eds.,
Functional Foods, Vol. 2, CRC Press, Boca Raton, Florida, 2002, pp. 138.
47. S. L. Abidi, J. Amer. Oil Chem. Soc., 80, 327333 (2003).
48. L. Velasco and F. D. Goffman, Plant Syst. Evol., 221, 7788 (2000).
49. J. Sabate, H. E. T. Bell, and G. E. Fraser, in G. A. Spiller, ed., CRC Handbook of Lipids in
Human Nutrition, CRC Press, Boca Raton, Florida, 1996, pp. 145151.
50. O. J. Catchpole, J. C. vonKamp, and J. B. Grey, Ind. Eng. Chem. Res., 36, 43184324
(1997).
51. O. J. Catchpole and J. C. vonKamp, Ind. Eng. Chem. Res., 36, 37623768 (1997).
52. A. Cert, W. Moreda, and M. C. Perez-Camino, J. Chromatogr. A, 881, 131148
(2000).
53. M. W. Formo, E. Jungermann, F. A. Norris, and N. O. Sonntag, in D. Swern, ed., Baileys
Industrial Oil and Fat Products, Vol. 1, 4th ed., Wiley, New York, 1979, pp. 198.
54. R. Celades, Rev. Franc. Corps Gras, 44, 105112 (1993).
55. T. Kawakami, T. Kawai, N. Takei, T. Kise, S. Eda, and M. R. Urist, Clin. Orthopaed. &
Related Res., 337, 261266 (1997).
56. K. Grob, A. Artho, and C. Mariani, Fat Sci. Technol., 94, 394399 (1992).
57. R. Bortolomeazzi, P. Berno, L. Pizzale, and L. S. Conte, J. Agric. Food Chem., 49, 3278
3283 (2001).
58. H. Pfander, Methods Enzymol., 213, 313 (1992).
59. IUPAC-IUB Commission on Biochemical Nomenclature, Pure Appl. Chem., 41, 407
431 (1975).
60. N. L. Puspitasari-Nienaber, M. G. Ferruzzi, and S. J. Schwartz, J. Amer. Oil Chem. Soc.,
79, 633640 (2002).
61. B. Gandul-Rojas and M. I. Minguez-Mosquera, J. Sci. Food Agric., 72, 3139 (1996).
62. E. Psomiadou and M. Tsimidou, J. Sci. Food Agric., 81, 640647 (2001).
63. J. Velasco and C. Dobarganes, Eur. J. Lipid Sci. Technol., 104, 661676 (2002).
64. T. Matsuno, Fisheries Sci., 67, 771783 (2001).
65. C. Galli, A. Petroni, and F. Visioli, Eur. J. Pharm. Sci., 2, 6768 (1994).
66. D. Boskou, Olive Oil Chemistry and Technology, AOCS Press, Champaign, Illinois,
1996, pp. 5283.
67. G. F. Montedoro, M. Servili, M. Baldioli, and E. Miniati, J. Agric. Food Chem., 40, 1571
1576 (1992).
68. G. F. Montedoro, M. Servili, M. Baldioli, R. Selvaggini, E. Miniati, and A. Macchioni, J.
Agric. Food Chem., 41, 22282234 (1993).
69. M. Brenes, F. J. Hidalgo, A. Garca, J. J. Ros, P. Garca, R. Zamora, and A. Garrido, J.
Amer. Oil Chem. Soc., 77, 715720 (2000).
70. M. Brenes, A. Garca, P. Garca, and A. Garrido, J. Agric. Food Chem., 48, 51785183
(2000).
71. R. W. Owen, W. Mier, A. Giacosa, W. E. Hull, B. Spiegelhalder, and H. Bartsch, Food
Chem. Toxicol., 38, 647659 (2000).
72. M. Servili and G. F. Montedoro, Eur. J. Lipid Sci. Technol., 104, 602613 (2002).
73. L. Cinquanta, M. Esti, and E. La Notte, J. Amer. Oil Chem. Soc., 74, 12591264 (1997).
74. A. Kamal-Eldin and L. A. Appelqvist, J. Amer. Oil Chem. Soc., 71, 149156 (1994).
75. Y. Fukuda M. Nagata, T. Osawa, and M. Namiki, J. Amer. Oil Chem. Soc., 83, 10271031
(1986).
76. G. Henon, K. Recseg, and K. Kovari, J. Amer. Oil. Chem. Soc., 78, 401 410 (2001).
77. C. Mariani and S. Venturini, Riv. Ital. Sost. Gras., 79, 4957 (2002).
78. B. Reiter and E. Lorbeer, J. Amer. Oil Chem. Soc., 78, 881888 (2001).
79. A. A. Carelli, L. M. Frizzera, P. R. Forbito, and G. H. Crapiste, J. Amer. Oil Chem. Soc.,
79, 763768 (2002).
80. K. Ward, R. Scarth, J. K. Daun, and C. T. Thorsteinson, J. Amer. Oil Chem. Soc., 71,
13271331 (1994).
82. J. Cmolik and J. Pokorny, Eur. J. Lipid Sci. Technol., 102, 472486 (2000).
83. R. Subramanian, M. Nakajima, A. Yasui, H. Nabetani, T. Kimura, and T. Maekawa,
84. A. Koris and G. Vatai, Desalination, 148, 149153 (2002).
85. L. G. Pan, A. Campana, M. C. Tomas, and M. C. Anon, J. Amer. Oil Chem. Soc., 77,
12731276 (2000).
86. K. Clausen, Eur. J. Lipid Sci. Technol., 103, 333340 (2001).
87. W. R. Morrison, in P. J. Barnes, ed., Lipids in Cereal Technology, Academic Press,
London, U.K., 1983, pp. 1132.
88. V. L. Youngs, M. Pueskuelcue, and R. R. Smith, Cereal Chem., 54, 803812 (1977).
90. K. T. N. Balagopala, V. S. Ramjan, R. B. V. S. Koppeswara, C. P. Prasad, V. Penumarthy,
K. Vijay, R. K. N. Prasanna, R. Ongole, B. P. Satya, and R. T. Chandrasekhara, Eur. J.
Lipid Sci. Technol., 104, 203211 (2002).
91. S. Vichia, A. I. Castelloteb, L. Pizzalea, L. S. Contea, S. Buxaderasb, and E. Lopez-
Tamames, J. Chromatogr. A, 983, 1933 (2003).
92. P. Pfnuer, T. Matsui, W. Grosch, H. Guth, T. Hofmann, and P. Schieberle, J. Agric. Food
Chem., 47, 20442047 (1999).
93. S. H. Lee and K. J. Joo, Korean J. Food Sci. Technol., 30, 260265 (1998).
94. S. Nakamura, O. Nishimura, H. Masuda, and S. Mihara, J. Agric. Biol. Chem., 53, 1891
1899 (1989).
REFERENCES 351
95. A. A. Elsawy, M. M. Soliman, and H. M. Fadel, Grasas y Aceites, 39, 160162 (1988).
96. M. M. Soliman, A. A. Elsawy, H. M. Fadel, and F. Osman, Acta Aliment., 15, 251263
(1986).
97. Y. G. Lee, S. U. Lim, and J. O. Kim, J. Korean Agric. Chem. Soc., 36, 407415 (1993).
98. N. Frega, M. Mozzon, and G. Lercker, J. Amer. Oil Chem. Soc., 76, 325329 (1999).
99. F. Gutierrez, M. A. Albi, R. Palma, J. J. Ros, and J. M. Olas, J. Food Sci., 54, 6870
(1989).
100. M. H. Gordon and I. A. Rahman, J. Amer. Oil Chem. Soc., 68, 577579 (1991).
101. C. C. Ho and M. C. Chow, J. Amer. Oil. Chem. Soc., 77, 191199 (2000).
102. R. A. Tietz and R. W. Hartel, J. Amer. Oil Chem. Soc., 77, 763771 (2000).
103. R. J. Sims, J. A. Fioriti, and M. J. Kanuk, J. Amer. Oil Chem. Soc., 49, 298301 (1972).
104. G. Blekas, M. Tsimidou, and D. Boskou, Food Chem., 52, 289294 (1995).
105. M. Deiana, A. Rosa, C. F. Q. Cao, F. M. Pirisi, G. Bandino, and M. A. Dessi, J. Agric.
Food Chem., 50, 43424346 (2002).
106. J. Cillard and P. Cillard, J. Amer. Oil Chem. Soc., 57, 3942 (1980).
107. J. Cillard, P. Cillard, M. Cormier, and L. Girre, J. Amer. Oil Chem. Soc., 57, 252255
(1980).
108. J. Cillard, P. Cillard, and M. Cormier, J. Amer. Oil Chem. Soc., 57, 255261 (1980).
109. M. D. Fuster, A. M. Lampi, A. Hopia, and A. Kamal-Eldin, Lipids, 33, 715722 (1998).
110. A. M. Lampi, L. Kataja, A. Kamal-Eldin, and V. Piironen, J. Amer. Oil Chem. Soc., 76,
749755 (1999).
111. M. Makinen, A. Kamal-Eldin, A. M. Lampi, and A. Hopia, Eur. J. Lipid Sci. Technol.,
103, 286291 (2001).
112. A. Kamal-Eldin, M. Makinen, A. M. Lampi, and A. Hopia, Eur. Food Res. Technol., 214,
5257 (2002).
113. N. V. Yanishlieva, A. Kamal-Eldin, E. M. Marinova, and A. G. Toneva, Eur. J. Lipid Sci.
Technol., 104, 262270 (2002).
114. D. Boskou, Grasas y Aceites, 49, 326330 (1998).
115. J. L. Perrin, Rev. Fran. Corps Gras., 39, 2532 (1992).
116. F. Gutierrezrosales, J. Garridofernandez, L. Gallardoguerrero, B. Gandulrojas, and M. I.
Minguezmosquera, J. Amer. Oil Chem. Soc., 69, 866871 (1992).
117. M. Rahmani and A. S. Csallany, J. Amer. Oil Chem. Soc., 75, 837843 (1998).
118. B. H. Chen and J. H. Huang, Food Chem. 62, 299307 (1998).
119. O. T. Kasaikina, V. V. Vedutenko, A. M. Kashkay, V. D. Kortenska, M. I. Boneva, and
N. V. Yanishlieva, Oxid. Commun., 25, 232243 (2002).
120. E. Psomiadou and M. Tsimidou, J. Agric. Food Chem., 50, 722727 (2002).
121. B. Le Tutour, C. Brunel, and F. Quemeneur, New J. Chem., 20, 707721 (1996).
122. Y. Endo, R. Usuki, and T. Kaneda, Agric. Biol. Chem., 48, 985989 (1984).
123. Y. Endo, R. Usuki, and T. Kaneda, J. Amer. Oil Chem. Soc., 61, 771784 (1984).
124. R. Usuki, T. Suzuki, Y. Endo, and T. Kaneda, J. Amer. Oil Chem. Soc., 61, 785788
(1984).
125. C. L. Tautorus and N. H. Low, J. Amer. Oil Chem. Soc., 70, 843847 (1993).
126. C. L. Tautorus and N. H. Low, J. Amer. Oil Chem. Soc., 71, 11231128 (1994).
127. S. Ramamurthi and N. H. Low, J. Agric. Food Chem. 43, 14791483 (1995).
128. T. Pagano, M. Sergio, L. Glisenti, W. Diano, and M. A. Grompone, Ingenieria Qumica,
19, 1119 (2001).
129. P. Falaras, F. Lezou, G. Seiragakis, and D. Petrakis, Clays Clay Minerals, 48, 549556
(2000).
130. K. K. Reddy, R. Subramanian, T. Kawakatsu, and M. Nakajima, Eur. Food Res. Technol.,
213, 212218 (2001).
131. G. W. Burton and K. U. Ingold, Science, 224, 569573 (1984).
132. C. S. Foote, in W. A. Pryor, ed., Free Radicals in Biology, Vol. 2, Academic Press,
New York, 1976, pp. 85133.
133. C. S. Foote, in H. H. Wasserman, and R. W. Murray, eds., Singlet Oxygen, Academic
Press, New York, 1979, pp. 139171.
134. E. C. Lee and D. B. Min, J. Agric. Food Chem., 38, 16301634 (1990).
135. M. M. Jung and D. B. Min, J. Amer. Oil Chem. Soc., 68, 653658 (1991).
136. D. Bellus, in B. Ranby and J. F. Rabek, eds., Singlet Oxygen Reactions with Organic
Compounds and Polymers, Wiley, Chichester, U.K., 1978, pp. 61110.
137. J. Terao, R. Yamauchi, H. Murkami, and S. Matsushita, J. Food Process Preserv., 4, 79
93 (1980).
138. P. Di Masico, S. Kaiser, and H. Sies, Arch. Biochem. Biophys., 274, 532538
(1989).
139. N. I. Krinsky, in H. H. Wasserman and R. W. Murray, eds., Singlet Oxygen, Academic
Press, New York, 1979, pp. 597641.
140. N. I. Krinsky, Free Radical Biol. Med., 7, 617635 (1989).
141. M. M. Mathews-Roth, Photochem. Photobiol., 43, 9193 (1986).
142. M. M. Mathews-Roth, Fed. Proc., 46, 18901893 (1987).
143. P. Palozza and N. I. Krinsky, Arch. Biochem. Biophys., 297, 184187 (1992).
144. N. I. Krinsky and S. M. Deneke, J. Natl. Cancer Inst., 69, 205210 (1982).
145. J. Terao, Lipids, 24, 659661 (1989).
146. I. Jialal, E. P. Norkus, L. Cristol, and S. M. Grundy, Biochim. Biophys. Acta, 1086, 134
138 (1991).
147. T. A. Kennedy and D. C. Liebler, J. Biol. Chem., 267, 46584663 (1992).
148. B. P. Lim, A. Nagao, J. Terao, K. Tanaka, T. Suzuki, and K. Takama, Biochim. Biophys.
Acta, 1126, 178184 (1992).
149. D. C. Liebler, Ann. N. Y. Acad. Sci., 691, 2031 (1993).
150. S. A. Everet, S. C. Kundu, S. Maddix, and R. L. Willson, Biochem. Soc. Trans., 23, 230s
(1995).
151. F. J. Petracek and L. Zechmeister, J. Amer. Chem. Soc., 78, 14271434 (1956).
152. M. Elahi and E. R. Cole, Nature, 203, 186187 (1964).
153. A. H. El-Tinay and C. O. Chichester, J. Org. Chem., 35, 22902293 (1970).
154. W. Boguth, R. Patzelt, and R. Repges, Int. J. Vitam. Res., 41, 2132 (1971).
155. T. A. Kennedy and D. C. Liebler, Chem. Res. Toxicol., 4, 290295 (1991).
156. G. J. Handelman, F. J. G. M. Van Kuijk, A. Chatterjee, and N. I. Krinsky, Free Rad. Biol.
Med., 10, 427 437 (1991).
REFERENCES 353
157. R. C. Mordi, J. C. Walton, G. W. Burton, L. Hughes, K. U. Ingold, and D. A. Lindsay,

Tetrahedron Lett., 32, 42034206 (1991).
158. S. Q. Alam and B. S. Alam, J. Nutr., 113, 26082614 (1983).
159. L. Lomnitski, M. Bergman, Schon, and S. Grossman, Biochim. Biophys. Acta, 1082,
101107 (1991).
160. N. V. Yanishlieva, V. G. Raneva, and E. M. Marinova, Grasas y Aceites, 52, 1016
(2001).
161. T. Toyosaki, Int. J. Food Sci. Nutr., 53, 419423 (2002).
162. Z. L. Li, J. Cui, X. Y. Hong, Y. C. Liu, and X. L. Liu, J. Phys. Org. Chem., 15, 314318
(2002).
163. N. M. Bandarra, R. M. Campos, I. Batista, M. L. Nunes, and J. M. Empis, J. Amer. Oil
Chem. Soc., 76, 905913 (1999).
164. M. Servili, M. Baldioli, E. Miniati, and G. F. Montedoro, Riv. Ital. Sost. Gras., 73, 5559
(1996).
165. R. Aparicio, L. Roda, M. A. Albi, and F. Gutierrez, J. Agric. Food Chem., 47, 41504155
(1999).
166. M. A. Dessi, M. Deiana, B. W. Day, A. Rosa, S. Banni, and F. P. Corongiu, Eur. J. Lipid
Sci. Technol., 104, 506512 (2002).
167. M. T. Satue, S. W. Huang, and E. N. Frankel, J. Amer. Oil Chem. Soc., 72, 11311137
(1995).
168. M. Baldioli, M. Servili, G. Perretti, and G. F. Montedoro, J. Amer. Oil Chem. Soc., 73,
15891593 (1996).
169. M. Tsimidou, Ital. J. Food Sci., 10, 99116 (1998).
170. V. Fogliano, A. Ritieni, S. M. Monti, M. Gallo, D. Della Medaglia, M. L. Ambrosino, and
R. Sacchi, J. Sci. Food Agric., 79, 18031808 (1999).
171. E. Pagliarini, B. Zanoni, and G. Giovanelli, J. Agric. Food Chem., 48, 13451351 (2000).
172. M. D. Salvador, F. Aranda, and G. Fregapane, J. Amer. Oil Chem. Soc., 76, 427 432
(1999).
173. D. W. Peterson, Proc. Soc. Exp. Biol. Med., 78, 143147 (1951).
174. O. J. Pollak, Circulation, 6, 459460 (1952).
175. O. J. Pollak, Circulation, 7, 702706 (1953).
176. O. J. Pollak and D. Kritchevsky, in T. B. Clarkson, D. Kritchevsky, and O. J. Pollak, eds.,
Monographs on Atherosclerosis, Karger, Basle, Vol. 10, 1981, pp. 1219.
177. K. Uchida, H. Takese, Y. Nomura, K. Takeda, N. Takeuchi, and Y. Ishikawa, J. Lipid Res.,
25, 236244 (1984).
178. R. S. Tilvis, M. D. Tatu, and M. D. Miettinen, Amer. J. Clin. Nutr., 43, 9297 (1986).
179. L. Laraki, X. Pelletier, and G. Derby, Ann. Nutr. Metab., 35, 221225 (1991).
180. L. Laraki, X. Pelletier, J. Mourot, and G. Debry, Ann. Nutr. Metab., 37, 129133 (1993).
181. X. Pelletier, S. Belbraouet, D. Mirabel, F. Mordret, J. Perrin, X. Pages, and G. Debry,
Ann. Nutr. Metab., 39, 291295 (1995).
182. P. J. H. Jones, D. E. Macdougall, F. Ntatnios, and C. A. Vanstone, Can. J. Physiol.
Pharmacol., 75, 217227 (1997).
183. T. J. Howell, E. D. MacDougall, and P. J. H. Jones, J. Lipid Res., 39, 892900 (1998).
184. G. W. Meijer and J. A. Weststrate, Eur. J. Clin. Nutr., 52, 334343 (1998).
185. J. A. Weststrate and G. W. Meijer, Eur. J. Clin. Nutr. 52, 334343 (1998).
186. J. Plat, E. N. M. van Onselen, and R. P. Mensink, Eur. Heart J. Suppl. 1, s58s63 (1999).
187. C. H. Hennekens and J. M. Gaziano, Clin. Cardiol., 16, 110115 (1993).
188. N. Rong, L. M. Ausman, and R. J. Nicolosi, Lipids, 32, 303309 (1997).
189. L. K. Earl, P. Baldrick, and P. A. Hepburn, Int. J. Toxicol., 21, 353359 (2002).
190. M. N. Vissers, P. L. Zock, G. W. Meijer, and M. B. Katan, Amer. J. Clin. Nutr., 72, 1510
1515 (2000).
191. H. Gylling and T. A. Miettinen, Metabolism, 48, 575580 (1999).
192. T. E. Strandberg, R. S. Tilvis, and T. A. Miettinen, Lipids, 24, 705708 (1989).
193. H. M. Evans and K. S. Bishop, Science, 56, 650651 (1922).
194. P. B. McCay, Ann. Rev. Nutr., 5, 323340 (1985).
195. F. A. Oski, New Eng. J. Med., 303, 454455 (1980).
196. L. J. Machlin, Vitamin E: A comprehensive Trestise, Marcel Dekker, New York, 1980.
197. L. J. Machlin, in L. J. Machlin, ed., Handbook of Vitamins: Nutritional, Biochemical and
Clinical Aspects, Marcel Dekker, New York and Basel, 1984, pp. 99145.
198. G. W. Burton, A. Joyce, and K. U. Ingold, Lancet, 2, 327 (1982).
199. G. W. Burton, A. Joyce, and K. U. Ingold, Arch. Biochem. Biophys., 221, 281290
(1983).
200. G. W. Burton, K. H. Cheesman, D. Doba, K. U. Ingold, and T. F. Slater, in R. Porter and
J. Whelan, eds., Biology of Vitamin E, Ciba Foundation Symposium 101, Pitman,
London, U.K., 1983, pp. 418.
201. R. S. Parker, J. Nutr., 119, 101104 (1989).
202. H. Sies, Angew Chem., 98, 10611076 (1986).
203. H. Sies and M. E. Murphy, J. Photochem. Photobiol.: B. Biol., 8, 211224 (1991).
204. H. Sies, W. Stahl, and A. R. Sundqvist, in E. Sauberlich and L. J. Machlin, eds., Beyond
Difficiency, New York Acta, Ann. N. Y. Acad. Sci., 669, 720 (1992).
205. S. A. B. E. van Acker, L. M. H. Koymans, and A. Bast, Free Rad. Biol. Med., 15, 311328
(1993).
206. J. L. Hess, in R. G. Alscher, and J. L. Hess, eds., Antioxidants in Higher Plants, CRC
207. J. C. Bauernfeind and I. D. Desai, CRC Crit. Rev. Food Sci. Nutr., 8, 337382 (1977).
208. J. C. Bauernfeind, in L. J. Machlin, ed., Vitamin E: A Comprehensive Treatise, Marcel
Dekker, New York, 1980, pp. 99167.
209. P. J. McLaughlin and J. L. Weihrauch, J. Amer. Dietitic Assoc., 75, 647665 (1979).
210. K. F. Gey and P. Puska, Ann. N. Y. Acad. Sci., 570, 268286 (1989).
211. K. F Gey, P. Puska, P. Jordan, and U. K. Moser, Amer. J. Clin. Nutr., 53, 326s334s
(1991).
212. M. J. Stampfer, C. H. Hennekens, J. E. Manson, G. A. Colditz, B. Rosner, and W. C.
Willett, N. Engl. J. Med., 328, 14441449 (1993).
213. E. B. Rimm, M. J. Stampfer, A. Ascherio, E. Giovannucci, G. A. Colditz, and W. C.
Willett, N. Engl. J. Med., 328, 14501456 (1993).
214. P. Knekt, A. Reunanen, R. Jarvinen, R. Seppanen, M. Heliovaara, and A. Aroma, Am. J.
Epidemiol., 139, 11801189 (1994).
REFERENCES 355
215. M. Torun, N. Avci, and S. Yardim, J. Clin. Pharmacol. Ther., 20, 335340 (1995).
216. J. M. Gaziano, J. E. Manson, J. E. Buring, and C. H. Hennekens, in S. He and M. L. J.,
eds., Beyond Deficiency: New Views on The Function and Health Effects of Vitamins,
Ann. N. Y. Acad. Sci., 669, 249259 (1992).
217. H. Esterbauer, M. Dieber-Rotheneder, G. Striegl, and G. Waeg, Amer. J. Clin. Nutr., 53,
314s321s (1991).
218. I. Jialal and S. M. Grundy, J. Lipid Res., 33, 899906 (1992).
219. P. D. Reaven, A. Khouw, W. F. Beltz, S. Parthasarathy, and J. L. Witztum, Arterioscler.
Thromb., 13, 590600 (1993).
220. J. D. Belcher, J. Balla, G. Balla, D. R. Jacobs, M. Gross, H. S. Jacob, and G. M.
Vercellotti, Arterioscler. Thromb., 12, 17791789 (1993).
221. M. Steiner, Clin. Cardiol., 16, 116118 (1993).
222. A. Azzi, D. Boscoboinik, D. Marilley, N. K. Ozer, B. Stauble, and A. Tasinato, Amer. J.
Clin. Nutr., 62, 1337s1346s (1995).
223. S. Devaraj, D. Li, and I. Jialal, J. Clin. Invest., 98, 756763 (1996).
224. J. G. Bieri and R. P. Evarts, Amer. J. Clin. Nutr., 27, 980986 (1974).
225. M. G. Traber and H. J. Kayden, Amer. J. Clin. Nutr., 49, 517526 (1989).
226. M. G. Traber, G. W. Burton, L. Hughes, K. U. Ingold, H. Hidaka, M. Malloy, J. Kane, J.
Hyams, and H. J. Kayden, J. Lipid Res., 33, 11711182 (1992).
227. H. J. Kayden and M. G. Traber, J. Lipid Res., 34, 343358 (1993).
228. Q. Jiang, S. Christen, M. K. Shigenaga, and B. Ames, Amer. J. Clin. Nutr., 74, 714722
(2001).
229. R. V. Cooney, A. A. Franke, P. J. Harwood, V. Hatch-Pigott, L. J. Custer, and L. J.
Mordan, Proc. Natl. Acad. Sci. USA, 90, 17711775 (1993).
230. R. V. Cooney, P. J. Harwood, A. A. Franke, K. Narala, A. K. Sundstrom, P. O. Berggren,
and L. J. Mordan, Free Radic. Biol. Med., 19, 259269 (1995).
231. S. Christen, A. A. Woodall, M. K. Shigenaga, P. T. Southwell-Keely, M. W. Duncan, and
B. N. Ames, Proc. Natl. Acad. Sci., 49, 32173222 (1997).
232. S. Christen, P. Gee, and B. N. Ames, Methods Enzymol., 269, 267278 (1996).
233. A. J. Brown, J. Nutr. Biochem., 7, 2939 (1996).
234. K. S. Williamson, S. P. Gabbita, S. Mou, M. West, Q. N. Pye, W. R. Markesbery, R. V.
Cooney, P. Grammas, U. Reimann-Philipp, R. A. Floyd, and K. Hensley, Nitric Oxide:
Biol. Chem., 6, 221227 (2002).
235. W. J. Wechter, D. Kantoci, E. D. Murray Jr., D. C. DAmico, and M. E. Jung, Proc. Natl.
Acad. Sci., 93, 60026007 (1996).
236. M. Ohrvall, S. Tengblad, and B. Vessby, Eur. J. Clin. Nutr., 48, 212218 (1994).
237. A. Kontush, T. Spranger, A. Reich, K. Baum, and U. Beisiegel, Atherosclerosis, 144,
117122 (1999).
238. M. Kristenson, B. Zieden, Z. Kucinskiene, L. Schafer-Elinder, B. Bergdahl, B. Elwing,
A. Abaravicius, L. Razinkoviene, H. Calkauskas, and A. G. Olsson, Brit. Med. J., 314,
629633 (1997).
239. T. Messner, U. B. Erkstam, I. B. Gustavsson, S. B. Nilsson, and B. Vessby, Int. J.
Circumpolar Health, 56, 2129 (1997).
240. S. R. Farwer, B. C. der Boer, E. Haddeman, G. A. Kivits, A. Wiersma, and B. H. Danse,

British J. Nutr., 72, 127145 (1994).
241. G. J. Handelman, W. L. Epstein, J. Peerson, D. Spiegelman, L. J. Machlin, and E. A.
Dratz, Amer. J. Clin. Nutr., 59, 10251032 (1994).
242. M. Clement and J. M. Bourre, Biochim. Biophys. Acta., 1334, 173181 (1997).
243. H. U. Melchert and E. Pabel, J. Amer. Oil Chem. Soc., 75, 213216 (1998).
244. M. Clement, L. Dinh, and J. M. Bourre, Biochim. Biophys. Acta., 1256, 175180 (1995).
245. W. A. Behrens and R. Madere, Nutr. Res., 3, 891897 (1983).
246. W. A. Behrens and R. Madere, J. Nutr., 117, 15621569 (1987).
247. I. Elmadfa, S. W. Kim, M. Reutlinger, and R. Siewert, Z. Ernahrungswiss, 28, 3648
(1989).
248. K. Yamashita, Y. Nohara, K. Katayama, and M. Namiki, J. Nutr., 122, 24402446 (1992).
249. K. Yamashita, Y. Iizuka, T. Imai, and M. Namiki, Lipids, 30, 10191028 (1995).
250. M. Sugano, K. Koba, and E. Tsuji, Anticancer Res., 19, 36513657 (1999).
251. A. A. Qureshi, W. C. Burger, D. M. Peterson, and C. E. Elson, J. Biol. Chem., 261, 544
550 (1986).
252. A. A. Qureshi, N. Qureshi, J. J. K. Wright, Z. Shen, G. Kramer, A. Gapor, Y. H. Chong,
G. Dewitt, A. S. H Ong, D. M. Peterson, and B. A. Bradlow, Amer. J. Clin. Nutr., 53,
s1021s1026 (1991).
253. M. N. Gould J. D. Haag, W. S. Kennan, M. A. Tanner, and C. E. Elson, Amer. J. Clin.
Nutr., 53, s1068s1070 (1991).
254. A. A. Qureshi, S. A. Sami, W. A. Salser, and F. A. Khan, J. Nutr. Biochem., 12, 318329
(2001).
255. A. A. Qureshi, S. A. Sami, W. A. Salser, and F. A. Khan, Atherosclerosis, 161, 199207
(2002).
256. D. T. Tan, H. T. Khor, W. H. S. Low, A. Ali, and A. Gapor, Amer. J. Clin. Nutr., 53, s1027
s1030 (1991).
257. D. Raederstorff, V. Elste, C. Aebischer, and P. Weber, Ann. Nutr. Metabol., 46, 1723
(2002).
258. R. A. Parker, B. C. Pearce, R. W. Clark, D. A. Gordon, and J. J. K. Wright, J. Biol. Chem.,
268, 1123011238 (1993).
259. H. T. Khor and T. T. Ng, Int. J. Food Sci. Nutr., 51, s3s11 (2000).
260. J. C. Bauernfeind, Carotenoids as Colorants and Vitamin A Precursors: Technical and
Nutritional Applications, Academic Press, New York, 1981.
261. K. L. Simpson, Proc. Nutr. Soc., 42, 717 (1983).
262. C. S. You, R. S. Parker, and J. E. Swanson, Asia Pacific J. Clin. Nutr., 11, s438s442
(2000).
263. A. Kassoff, et al., Arch. Ophthalmol. Chic., 119, 14171436 (2001).
264. A. Kassoff et al., Arch. Ophthalmol. Chic., 119, 14391452 (2001).
265. H. Esterbauer, G. Striegel, H. Puhl, S. Oberreither, M. Rothenneder, M. El-Saadani, and
G. Jurgens, Ann. N. Y. Acad. Sci., 570, 254267 (1989).
266. M. K. Biyani and D. S. Sheorey, J. Assoc. Physicians India., 42, 899903 (1994).
267. G. Mozsik, T. Javor, G. Toth, T. Zsoldos, and A. Tigyi, Acta Phys. Hung., 64, 315319
(1984).
REFERENCES 357
268. A. M. Y. Nomura, G. N. Stemmermann, L. K. Heilbrun, R. M. Salkeld, and J. P.

Vuilleumier, Cancer Res., 45, 23692372 (1985).
269. M. S. Menkes, G. W. Comstock, J. P. Vuilleumier, K. J. Helsing, A. A. Rider, and
R. Brookmeyer, N. Engl. J. Med., 315, 12501254 (1986).
270. K. F. Gey, G. B. Brubacher, and H. B. Stahelin, Amer. J. Clin. Nutr., 45, 13681377
(1987).
271. H. Nishino, J. Cell Biochem., 22, 231235 (1995).
272. A. Bendich, Ann. N. Y. Acad. Sci., 691, 6167 (1993).
273. R. H. Prabhala, H. S. Garewal, F. L. Meyskens, and R. R. Watson, Nutr. Res., 10, 1473
1486 (1990).
274. R. H. Prabhala, M. Braune, H. S. Garewal, and R. R. Watson, Ann. N. Y. Acad. Sci., 691,
262263 (1993).
275. R. G. Cutler, Proc. Natl. Acad. Sci. USA, 87, 76277631 (1984).
276. M. M. Mathews-Roth, Pure Appl. Chem., 63, 147156 (1991).
277. J. A. Olson and N. I. Krinsky, FASEB J., 9, 15471550 (1995).
278. K. J. Yeum and R. M. Russell, Ann. Rev. Nutr., 22, 483504 (2002).
279. W. Miki, Pure Appl. Chem., 63, 141146 (1991).
280. J. A. Olson, in M. E. Shils, J. A. Olson, and M. Shike, eds., Modern Nutrition in
Health and disease, 8th ed., Lea and Febiger, Philadelphia, Pennsylvania, 1993, pp. 287
307.
281. N. V. Dimitrov and D. E. Ullrey, in N. I. Krinsky, M. M. Mathews-Roth, and R. F., Taylor,
eds., Carotenoids: Chemistry and Biology, Plenum Press, New York, 1989, pp. 269277.
282. R. S. Parker, in J. E. Kinsella, ed., Advances in Food and Nutrition Research, Vol. 33,
Academic Press, New York, 1989, pp. 157232.
283. A. Bendich and J. A. Olson, FASEB J., 3, 19271932 (1989).
284. A. D. Lascari, Clin. Pediat., 20, 2529 (1981).
285. F. Visioli and C. Galli, J. Agric. Food Chem., 46, 42924296 (1998).
286. F. Visioli and C. Galli, Crit. Rev. Food Sci. Nut., 42, 209221 (2002).
287. F. Visioli, D. Caruso, C. Galli, S. Viappiani, G. Galli, and A. Sala, Biochem. Biophys.
Commun., 278, 797799 (2000).
288. S. Gorinstein, H. Leontowicz, A. Lojek, M. Leontowicz, M. Ciz, R. Krzeminski,
M. Gralak, J. Czerwinski, Z. Jastrzebski, S. Trakhtenberg, M. N. Grigelmo,
R. Soliva-Fortuny, and O. Martin-Belloso, J. Agric. Food Chem., 50, 61026108 (2002).
289. M. Fito, E. Gimeno, M. I. Covas, E. Miro, M. C. Lopez-Sabater, M. Farre, R. de la Torre,
and J. Marrugat, Lipids, 37, 245251 (2002).
290. J. L. Quiles, A. J. Farquharson, D. K. Simpson, I. Grant, and K. W. J. Wahle, British
J. Nutr., 88, 225234 (2002).
291. K. Yamashita, Y. Kawagoe, Y. Nohara, M. Namiki, T. Osawa, and S. Kawagishi, J. Japan.
Soc. Nutr. Food Sci., 43, 445 446 (1990).
292. N. Hirose, T. Inoue, K. Nishihara, M. Sugano, K. Akimoto, S. Shimizu, and H. Yamada,
J. Lipid Res., 32, 629638 (1991).
293. A. Kamal-Eldin, D. Pettersson, and L. A. Appelqvist, Lipids, 30, 499505 (1995).
294. A. Kamal-Eldin, J. Frank, A. Razdan, S. Tengblad, S. Basu, and B. Vessby, Lipids, 35,
427435 (2000).
295. R. S. Parker, T. J. Sontag, and J. E. Swanson, Biochem. Biophys. Res. Commun., 277,
531534 (2000).
296. T. J. Sontag and R. S. Parker, J. Biol. Chem., 277, 2529025296 (2002).
297. S. Ikeda, T. Tohyama, and K. Yamashita, J. Nutr., 132, 961966 (2002).
298. M. Sugano, T. Inoue, K. Koba, K. Yoshida, N. Hirose, Y. Shinmen, K. Akimoto, and
T. Amachi, Agric. Biol. Chem., 54, 26692673 (1990).
299. S. Shimizu, K. Akimoto, Y. Shinmen, H. Kawashima, M. Sugano, and H. Yamada,
Lipids, 26, 512516 (1991).
300. T. Utsunomiya, S. R. Chavali, W. W. Zhong, and R. A. Forse, Amer. J. Clin. Nutr., 72,
804808 (2000).
301. T. Ide, L. Ashakumary, Y. Takahashi, M. Kushiro, N. Fukuda, and M. Sugano, Biochim.
Biophys. Acta, 1534, 113 (2001).
302. R. Umeda-Sawada, M. Ogawa, and O. Igarashi, Lipids, 34, 633637 (1999).
303. U. Bracco, W. Rostagno, and R. H. Egli, Int. Chocolate Rev., 25, 4148 (1970).
304. V. V. S. Mani and G. Lakshminarayana, J. Oil Technol. Assoc. India, 5, 184193 (1976).
305. A. M. Abu-Hadeed and A. R. Kotb, J. Amer. Oil Chem. Soc., 65, 19221926 (1988).
306. D. Firestone, J. Assoc. Off. Anal. Chem. Intern., 84, 176180 (2001).
307. F. Ulberth and M. Buchgraber, Eur. J. Lipid Sci. Technol., 105, 3242 (2003).
308. J. B. Rossell, Fat Sci. Technol., 93, 526531 (1991).
309. M. H. El-Mallah, T. Murui, and S. El-Shami, Grasas y Aceites, 50, 715 (1999).
310. G. Morchio, A. Pellegrino, C. Mariani, and G. Bellan, Rivista Ital. Sostanze Grasse, 76,
115128 (1999).
311. M. H. Gordon and R. E. Griffith, Food Chem., 43, 7178 (1992).
312. K. Grob, A. M. Giuffre, U. Leuzzi, and B. Mincione, Fat Sci. Technol., 96, 286290
(1994b).
313. A. Johansson and L. B. Croon, Lipids, 16, 306314 (1981).
314. E. Schulte, Fat Sci. Technol., 96, 124128 (1994).
315. K. Grob, A. M. Giuffre, M. Biedermann, and M. Bronz, Fat Sci. Technol., 96, 341345
(1994).
316. C. Mariani, G. Bellan, G. Morchio, and A. Pellegrino, Riv. Ital. Sost. Gras., 76, 5967
(1999).
317. C. Mariani, G. Bellan, G. Morchio, and A. Pellegrino, Riv. Ital. Sost. Gras., 76, 297305
(1999).
318. R. K. Chevan, V. S. Ghole, and M. Dikshit, J. Food Sci. Technol. India, 31, 433 434
(1994).
319. B. Gandul-Rojas, M. R. L. Cepero, and M. I. Minguez-Mosquera, J. Amer. Oil Chem.
Soc., 77, 853858 (2000).
320. F. Ulberth and M. Buchgraber, Eur. J. Lipid Sci. Technol., 102, 687694 (2000).
321. M. Kates, Techniques in Lipidology, 2nd ed., Elsevier Publ. Comp., Amsterdam, The
Netherlands.
322. N. Frega, F. Bocci, and G. Lercker, J. Amer. Oil Chem. Soc., 69, 447 450 (1992).
323. N. Frega, F. Bocci, G. Giovannoni, and G. Lercker, Chromatographia, 36, 215217
(1993).
REFERENCES 359
324. J. Giacometti, Analyst, 126, 472475 (2001).

325. V. Baeten, P. Dardenne, and R. Aparicio, J. Agric. Food Chem., 49, 50985107 (2001).
326. A. Kamal-Eldin, L. A. Appelqvist, G. Yousif, and G. M. Iskander, J. Sci. Food Agric., 59,
327334 (1992).
327. P. Hostmann and A. Montag, Fette Seifen Anstrichm., 88, 262264 (1986).
328. A. C. Bello, J. Amer. Oil Chem. Soc., 75, 11201123 (1992).
329. B. Ham, B. Butler, and P. Thionville, LC/GC, 18, 11741179 (2000).
330. P. Breinholder, L. Mosca, and W. Lindner, J. Chromatogr., B777, 6782 (2002).
331. A. Kamal-Eldin, S. Gorgen, J. Pettersson, and A. M. Lampi, J. Chromatogr. A, 881, 217
227 (2000).
332. R. Yamauchi, H. Noro, M. Shimoyamada, and K. Kato, Lipids, 37, 515522 (2002).
333. M. F. Ramadan and J. T. Morsel, Eur. Food Res. Technol., 214, 521527 (2002).
334. E. Gimeno, E. Calero, A. Castellote, R. M. Lamuela-Raventos, M. C. de la Torre, and
M. C. Lopez-Sabater, J. Chromatogr. A, 881, 255259 (2000).
335. D. Darnoko, M. Cheryan, E. Moros, J. Jerrel, and E. G. Perkins, J. Liquid Chromatogr.
Rel. Technol., 23, 18731885 (2000).
336. Y. A. Tan, K. S. Low, and C. L. Chong, J. Sci. Food Agric., 66, 479484 (1994).
337. K. Ward, R. Scarth, J. K. Daun, and C. T. Thorsteinson, J. Amer. Oil Chem. Soc., 71, 931
934 (1994).
338. K. Ward, R. Scarth, J. K. Daun, and C. T. Thorsteinson, J. Amer. Oil Chem. Soc., 71, 811
815 (1994).
339. G. Lercker and M. T. Rodriguez-Estrada, J. Chromatogr. A, 881, 105129 (2000).
340. P. M. Panagiotopoulou and M. Tsimidou, Grasas y Aceites, 53, 8495 (2002).
341. A. Bendini, M. Bonoli, L. Cerretani, B. Biguzzi, G. Lercker, and T. G. Toschi, J.
Chromatogr. A, 985, 425433 (2003).
342. S. Gomez-Alonso, M. D. Salvador, and G. Fregapane, J. Agric. Food Chem., 50, 6812
6817 (2002).
343. C. Capannesi, I. Palchetti, M. Mascini, and A. Parenti, Food Chem., 71, 553562 (2000).
13
Lecithins
Bernard F. Szuhaj
Szuhaj & Associates LLC
Fort Wayne, Indiana
1. INTRODUCTION
Phospholipids are lipids containing a phosphoric acid residue; they are natures
principal surface-active agents. They are found in all living cells, whether of animal
or plant origin. In humans and in animals, the phospholipids are concentrated in the
vital organs, such as the brain, liver, and kidney; in vegetables, they are highest in
the seeds, nuts, and grains. As constituents of cell membranes, and active partici-
pants in metabolic processes, they are essential to life (18).
The commercial term lecithin is very general, and it describes a composition
of lipid constituents and surface-active compounds present in the product rather
than in the chemical entity: phosphatidylcholine (PC). Thus, in general usage,
lecithin refers to a complex, naturally occurring mixture of polar lipids obtained
by water-degumming crude vegetable oils and separating and drying the hydrated
gums (8). It is, however, the phospholipid portion of lecithin that is mainly respon-
sible for giving form and function to commercial lecithin (6).
Commercial lecithin is an important coproduct of edible oil processing because
of its dietary significance and multifaceted functionality in food systems and indus-
trial applications. Unless indicated otherwise, the term lecithin will denote the
commercial designation throughout the text in this chapter.
Lecithin has a long history of use in foods, dating back more than 60 years (9).
The 1930s brought widespread use of commercial solvent extraction techniques for
361
362 LECITHINS
vegetable oil production, and because degumming of crude vegetable oil became
necessary for shipping stability, a large supply of crude lecithin gums was pro-
duced. These gums were obtained in sufficient quantity to necessitate their becom-
ing an item of commerce (9). In the ensuing years, there was extensive research into
developing new lecithin applications, as well as product refinements and modifica-
tions. Lecithin ingredients are now recognized as valuable products that have both
nutritional value and commercial, i.e., food/feed/industrial utility (2, 5).
Two of the earliest edible applications of lecithin, viscosity reduction in choco-
late and confectionery products, and emulsification/antispatter properties in margar-
ine, still enjoy wide popularity and represent outlets for large volumes of lecithin
products. In addition, other early uses such as in bakery goods, pasta, textiles, insec-
ticides, and paints, among others, are still active today.
2. SOURCES OF PHOSPHOLIPIDS
2.1. Human/Animal Tissues

Almost all body cells contain phospholipids. The common animal phospholipids
are made of sphingomyelin, phosphatidylcholine (PC), phosphatidylethanolamine
(PE), phosphatidylserine (PS), phosphatidylinositol (PI), and other glycerol phos-
pholipids of complex fatty acid composition. PC, formerly referred to as lecithin,
PE, formerly referred to as cephalin, and PS are by far the most predominant
phospholipids from most animal sources. As constituents of cell walls and active
participants in metabolic processes, they appear to be essential to life (8).
The exact composition of human/animal phospholipids depends on the source
and the method of extraction and purification. The central nervous system espe-
cially has a high phospholipid content. The liver is the site for their biosynthesis,
and the lipids of the mitochondria, which are the regulators of cell metabolism and
energy production in the body, consist of up to 90% phospholipids (10).
A survey of improved fractionation and analytical methods for elucidating the
molecular species of these complex animal phospholipids, and the phospholipids
present in primary and processed foods, has been published by Kuksis (11). This
survey also includes information on the quantitative analysis of phospholipids,
peroxidation products of phospholipids, and the composition of selected animal
phospholipids.
Only egg yolk, milk, and brain have served as major animal sources of commer-
cial lecithin. In some instances, isolated and purified lecithins have been developed
for clinical nutritional uses. Weihrauch and Son (12) present a concise review of
phospholipid composition in various foods.
The following brief discussion covers phospholipids from animal sources that
have some commercial significance.
Egg phospholipids. At one time, eggs, which possess a relatively high phospho-
lipid content, served as a commercial source until soybean technology made it
uneconomical to produce. The phospholipids in eggs are mainly in the yolk, where
at least a portion of them are combined with protein and carbohydrates.
SOURCES OF PHOSPHOLIPIDS 363
TABLE 1. Composition of Soy and Egg Lecithin (14).
Polar Lipids Soy Egg
Phosphatidylcholine 2022 6872

Phosphatidylethanolamine 2123 1216
Phosphatidylinositol 1820 02
Phosphatidic acid 48
Sphingomyelin 24
Other phospholipids 15 10
Glycolipids 912
Compositional data on commercial egg products and various lipid extracts from
egg yolk have been compiled by Gornall and Kuksis (13), Kuksis (11), Schneider
(14), and Satirhos et al. (15). Tables 1 and 2 (14, 16) compare the composition of
soy and egg lecithins and their fatty acids, and Table 3 (1621) shows the distribu-
tion of phospholipid classes in egg yolk.
Besides phospholipid composition, the main difference between plant/legume
lecithin (e.g., soy) and lecithin in egg yolk is that the former has a higher unsatu-
rated fatty acid content and no cholesterol. Egg lecithin as a commercial ingredient,
with the exception of some medical feeding programs, is too expensive for routine
use in food (10). In some infant formulas, egg yolk lipids and egg lecithin are
used (22).
Milk phospholipids. Milk has a phospholipid content of about 0.035%, which is
associated with the fat by virtue of being part of a colloidal membrane that sur-
rounds the fat globule. Wittcoff (4), Morrison et al. (23), and Privett et al. (24)
have reported the results of TLC analysis of the polar lipids of various milk frac-
tions. Recently, Jensen (25) published extensive high performance liquid chroma-
tography (HPLC) results of the phospholipid composition of cows milk during
lactation, and typical data on phospholipid, sphingolipid, and glycosphingolipid
classes in bovine milk. The feeding regime of fatty acids in the diet affects the fatty
acid composition of the milk lipids to a certain extent. It might be expected that the
fatty acid composition of the phospholipids will be affected as well.
TABLE 2. Fatty Acid Composition (%) (14).
Type of Acid Soy Egg
Saturated
Palmitic 1518 2729
Stearic 36 1417
Unsaturated
Oleic 911 3538
Linoleic 5660 1518
Linolenic 69 01
Arachidonic 0 35
364 LECITHINS
TABLE 3. Distribution of Phospholipid Classes in Egg Yolk (16).
Analyses
a b c d e,f e,g
Phospholipid Classes
% Total lipid (w/w) 23

PC 6676 69.1 77.0 69 82.6 87.1
PE 1524 23.9 18.0 24 9.1 7.8
PS 1 2.7 Trace
CL 1 3.2 Trace
SPH 36 1.0 2.3 3 1.8 2.5
LPC LPE 36 2.5 3 6.5 2.5
Unidentified
a
Privett et al. (17).
b
Noble and Moore (18).
c
Cook and Martin (19).
d
Gornall and Kuksis (20).
e
Connelly and Kuksis (21).
f
Commercial sample.
g
Intralipid.
Skim milk and milk serum have the highest portion of polar lipids as percent of
the total lipids, and whole milk and cream have the least. Of the polar lipids, phos-
phatidylethanolamine constitutes the largest component, with phosphatidylcholine
and sphingomyelin (being present in about equal proportions) at a significantly
lower level (Table 4) (16, 23, 24, 26).
Brain phospholipids. The brain is a rich source of phospholipids, and together
with the spinal cord, it probably possesses the highest phospholipid content of
any of the organs. There are many different types of phospholipids in the central
nervous system. As they bypass the blood-brain barrier, adequate nutrition
(biosynthesis) of the nerve cells is assured with these substances. Special
TABLE 4. Distribution of Phospholipid Classes in Various Milk Fractions (weight %) (16, 24).
Whole Plastic Nonfat
Lipid Classes Milk Skim Skim Serum Cream Milka Serumb
% Total lipid 2.0 19.0 32.0 22.0 1.02.0

Ceramide monohexoside Trace 19.5 14.1 7.8 17.9 3.0
Ceramide dihexodide Trace 10.1 3.0
Phosphatidylethanolamine 36.4 27.5 45.1 32.4 25.3 30.0 26.9
Plasmalogen 1.0
Phosphatidylcholine 27.0 7.3 16.4 23.2 19.9 28.0 29.3
Plasmalogen 3.0
Sphingomyelin 29.0 18.2 14.5 26.9 21.6 19.0 31.3
Phosphatidylserine Trace 1.9 7.6 3.5 Trace 8.0 5.0
Phosphatidylinositol Trace Trace Trace 1.1 Trace 5.0 5.9
Ganglioside (unknown) 7.6 15.54 2.33 5.14 15.34 0.0 1.0
a
Morrison et al. (23).
b
Santha and Narayanan (26).
TABLE 5. Distribution of Phospholipid Classes in Brain of Different Animal Species

(weight %) (16).
Animal Species
a,b a,b b,c
Lipid Classes Human Bovine Bovine Humanc,d Bovinec,e Humanc,e Sheepf Ratg
% Total lipid
PC 21.8 18.4 32.4 33.2 48.2 47.6 37.3 36.8
Plasmalogen 0.9
PE 35.4 36.1 23.5 25.2 24.2 17.8 7.7 36.4
Plasmalogen 16.5
PS 18.8 18.0 11.0 10.7 6.7 9.3 9.2 11.8
PI 1.8 1.8 4.3 4.8 7.1 5.0 2.1 3.1
PA 1.1 1.7 0.9 0.3 1.3 1.2 2.6 1.2
CL 1.2 1.0 NDh 0.3 2.0 2.2
PG 0.9 NDh 0.4 0.6
LPC (LPE) 2.0 0.2 0.9 1.0 1.0 2.5
LbisPC 1.0 0.4 2.1 0.2 0.7 0.2
SPH 16.3 15.0 20.4 17.0 4.9 10.7 12.9 5.7
Unidentified
a
Siakotos et al. (28).
b
Myelin.
c
Siakotos et al. (27).
d
Edothelial cells.
e
Nuclei.
f
Scott et al. (29).
g
Wuthier (30).
h
ND not detected.
enzyme systems see to it that the most efficient functioning is accomplished at all
times (4).
The composition of brain phospholipids has been extensively investigated by
adsorption column- and thin-layer chromatography (TLC). Table 5 lists the major
classes of brain phospholipids from different animal species, as compiled by Kuksis
(16, 2730). At the end of the 1990s, the fear of mad cow disease (BSE) may have
addressed the purity criteria for the applications of brain phospholipids from cows.
Phospholipids in liver, kidney, muscle, and other tissues. Organ meats such as
liver, kidney, and muscles are a major source of dietary phospholipids. The reader
is referred to Kuksis (16) for the distribution of various phospholipid classes in the
liver, kidney, muscles (heart and skeletal), spleen, lung, blood cells, bile, and adi-
pose tissue of different animal species. Compositional data of fatty acids for these
tissues and fluids are also given.
In blood, phosphatidylcholine is quantitatively the most important phospholipid.
Sphingomyelin is present in varying amounts in perhaps all of the animal organs,
most of it in the soft organs, and to a lesser degree in skeletal muscles and eggs (4).
Total blood contains about 0.2% to 0.3% phospholipids. In plasma and serum, phos-
phatidylcholine predominates, whereas in corpuscles, phosphatidylethanolamine
and sphingomyelin constitute the bulk of phospholipids. Most workers have found
366 LECITHINS
that the phospholipid content is greater in red blood cells than in plasma, and it
constitutes 60% to 65% of the total lipids in these cells (4).
2.2. Soybean
Although the highest concentrations of phospholipids occur in animal products, i.e.,
meat, poultry, fish, eggs, and milk/cheese, the major commercial source is the soy-
bean, which contains 0.3% to 0.6%. Nevertheless, phospholipids from other vege-
table oilseeds, i.e., corn, cottonseed, linseed, peanut, rapeseed, safflower, and
sunflower, and plants have also been studied and used (5).
Standard-grade, commercial lecithin from the soybean is a complex mixture. It
comprises phospholipids, triglycerides, and minor amounts of other constituents
(i.e., phytoglycolipids, phytosterols, tocopherols, and fatty acids). The composition
and molecular arrangement of this heterogeneous mixture of compounds defines
a product that is low in apparent polarity and has a strong tendency to promote
water-in-oil (w/o)-type emulsions (31).
A wide range of data has been published showing the variability in the composi-
tion of phospholipids and fatty acids in soybean lecithin (Tables 6 and 7) (32).
Older data were often determined by qualitative TLC, whereas today 31P-NMR,
quantitative Li-Sc HPLC, and HPTLC methods have been developed.
Soy lecithin is a coproduct of oil processing. As a result, purification steps used
to produce quality oil may affect the lecithin components. Also, soybeans exposed
to frost damage, or subjected to prolonged storage, have reduced lecithin yields
(33). Phospholipases, which produce phosphatidic acid, are active during storage
and may reduce the yield of lecithin (34). During the maturation process, the major
phospholipids (PC, PE, and PI) increase, and others decrease or remain constant
(32).
A change in the relative concentration of any of these components, or an altera-
tion of their chemical structures, may cause some change in the physical or
chemical properties of commercial lecithins. Lecithins can exist as liquids, plastics,
TABLE 6. Components (%) of Soybean Lecithin (32).
Range of Composition
Component Low Intermediate High
Phosphatidylcholine 12.021.0 29.039.0 41.046.0

Phosphatidylethanolamine 8.09.5 20.026.3 31.034.0
Phosphatidylinositol 1.77.0 13.017.5 19.021.0
Phosphatidic acid 0.21.5 5.09.0 14.0
Phosphatidylserine 0.2 5.96.3
Lysophosphatidylcholine 1.5 8.5
Lysophosphatidylinositol 0.41.8
Lysophosphatidylserine 1.0
Lysophosphatidic acid 1.0
Phytoglycolipids 14.315.4 29.6
TABLE 7. Fatty Acids (%) of Soybean Lecithin (32).
Range of Composition
Fatty Acid Low Intermediate High
Myristic (C14:0) 0.31.9

Palmitic (C16:0) 11.718.9 21.526.7 42.7
Palmitoleic (C16:1) 7.08.6
Stearic (C18:0) 3.74.3 9.311.7
Oleic (C18:1) 6.89.8 17.025.1 39.4
Linoleic (C18:2) 17.120.0 37.040.0 55.060.8
Linolenic (C18:3) 1.6 4.06.2 9.2
Arachidic (C20:0) 1.42.3
or free-flowing solids. Their color, solvent solubility, surfactant properties, and

chemical reactivity all can be modified. These modifications, in turn, will alter
the functional properties of the lecithin in a given application (31).
The ensuing section is a brief review of phospholipid plant sources other than
those from the soybean, which have current or potential commercial applications.
Soybean lecithin then will be discussed in more detail later throughout the subse-
quent sections in this chapter.
2.3. Corn
Weeks and Walters (35) have found that 2.5% to 4.5% of the phosphorus in corn is
in the form of phospholipids, depending on the variety involved.
The first detailed analysis of commercial corn phospholipids was published by
Scholfield et al. (36). Unlike the phenomenal growth in demand for corn sweeteners
TABLE 8. Distribution (%) of Polar Lipids in Corn and Soybean Lecithin (32).
Polar Lipid Corn Soybean
Sterylglycoside ester 15.0 4.3

Monogalactosyldiglyceride 1.8 0.8
Digalactosyldiglyceride 3.7 3.0
Other glycolipids 9.8 6.4
N-Acyl phosphatidylethanolamine 2.6 2.2
N-Acyl lysophosphatidylethanolamine 3.7 10.4
Phosphatidylethanolamine 3.2 14.1
Phosphatidylglycerol 1.4 1.0
Phosphatidylcholine 30.4 33.0
Phosphatidylinositol 16.3 16.8
Phosphatidic acid 9.4 6.4
Phosphatidylserine 1.0 0.4
Lysophosphatidylethanolamine Trace 0.2
Lysophosphatidylcholine 1.7 0.9
368 LECITHINS
TABLE 9. Fatty Acid Composition (%) of Corn and Soybean Lecithin (32).
Composition
Fatty Acid Corn Soybean
Palmitic (C16:0) 17.7 17.4

Stearic (C18:0) 1.8 4.0
Oleic (C18:1) 25.3 17.7
Linoleic (C18:2) 54.2 54.0
Linolenic (C18:3) 1.0 6.8
and other products of the corn-refining industry, the commercial exploitation of the
coproduct corn lecithin has not taken place in large quantities. Tables 8 and 9 illus-
trate the distribution of polar lipids and fatty acids in corn lecithin compared with
those in soybean lecithin (32).
Similar compositions were noted for corn and soy PC and PI. Phosphatidic acid
and glycolipids represent a higher proportion of polar lipids in corn than in soybean
lecithin. Cherry (37) and Cherry and Kramer (32) also stated that the percentage of
minor components in corn, steryl-glycoside ester, and other glycolipids are more
than twice that found in soybean. The physical properties, particularly the emulsi-
fying properties of corn lecithin, differ from those of soybean lecithin because of
the higher proportion of glycolipids in the corn lecithin.
Both the glycolipids and the phospholipids of corn have lower percentages of
linolenic acid (18:3) and are more saturated than those in the soybean. In general,
crude corn and soybean lecithins are equal in linoleic acid (18:2) content, but lino-
leic acid in corn varies from 42% to 70% depending on the variety of corn. Phytic
acid, 88% of which is in the corn germ, is extracted as part of the lecithin fraction
(32, 37). Elimination of phytic acid in corn is desirable because it binds zinc,
magnesium, and calcium.
2.4. Cottonseed
The phospholipids in cottonseed are similar in many respects to those of soybeans,
with the exception of their lower level of linolenic and higher level of saturated
fatty acid content (38). Cherry and Kramer (32) compiled Table 10 to show the
composition of cottonseed lecithin.
Lecithin can be fractionated from cottonseed as phospholipids and glycolipids.
Cottonseed lecithin shows flavor and color deterioration when blended with other
vegetable oils. The saturated/unsaturated fatty acid ratio of cottonseed phospholi-
pids is approximately 1:2 (39). Palmitic acid constitutes 90% of the total saturated
fatty acids (36%), and linoleic acid is approximately 80% of the total unsaturated
fatty acids (64%). Gossypol binds to lecithin during oil extraction from glanded cot-
tonseed (approximately 9% in crude phospholipids). This economically negates its
TABLE 10. Composition (%) of Cottonseed Lecithin (32).
Component Extract Phosphatidylcholine Phosphatidylethanolamine Phosphatidylserine
Phospholipids 1.82.2 34.935.9 13.720.1 7.026.0

Fatty acids
Myristic (C14:0) 0.4 0.3 0.4 0.6
Palmitic (C16:0) 32.9 31.1 33.7 33.3
Palmitoleic (C16:1) 0.5 0.3 0.3 0.6
Stearic (C18:0) 2.7 2.8 2.2 0.3
Oleic (C18:1) 13.6 11.5 11.5 14.4
Linoleic (C18:2) 50.0 54.0 49.0 50.4
Total gossypol 9.13 2.34 22.43 19.90
Free gossypol 0.02 2.24 0.05 0.01
use as a commercial source. Cultivars of glandless or gossypol-free cottonseed may

have some potential for providing commercial edible lecithins (37).
The composition of cottonseed lecithin and the composition of the phospholipid
fraction from hexane-defatted glandless cottonseed oil are summarized in Tables 10
and 11 (32, 37, 4042).
As cottonseed lecithin contains only trace amounts of fatty acids with more than
two double bonds (linolenic acid), it is more stable to oxidation and rancidity than
soybean lecithin. Cottonseed phospholipids are relatively high in phosphatidylcho-
line, which could provide good emulsifying properties in foods (32, 37).
TABLE 11. Composition of Phospholipid Fraction from Glandless

Cottonseed Oil (37, 40, 41).
Phospholipida Composition (% of Total Phosphorus)
Origin 4.12
Lysophosphatidylcholine 2.56
Phosphatidylinositol 13.41
Phosphatidylserine 2.38
Phosphatidic acid 8.76
Phosphatidylcholine 23.16
Phosphatidylethanolamine 13.46
Phosphatidylglycerol 7.62
Lysophosphatidylserine NDb
Lysophosphatidylethanolamine ND
Unknown (sum: 6 TLC spots) 25.30
a
Water (24%) was added to hexane-extracted glandless cottonseed oil, the resulting
mixture stirred 30 minutes at 70 C, and centrifuged to separate the oil and
phospholipid-containing fraction. The phospholipids were separated by two-
dimensional thin-layer chromatography (TLC) on Silica gel-60 plates. Dimen-
sion I chloroform: methanol:7N NH4OH (65:30:1); Dimension II chloroform:
methanol:acetic acid:water (170:25:25:4). Quantitation of the phsopholipids was
according to El-Sebaiy et al. (42).
b
ND not detected.
370 LECITHINS
2.5. Rapeseed
Rewald (43) found approximately 20% phospholipids in rapeseed. Rapeseed
lecithin has been reviewed, and an extensive bibliography has been compiled
(44). Table 12 (44) shows the composition of rapeseed and soybean gums.
The major phospholipids present in rapeseed lecithin are phosphatidylcholine,
phosphatidylethanolamine, and phosphatidylinositol. The relative proportion of
these components does not differ significantly from that of soybean lecithin. In
lecithin from high erucic oils, the long-chain fatty acids (C20C22) are present
only in small amounts, and in low erucic oil lecithin (including canola), the fatty
acid composition is not markedly different from that of the parent oil, except for a
somewhat higher content of C16:0, leaving a slightly higher C18:1 and C18:3 fatty
acid level in the degummed oil (44).
Solvent-extracted rapeseed oil has been found to contain the highest level of
phosphorus. For this reason, it is common practice to degum solvent-extracted
oil or the mixed crude oil from pressing and subsequent solvent-extraction. As
the double-zero rapeseed varieties such as canola became available, the applications
of rapeseed lecithin have developed positively. Where at first rapeseed lecithin
was applied as an emulsifier and energy component in animal feed, the recent
concerns about GMO soybean varieties in some parts of the world have increased
the market value of the softseed lecithins for food applications (45). The pho-
spholipid composition is similar to soybean lecithin with variations due to crop
and processing conditions. The rapeseed phospholipid compositions in Table 12
have been confirmed by recent data, whereas the soybean lecithin composition in
TABLE 12. Composition of Rapeseed and Soybean Gums (%) (44).
Rapeseed Soybean
a b b b c
Components (9) (9) (10) (11) (8) (8)c
Water 24
Nonlipid 9
Triglycerides or neutral lipids 16 38.1 5.6 29 35
Phospholipids (total) 51
Phosphatidylcholine 22 16.2 24.6 20 21
Phosphatidylethanolamine 15 17.5 22.1 16 8
Phosphatidylinositol 18 7.6 14.7d 8 20
Lysophosphatidylcholine 19.4 1 0
Lysophosphatidylethanolamine 2.0
Sterol glycosides or glycolipids 9e 7.9 13.6 11 0
Unidentified 36f 10.7 15 11
Unaccounted for 5
a
Complete gum sample, including water.
b
Acetone precipitate from gum sample.
c
Dried gum sample.
d
Tentatively identified.
e
Tentatively: phytoglycolipids.
f
Tentatively: 16% acidic phospholipids (plus 20% unidentified).
the last column is not representative because of the low phosphatidylethanolamine

content.
Sosada (46) determined the optimum conditions for fractionation of rapeseed
lecithin with alcohols to improve purified lecithin yield and phosphatidylcholine
enrichment.
2.6. Sunflower and Peanut

Although sunflower lecithin currently is not used to any great extent, as sunflower
oil production is increased, the availability of lecithin from this oil may be a pos-
sibility (47). Because of its high phosphatidylcholine content, sunflower lecithin
can be used in foods and feedstuffs. Its use in the manufacture of foods and cos-
metics can be increased by refining and fractionation and/or modifications (48).
Sunflower lecithin has a mild taste and similar emulsifying properties as soybean
lecithin. The crude lecithin is pastier than soybean lecithin, because of waxes, but
exact adjustment of the oil and acid content in refined sunflower lecithin results in
an effective emulsifier with good handling properties. This makes sunflower
lecithin interesting for food manufacturers, particularly in Europe, the biggest
sunflower seed producing/processing continent.
The percentage of phospholipids in sunflower oil ranges from 0.02% to 1.5%,
with an average of around 0.75%. The composition of the phospholipids is
similar to soybean lecithin, with a tendency to higher phosphatidylcholine and
lower phosphatidylethanolamine ratios, which might be caused by crop varieties
and processing conditions. Hollo et al. (49) report positive results of physical
and enzymatic modification of sunflower lecithin for improving the emulsifying
properties.
Hilditch and Zaky (50) found that phospholipids in peanuts are considerably less
unsaturated than those in soybean and cottonseed. Rewald (51) has shown 35.7%
phosphatidylcholine and 64.3% phosphatidylethanolamine content in peanut phos-
pholipids.
2.7. Other Plants and Micro-Organisms

Table 13 shows the phospholipid composition of selected plant sources (32).
Parsons and Price (52) have published compositional data on the phospholipids
of barley grain based on thin-layer chromatography.
Other than in animal tissues, egg, and oilseeds, quantitative data on the phospho-
lipids in plants are meager because of the difficulty involved in their isolation. The
phospholipids in wheat are about 80% PC and 20% PE (4). Rye, barley, and other
grains, vegetables, and fruits all contain small amounts of phospholipids. Micro-
organisms, especially those that are acid-fast, and lower plants, also contain large
amounts of lipids, including phospholipids; these entities are of interest for clinical
research (4). A survey of microbiological sources of phospholipids has been pub-
lished by Ratledge (53).
372 LECITHINS
TABLE 13. Phospholipids (%) of Selected Plant Sources (32).
Phospholipids
Source Phosphatidylcholine Phosphatidylethanolamine Phosphatidylinositol
Rapeseed 16.2; 20.024.6 15.017.5; 22.1 7.68.0; 14.718.0

Sunflower seed 12.726.8; 42.264.2 9.929.4; 46.6 3.721.4; 24.036.6
Peanut seed 49.0 16.0 22.0
Cucurbit seed 55.874.9 10.518.7 13.717.2
Rice bran 20.423.1 17.820.2 5.86.6
Barley seed 44.344.4 7.68.8 1.11.3
Olive fruit 47.358.9 5.38.0 18.023.9
Avocado fruit 37.044.9 12.019.5 12.118.0
Palash seed 44.6 14.8 27.0
Jangli badam seed 30.0 23.0 40.6
Papaya seed 28.1 18.7 34.0
Coriander seed 44.0 29.3 23.1
Carrot seed 29.1 35.4 23.1
3. NOMENCLATURE, CLASSIFICATION, STRUCTURE

AND COMPOSITION, AND CHEMICAL/PHYSICAL PROPERTIES
3.1. Definition
According to Wittcoff (4), three distinct polymeric alcohols provide the basic con-
stituents for the various phospholipids. The first of these is glycerol, and the phos-
pholipids containing it are referred to as glycerophospholipids. Included herein,
in addition to PC, PE, and PS, are the acetalphospholipids or plasmalogens (in
body fluids, muscles, and egg), the lysophospholipids, and the phosphatidic acids.
The second polyhydric alcohol is the aminodihydroxy compound sphingosine,
which is the basis for not only sphingomyelin (in the brain and spinal cord), but
also for other glycolipids. All of these compounds based on sphingosine are also
referred to as sphingolipids. The third polyhydric alcohol is inositol, which is
included in PI.
Phospholipids also form complexes with proteins (e.g., vitellin in egg yolk, animal
and plant tissues, lipoproteins in blood serum, and milk), carbohydrates, glycosides,
alkaloids, minerals, enzymes, cholesterol, and other substances. Lysophospholipids
represent a special class of compounds resulting from the chemical or enzymatic
hydrolysis of phospholipids. The role of phospholipases in normal and pathological
conditions as well as in cell metabolism is of great biological significance (4).
For the elucidation, synthesis, chemical properties, physical chemistry, com-
position, and analytical determination of the various individual phospholipid struc-
tures in animal and plant sources, the reader is referred to Wittcoff (4). Schneider
(14) discusses the nomenclature used for phospholipids in more detail and provides
compositional data on commercial lecithins (Table 14).
Because of the commercial significance of soybean lecithin, this chapter will
focus primarily on the structure, composition, analytical determination, properties,
and applications of this product.
NOMENCLATURE, CLASSIFICATION, STRUCTURE AND COMPOSITION 373
TABLE 14. Composition of Commercial Lecithins (%) (on Oil-Free Basis) (14).
Lecithin Soy Corn Sunflower Rapeseed Egg Bovine Brain
Phosphatidylcholine 21 31 14 37 69 18
Phosphatidylethanolamine 22 3 24 29 24 36
Phosphatidylinositol 19 16 13 14 2
Phosphatidic acid 10 9 7 2
Phosphatidylserine 1 1 3 18
Sphingomyelin 1 15
Glycolipids 12 30 20
The U.S. Food Chemical Codex (54) defines lecithin as follows:
Food-grade lecithin is obtained from soybeans and other plant sources. It is a complex
mixture of acetone-insoluble phosphatides that consist chiefly of phosphatidylcholine,
phosphatidylethanolamine, and phosphatidylinositol, combined with various amounts
of other substances such as triglycerides, fatty acids and carbohydrates. Refined
grades of lecithin may contain any of these components in varying proportions and
combinations depending on the type of fractionation used. In its oil-free form, the
preponderance of triglycerides and fatty acids is removed and the product contains
90% or more of phosphatides, representing all or certain fractions of the total phospha-
tide complex. The consistency of both natural grades and refined grades of lecithin
may vary from plastic to fluid, depending on the free fatty acid and oil content, and
upon the presence or absence of other diluents. . .
3.2. Classification of Commercial Soybean Lecithin Products

The simplest method for modifying natural (crude) lecithin is the addition of a non-
reactive substance. Plastic lecithins are converted to fluid forms by adding 2% to
5% fatty acids and/or carriers such as soybean oil. If the additives react with the
lecithin to alter the chemical structure of one or more of the phospholipid compo-
nents, the resulting product is referred to as a chemically modified lecithin. Mod-
ification can also be achieved by subjecting lecithin to partial controlled enzymatic
hydrolysis. Finally, refined lecithin products can be obtained by fractionating the
various phospholipid components.
A method for classifying lecithin to include modified and refined forms has been
proposed by Cowell et al. (55). This classification distinguishes between natural
(crude) lecithins and those modified by either custom blending or chemical/enzy-
matic treatment, e.g., hydroxylation, hydrogenation, acetylation, or refining by
acetone or alcohol fractionation. These latter products reflect the state of the art
regarding the availability of the various lecithin products on the market and have
enhanced properties for specific uses. A listing of soybean lecithin classifications
follows (56).
374 LECITHINS
I. Crude commercial lecithin

A. Plastic
1. Unbleached
2. Single bleached
3. Double bleached
B. Fluid
1. Unbleached
2. Single bleached
3. Double bleached
II. Compounded
III. Chemically modified
IV. Refined
A. Deoiled
1. As is
2. Custom blended
B. Fractionated
1. Alcohol soluble
a. As is
b. Custom blended
2. Alcohol insoluble
a. As is
b. Custom blended
C. Purified phosphatides
Natural (crude) lecithins. Specifications as defined by the National Soybean

Processors Association (19861987) for natural (crude) lecithins is presented in
Table 15 (57). Specifications have also been published by the Food Chemicals
Codex (1996) (54).
Phospholipid content is specified in terms of acetone-insolubles (AI); product
clarity, and purity in terms of hexane-insolubles (HI). The lecithins are classified
as plastic or fluid in consistency, and they are further subdivided on the basis of
manufacturing procedure as natural color, bleached, or double bleached. Acidity
of phospholipids plus acidity of the carrier (i.e., the oil and fatty acids) is given
by the acid value (AV), i.e., milligram of potassium hydroxide required to neutralize
the acids in 1 g of the lecithin sample. Crude lecithin can be filtered for utmost
purity and clarity. Filtration removes hexane-insoluble (HI) matter. Such products
are in demand for encapsulated nutritional supplements, for pharmaceutical grades,
and for advanced technology industrial uses requiring a high level of purity.
Compounded lecithins. Compounded lecithins are blended products. Lecithin
combined with selected additives can exhibit modified properties and functional-
ities. Lecithin may have a synergistic action with some additives or, simply, be
TABLE 15. Soybean Lecithin Specifications (57).

Grade
Fluid Plastic
Fluid Fluid Double- Plastic Plastic Double-
Unbleached Bleached Bleached Natural Bleached Bleached
Analysis Lecithin Lecithin Lecithin Lecithin Lecithin Lecithin
Acetone insoluble, min. 62% 62% 62% 65% 65% 65%

Moisture, max.a 1% 1% 1% 1% 1% 1%
Hexane insoluble, max. 0.3% 0.3% 0.3% 0.3% 0.3% 0.3%
Acid value,
(mg KOH/g), max. 32 32 32 30 30 30
Color, Gardner, max.b 18 14 12 18 14 12
Viscosity, centipoise,
at 25 C, max.c 15,000 15,000 15,000
Penetration, max.d 22 mm 22 mm 22 mm
a
By Karl-Fischer Titration (AOCS Method Tb 264).
b
Undiluted basis.
c
By any appropriate conventional viscosimeter, or by AOCS Bubble Time Method Tq 1A-64, assuming density
to be unity. Fluid lecithin having a viscosity less than 7,500 centipoises may be considered a premium grade.
d
Using Precision cone 73525, Penetrometer 73510; sample conditioned 24 hours at 25 C.
compounded with ingredients for making it more compatible in a particular system.

Common additives include special oils, polysorbates, monoglycerides and modified
monoglycerides, lanolin derivatives, solvents, plasticizers, or other surfactants.
These are added either to the wet gums prior to drying or are blended with dry fluid
lecithin at elevated temperatures (7, 58).
Modified lecithins. Lecithins may be modified chemically, e.g., hydrogenation,
hydroxylation, acetylation, and by enzymatic hydrolysis, to produce products
with improved heat resistance, emulsifying properties, and increased dispersibility
in aqueous systems (7, 58, 59). One of the more important products is hydroxylated
lecithin, which is easily and quickly dispersed in water and, in many instances, has
fat-emulsifying properties superior to the natural product. Hydroxylated lecithin is
approved for food applications under Title 21 of the Code of Federal Regulations
172.814 (1998) (60).
Fractionated and oil-free lecithins. When crude lecithin is further refined by
various fractionation methods to selectively separate its components, acetone and
ethanol are the most common solvents used. Fractionation of crude lecithin yielding
phosphatidylcholine of greater than 90% purity is done commercially.
A commercial, nearly oil-free lecithin is prepared by acetone extraction of
natural lecithin, which removes all but 24% oil and free fatty acids. Then an
optional alcohol fractionation step can separate the oil-free lecithin into an alco-
hol-soluble lecithin enriched in phosphatidylcholine and an alcohol-insoluble frac-
tion enriched in phosphatidylinositol. The choline fraction is an excellent emulsifier
for oil-in-water (o/w) emulsions and the inositol fraction for water-in-oil (w/o)
emulsions.
376 LECITHINS
O O
CH2OCR1 CH2OCR1
O O
CHOCR2 CHOCR2
O O
CH2O P O CH2CH2N(CH3)3 CH2O P O CH2CH2NH3
O O
Phosphatidylcholine Phosphatidylethanolamine
O
CH2OCR1 R1 and R2 = C15C17
O
Hydrocarbon
CHOCR2 chains
O OH OH
CH2O P O
O HO
OH
OH
Phosphatidylinositol
Figure 1. Three principal components of soybean lecithin (7).
3.3. Structure of Phospholipids in Commercial Lecithins

Chemical structures for the most commonly occurring phospholipids in commercial
soybean lecithin are shown in Figure 1 (7). PC and PE are cationic and anionic at
the same time; that is, they are zwitterions, and thus they can have some buffering
action for both bases and acids. PI, however, is a relatively strong acid and, there-
fore, is anionic. The classes of compounds in commercial lecithin are as follows
(31):
Phospholipids Glycolipids
Anionic Steryl glucosides
Zwitterionic Esterified steryl
glucosides
Galactosyl glycerides
The reader is referred to Horrocks (61) for more specific discussion on the
nomenclature and structure of phospholipids.
3.4. Composition
Specification ranges, chemical and fatty acid compositions for commercial natural
lecithins, along with approximate compositional data for commercially refined
lecithin fractions are given in Tables 1619 (8, 6265), respectively.
TABLE 16. Specifications for Commercial Soybean Lecithin (62, 63).
Grade
Fluid Fluid Plastic Plastic
Natural Fluid Double- Natural Plastic Double-
Color Bleached Bleached Color Bleached Bleached
Analysis Lecithin Lecithin Lecithin Lecithin Lecithin Lecithin
Acetone insoluble, % min. 62 62 62 65 65 65

Moisture, % max.a 1 1 1 1 1 1
Benzene insolubles, % max. 0.3 0.3 0.3 0.3 0.3 0.3
Acid value, max. 32 32 32 30 30 30
Color, Gardner, max.b 10 7 4 10 7 4
Viscosity, poises,
at 25 C, max. 150 150 150
Penetration, max., in mmc 22 22 22
a
By toluene distillation for 2 hr or less.
b
As a 5% solution in colorless mineral oil.
c
By specified cone penetrometer test.
Soybean oil contains 1.53.0% phospholipids (71). Crude soybean lecithin has an
oil content of about 30%. PC is present at a level of about 16%., PE about 14%, and
inositol phospholipids about 12% (7). As can be seen in Table 18 (8), the fatty acid
compositions of soybean phospholipids are rich in polyunsaturated fatty acids. Mis-
cellaneous low-level constituents include water, phosphatidic acid, pigments, galac-
tosyl glycerides, various glycolipids, phosphatidylserine, carbohydrates, sterols,
and tocopherols. Phosphorus content of crude soybean oil extracted from flours
can vary depending on extraction temperature and flour moisture (72).
3.5. Chemical/Physical Properties

In practice, commercial lecithin products are not marketed by phospholipid content,
but rather by a set of unique chemical and physical properties. These properties, as
indicated by product specifications, must be understood because they are used to
characterize specific lecithin types.
TABLE 17. Approximate Chemical Composition of Natural Commercial

Soybean Lecithin (64).
Fraction %
Soybean oil 35
Phosphatidylcholine 16
Phosphatidylethanolamine 14
Phosphatidylinositol 10
Phytoglycolipids and other minor phosphatides and constituents 17
Carbohydrates 7
Moisture 1
378 LECITHINS
TABLE 18. Composition Fatty Acids of Soybean Lecithin (%) (8).
2022
Reference 14:0 14:1 16:0 16:1 18:0 18:1 18:2 18:3 20:0 Unsaturated
Hilditch and Zaky (50) 11.7 8.6 4.0 9.8 55.0 4.0 1.4 5.5
Rzhekhin et al.a (66) 18.1 3.7 22.4 40.0 5.0 2.3 6.2
Vijayalakshmi and
Raob (67) 42.7 7.0 11.7 17.0 20.0 1.6
Daga (68) 1.9 Trace 26.7 9.3 25.1 37.0
Daga (69) 0.3 1.2 25.5 10.3 39.4 17.1 6.2
Rydhag and Wiltonc (70) 21.5 4.3 7.2 60.9 6.1
Rydhag and Wiltond (70) 18.9 4.1 6.8 60.8 9.2
a
Also 2.3% unidentified.
b
CHCl3/CH3OH extraction.
c
Acetone-precipitated.
d
Granulated.
Commercial soybean lecithin, being a complex mixture of polar lipids, performs

as a wetting and emulsifying agent. Stanley (1) states that in heterogeneous systems
such as oil and water, the phospholipid molecules arrange themselves in monomo-
lecular layers with the fatty acid portion facing the oil surface and the phosphoric
acid portion facing the water surface. The arrangement lowers the interfacial
tension of the oilwater boundaries with resultant benefits such as rapid wetting,
lowering of viscosity, and better and more stable emulsions or dispersions.
Soybean lecithin is soluble in aliphatic and aromatic hydrocarbon solvents,
partially soluble in ethyl alcohol (principally the inositol fraction), and practically
insoluble in acetone (less than 0.003% weight/volume at 5 C) and in water (73).
When mixed with water, soybean lecithin hydrates to a thick emulsion that can
be thinned with water to almost any desired dilution. Acetone does dissolve readily
in lecithin and will form a thin, uniform imbibition as long as the quantity of
acetone is insufficient to precipitate the phospholipids. Lecithin is soluble in
TABLE 19. Approximate Composition of Commercially Refined Lecithin Fractions (%) (65).
Alcohol-Soluble Alcohol-Insoluble
Fraction Oil-Free Lecithin Lecithin Lecithin
Phosphatidylcholine 29 60 4
Cephalin 29 30 29
Inositol and other 32 2 55
phosphatides, including
glycolipids
Soybean oil 3 4 4
Other constituentsa 7 4 8
Emulsion type favored Either oil-in-water or
water-in-oil Oil-in-water Water-in-oil
a
Includes sucrose, raffinose, stachyose, and about 1% moisture.
mineral oils and fatty acids, and practically insoluble in cold vegetable and animal
oils, but it will dissolve in hot oils.
Soybean lecithin has a brown to light yellow color, depending on the conditions
used in its manufacture and the degree of bleaching.
Identification and characterization of phospholipids. For additional information
on various techniques and methods used in the identification and characterization of
phospholipids in general, the reader is referred to Kramer et al. (74).
Fractionation and purification of lecithin. Because of space limitations, it is not
possible to discuss fractionation and purification processes for all vegetable and ani-
mal lecithins in this chapter. The reader is referred to Schneider (14) who described
the fractionation and purification of various vegetable lecithins and those from egg
in considerable detail. Small-scale fractionation processes may include separation
of neutral oil and polar lipids (deoiling) including the use of acetone; the adsorption
of a hexane solution of lecithin on a silica column; separating neutral and polar
lipids from a hexane solution with the aid of membranes; treatment of lipid mix-
tures with supercritical gases or gas mixtures, e.g., carbon dioxide or propane
carbon dioxide; fractionation of neutral oil containing lecithins by solvent treat-
ment, e.g., aqueous methanol, ethanol, and propanol; fractionation of de-oiled
lecithins by solvent treatment, e.g., ethanol; solvent treatment after chemical mod-
ification, e.g., acylation prior to acetone de-oiling; precipitation methods, e.g., salt;
ultrafiltration methods; and many chromatographic processes, mainly for polar lipid
separation but also for separation focused on the degree of unsaturation. The com-
mercial manufacture of fractionated soybean lecithins will be covered later in this
chapter.
Synthesis and modification of phospholipids. For an excellent review of the
synthesis and modification of phospholipids, the reader is referred to Ghyczy
(75). According to the review, depending on the starting material used, there are
two ways to synthesize phospholipids. In the partial synthesis, phospholipids are
isolated from natural sources and the individual constituents, fatty acids, and
head groups are exchanged to obtain a certain phospholipid. In the total synthesis,
phospholipids are produced from fully synthesized, available, basic molecules that
were not obtained from phospholipids. Both methods are of importance today
because each manufacturing process has certain advantages with regard to definite
products and fields of application.
The partial synthesis may involve several synthetic steps, depending on the
basic phospholipid used, the enzyme, the final product desired, and the type and
position of the phospholipid constituents to be exchanged. For example, the partial
synthesis may avail of the reacylation of 3-sn-glycerophosphorylcholine (GPC).
Alternatively, by a deacylation step, GPC can be obtained from PC in soybean
lecithin (75).
Partial synthesis may include synthesizing PC with mixed fatty acids from GPC
as the starting material. Other types of phospholipids yield compounds, after
deacylation, which have certain functional groups, e.g., amino groups from PE.
PI can be manufactured by using the enzyme PL-D, using phospholipids from
soy lecithin (75).
380 LECITHINS
The most suitable starting materials, for the total synthesis of phospholipids, are
optically active derivatives from glycerol, called chiral C3 building blocks. In
addition to proper configuration, an early differentiation of the hydroxyl groups
is also necessary to shorten the process of synthesis (75).
Transesterification has also been investigated as a means for preparing polyun-
saturated phospholipids from soy phospholipids (76).
Specifications for soybean lecithin. The following methods are routinely used for
determining whether the specifications for given products are met:
AI. The amount of acetone-insoluble matter (%AI) is a measure of the polar

material found in lecithin. In soybean lecithin, the acetone-insolubles typi-
cally contain 7075% phospholipids, with the remaining portion consisting of
glycolipids, carbohydrates, and a small amount of residual triglyceride oil.
The amount of acetone-insoluble matter is determined by the AOCS Official
Method Ja-4-46 (77).
AV. The AV is the number of milligrams of potassium hydroxide necessary to
neutralize the acids in 1 g of lecithin (62). A products AV is representative of
the acidity contributed by both the phospholipids and any free fatty acids that
are present. The AV is usually not indicative of pH, as the chemical nature of
the phospholipid imparts buffering qualities to most systems. Lecithins
typically exhibit a neutral pH value in aqueous media. An AV above 36
may indicate degradation of the lecithin because of improper processing or
substandard quality soybeans. AV should not be confused with free fatty
acid content, pH, or mineral acids. The correct method to assay for free
fatty acids is to titrate only the acetone-soluble portion of the lecithin, whereby
any contribution from the phospholipids in the acetone-insoluble portion is
eliminated. AV is determined by the AOCS Official Method Ja 6-55 (77).
Moisture. The water content of lecithin products is usually less than 1.0%. As a
consequence of lecithins essentially moisture-free state, lecithin products
have very low water activity and do not adversely contribute to the micro-
biological profile of most food systems. Most lecithin products are preserved
well in storage. Higher moisture levels usually indicate a greater potential for
spoilage or chemical degradation. Moisture is determined by AOCS Official
Method Ja 2b-87 (77). A less accurate moisture level can also be determined
by azeotropic toluene distillation (AOCS Official Method Ja 2-46) (77). One
cannot determine lecithin moisture by vacuum oven methods. These methods
are known to degrade lecithin products and yield false moisture levels.
HI. The level of HI matter is one measure of the purity of lecithin products. HI
matter usually consists of residual fiber, but also particulate contaminants that
may be introduced during processing (e.g., filter aids). The level of HI matter
in crude lecithin should never exceed 0.3% and rarely exceeds 0.1%. HI
matter in lecithin is detrimental to clarity and use in specific applications. HI
is measured by an official Food Chemicals Codex (FCC) (1996) method (54)
or by AOCS Official Method Ja 3-87 (77).
Color. Commercial liquid lecithins vary in color from light honey to dark brown
(62). De-oiled lecithins are typically a shade of yellow. Historically, lecithins
have been color graded as unbleached, single-bleached, and double-bleached.
The color of lecithin is commonly determined with the use of a Gardner-
Hellige Varnish Comparator, or simply Gardner Liquid Color Standards. The
color of various lecithin products is generally in the range of Gardner 918 in
an undiluted form (AOCS Official Method Ja 9-87) (77).
Other physical/chemical properties and quality criteria.
Consistency. Lecithins are available in both fluid and plastic (solid) forms. Fluid
lecithins generally follow Newtonian flow characteristics. The viscosity
profile of lecithins is a complex function of acetone-insoluble content,
moisture, mineral content, acid value, and the combined effects of assorted
additives such as vegetable oils and surfactants. Generally, higher AI and/or
moisture content yields higher viscosity, whereas an increased AV often
decreases viscosity. Certain divalent minerals, such as calcium and others,
can also adjust the viscosity level.
Clarity. In some soy processing plants, high levels of HI may partition with the
lecithin gums on separation from the oil. This lipid-insoluble material can
cause haziness in fluid lecithins. With modern miscella and oil filtration
techniques, lecithins with very low HI contents can be produced. Conse-
quently, modern lecithins are clear. Additionally, moisture can also contribute
to lack of clarity. Generally, moisture levels over 1% can cause haziness.
Besides being an aesthetic problem, if haziness is caused by HI material, it
can result in sediment over time; solid particles may appear on the bottom of
an otherwise clear liquid product containing lecithin.
For a more detailed review of industrial methods of analysis, the reader is

referred to Lantz (62). A review of traditional and novel approaches to the analysis
of plant phospholipids has been prepared by Marmer (78). Ackman (79) has
reviewed the early developments and practical applications of GLC analysis.
Modern phospholipid analysis is typically accomplished by high-performance
liquid chromatography (HPLC) methods (80) or by nuclear magnetic resonance
(NMR) techniques (81).
Chemistry and reactivity of phospholipids. The chemistry of the phospholipids is
generally that of their ester linkages, unsaturated fatty acids, and other reactive
groups. Most of the applicable reactions of organic chemistry have been employed
in their study (82). Baer and Kates (83), Brockerhoff (84), Hanahan (85), Pryde
(86), Scholfield (8, 82), Strickland (87), Verheij (88), and Wittcoff (4) provide
major reviews of phospholipid chemistry and reactivity under various conditions.
The latter covers hydrolysis, hydrogenolysis, acetolysis, hydroxylation, thermal
decomposition, hydrogenation, autoxidation, browning reaction, and other reactions
(e.g., bromination and complexing with various substances).
Lecithin interaction with other food ingredients. Food systems are usually het-
erogeneous mixes of components, in which the interaction of ingredient classes
382 LECITHINS
TABLE 20. Alteration of Lecithin Form/Function (31).
Action Technique Utility
Decrease viscosity Add special diluents Sprayable

Easier handling
Increase hydrophilic properties Solvent fractionation o/w emulsifiers
Chemical modification Wetting agents
Enzyme modification
Compounding
Reduce color Process controls Light-colored foods
Oxidative bleaching
Convert to powder form De-oil Dry blendable
Mix with carrier
(e.g., proteins, starches, fats, surfactants) can be important to finished product qual-
ity, shelf life, and nutritional value.
The most common modifications of lecithin and the intended physical/functional
alterations are shown in Table 20 (31). The range of physical/functional properties
available in commercial lecithins is listed in Table 21 (31). These changes in
lecithin allow for the basic lecithin obtained from soybean oil to be converted to
various emulsifier products having a wide variety of food, feed, and industrial appli-
cations. Reviews describing chemical reactions for phospholipid modifications
intended to obtain specific functionalities include those of Eichberg (89), Hawthorn
and Kemp (90), Kuksis (91), Pryde (86), Snyder (92), Strickland (87), and Van Dee-
nen and DeHaas (93).
Model studies have given some insight into the mechanism of protein/phospho-
lipid interactions. The interactions of soy globulins and phosphatidylcholine were
reported by Kanamoto et al. (94). The results of these studies suggested that
high-energy input is necessary to the formation of stable phospholipid/protein
complexes. Interacting PC vesicles with 7S and 11S soy globulins, Beckwith
(95) demonstrated that the extent of protein/phospholipid interaction was dependent
on both the ratio of the reactants and the specific globulin.
TABLE 21. Physical/Functional Properties of Commercial Lecithin (31).
Property Commercial Range of Values
Viscosity 100 centipoise to plastic

Color Light honey to dark amber
HLB 212
pH 58
Flavor/odor Slightly nutty to moderate bitter, pungent
Solubility
Nonpolar Soluble
Lower alcohols Partially soluble to soluble
Glycerine Partially soluble to soluble
Water Insoluble to dispersible
Chen and Soucie (96) showed that treatment of soy protein isolate with hydro-
xylated lecithin lowered the isoelectric point, increased electrophoretic mobility,
and significantly increased protein dispersibility and suspension stability. Nielsen
(97) investigated the interaction of peroxidized phospholipids with several proteins
under N2. His findings demonstrated a covalent attachment of phospholipids to
proteins whose molecular size is increased.
The interaction of lecithin with starch can also have great functional significance
in food systems. Not surprisingly, the structure of the lecithins involved determines
their reactivity and hence functionality. Hydrolyzed lecithins have been shown to
complex with starch, retarding starch crystallization, and thus slowing staling in
yeast-raised baked goods (98, 99).
The absorption isotherms of several emulsifiers to fat and sugar crystals dis-
persed in oils have been examined (100). Unsaturated monoglycerides and phos-
pholipids cause a decrease in adhesion for all concentrations examined.
Phospholipids reduce the adhesion between sugar crystals, resulting in much denser
sediments.
The influence of soybean lecithins on the spontaneous solidification of different
model fats has been studied in the presence and absence of water (101). Lecithins
added to dry fat do not affect crystallization, but in the presence of water, they
clearly delay it.
Lecithins as antioxidants. The literature is replete with references to the antiox-
idant properties of lecithins. For example, Pokorny (102) claimed that the addition
of soybean phospholipids reduced the rate of autoxidation of sunflower oil and pro-
longed the induction period. Hudson and Ghavani (103) published data showing
that the addition of 0.3% dipalmitoyl phosphatidylethanolamine (DPE) to refined
soybean oil increased the induction time during Rancimat analysis from 8.8 hours
to 19.3 hours. Hildebrand et al. (104), and Jung et al. (105), also published data
demonstrating the antioxidant properties of various phospholipids and commercial
lecithins.
Although lecithins may act as antioxidants in some systems, they also have a
strong synergistic effect in combination with other antioxidants. Hudson and
Mahgoub (106) found that although 98% PC and 98% PE acted as pro-oxidants
in lard model systems, in combination with D-alpha-tocopherol and/or quercetin,
they acted as powerful synergists for antioxidant activity. Hudson and Lewis
(107) confirmed that PE and PC alone have negligible activity as antioxidants in
lard, but they showed that PE is a very effective synergist when used in combination
with polyhydroxy flavonoids at levels of 0.1% or more. Hamilton et al. (108) stated
that ascorbyl palmitate/lecithin and lecithin/tocopherol binary mixtures were
strongly synergistic in delaying peroxidation in fish oils, with ternary blends of
ascorbyl palmitate, lecithin, and tocopherols providing the greatest protection
against autoxidation.
Various mechanisms have been proposed for the mode of action of phospholi-
pids as antioxidants or antioxidant synergists, including chelation of pro-oxidant
metals (106, 108), the release of protons that bring about the rapid decomposition
of hydroperoxides without generating free radicals (107), and the regeneration of
384 LECITHINS
the primary antioxidant (107, 108). Saito and Ishihara (109) provide evidence for
the mechanism of action behind the antioxidant activity of PE and PC in purified
sardine oil. They attributed the antioxidant activity to the basicity of the amino por-
tion of these molecules. The base donates a pair of electrons to an oxygen of a
hydroperoxide molecule, binds to this oxygen, and degrades the hydroperoxide to
an alcohol.
McLean et al. (110) have examined the role of lipid structure in the activation of
phospholipase A2 by peroxidized phospholipids. Results showed that the increase in
rate of hydrolysis of peroxidized phospholipid substrates catalyzed by phospholi-
pase A2 is largely because of a preference for peroxidized phospholipid molecules
as substrates, and that peroxidation of the host lipid does not significantly increase
the rate of hydrolysis of nonoxidized lipids.
The reader is referred to Pryde (86) for a more thorough discussion on the
kinetics of autoxidation of phospholipids; their forming metal ion, iodine, and other
complexes; halogen addition; and their behavior during hydration, hydrogenation
(with heterogeneous and homogeneous catalysts), hydrolysis and alcoholysis,
hydroxylation, oxidation, radical, and other reactions.
4. MANUFACTURE, FRACTIONATION, AND PURIFICATION

OF LECITHINS
4.1. Manufacture of Crude Lecithin

Commercial soybean lecithin is obtained in the traditional manner by hexane
extraction of the crude oil from the soybean flake and then water degumming the
oil to yield a viscous fluid product.
The degumming of soybean oil is not an industry-wide practice. Brian (111) esti-
mated that only about one-third of the soybean oil produced in the United States
needs to be degummed to meet the U.S. needs for soybean lecithin production.
Removal of all phospholipids and gums is a necessary part of the steam-refining
process. However, this process has not yet been developed to the point where it can
produce refined soybean oil that meets U.S. competitive requirements. Studies have
been carried out to fulfill this objective (112, 113).
Based on a series of samples obtained from commercial processors, degumming
of the oil from undamaged soybeans removed 79% to 98% of the phosphorus.
Phosphorus content of the oil was lowered from 500 to 900 ppm in the crude to
12 to 170 ppm in the degummed oil (112). Water-degummed oil from damaged
beans may have an abnormally high phosphorus level, and degumming these oils
poses a difficult problem.
Oil degumming. All stages of oilseed processing affect the quality of soy
lecithin. Prior to degumming, the seed quality, cleaning of the beans, extraction
of the oil, and the handling of the miscella and crude oil all have an important
role in making a good quality lecithin.
Hexane extraction removes about 50% of phospholipids from the meal (114).
The presence of fines in the miscella is undesirable for making lecithin and should
MANUFACTURE, FRACTIONATION, AND PURIFICATION OF LECITHINS 385
be kept to a minimum (0.2% or less), and the crude miscella coming from the
extractor should be filtered. Several miscella filtration methods are available. The
crude oil also can be filtered to produce lecithin.
The method of hexane removal from the miscella is also very important in that
dark colors in lecithin are believed to be caused by an aldehydeamine reaction lar-
gely formed by heating the oil during the solvent stripping operation. Most U.S.
processors employ a dual-stage evaporator followed by a low-pressure stripping
system (115).
Hydration causes most of the phospholipids and gums present in a crude oil to
become insoluble in the oil. Such hydration can come about from water added to
the oil in the degumming step or from moisture picked up from the air by the oil
during storage (116).
Current commercial practices. If the oil is recovered by solvent (hexane) extrac-
tion, some mills allow a portion of the steam blown through the oil for removal of
last traces of solvent to condense and thus hydrate the gums. Sometimes operators
find it difficult to closely control the moisture addition with direct steam and prefer
to add hot water in controlled amounts.
Crude oil from which the lecithin is to be recovered is usually filtered prior
to degumming to remove residual meal fines and seed fragments. Although
more difficult to accomplish, dry lecithin can also be filtered. Careful filtration
results in a highly clarified lecithin with little or no residual hexane insoluble
matter (33).
Brian (111) describes two methods of miscella filtration, one with and one with-
out filter aids, but both result in a lecithin that still remains somewhat cloudy.
Highly clarified lecithin products can be obtained only by filtering the crude oil,
usually with the aid of vertical leaf or plate and frame filters, wherein the dry oil
is heated to 82 C and 0.1% filter aid is added (33).
Two principal degumming methods are employed: batch and continuous. The
batch degumming process is shown in Figure 2 (115). A flow sheet for the contin-
uous degumming of soybean oil and production of soybean lecithin is shown in
Figure 2. Batch degumming system for lecithin production (115).

386 LECITHINS
Figure 3. Flowsheet for degumming soybean oil and crude lecithin production (111).
Figure 3 (111). Recommended conditions for degumming of crude vegetable oils

can vary greatly as shown in Table 22 (111, 117124).
In a batch degumming system, crude soybean oil is typically heated to about
70 C in a large tank fitted with an agitator. Water is added (about 2% by volume),
and the hydrated oil is agitated for up to one hour. The hydration of soybean
TABLE 22. Degumming Conditions from the Literature (8, 118).
Parameter Quantity Reference
Water 75% wt of gums Crauer (119)

12.5% Brian (111)
23% Van Niewenhuyzen (59)
3% Bernardini (121)
1% Norris (122)
2% Carr (123)
Equal to wt gums Braae (117)
25% Andersen (124)
Temperature 3249 C Norris (122)
5070 C Van Niewenhuyzen (59)
6575 C Bernardini (121)
70 C Carr (123)
95 C Andersen (124)
Agitation Vigorous Bernardini (121)
Mechanical agitation Carr (123)
Time 3060 min. Carr (123)
1015 min. Braae (117)
phospholipids proceeds rapidly, and for all practical purposes, 15 minutes is ade-
quate for batch systems. The hydrated gums or lecithin emulsion are then
removed by continuous centrifugation. This step is then followed by drying in a
batch or film dryer. The gums are usually dried to a moisture content of less than
1%, typically 0.30.75% (3, 33).
In continuous systems, preheated crude oil (80 C) and water are metered into an
in-dwell pipeline agitator, or a large agitated tank, and held only for a short period.
In both systems, the oil is then pumped to a centrifuge for separation of the lecithin
sludge from the oil (33, 118, 125126). Water with a low concentration of calcium
and magnesium is preferred (115).
In commercial processes, the amount of degumming water required (1.52.0%)
is roughly equivalent to the phospholipid content of the crude oil (118). Too little
water will result in a dark, viscous gums phase and hazy degummed oil that
contains unhydrated phospholipids. Too much water will result in a three-phase
system consisting of free water, a fluid yellowish-brown gums phase, and a hazy
degummed oil phase after centrifugal separation (33, 118, 119, 126).
Flider (33) points out that the AI content of the gums is enhanced by raising the
temperature of degumming. For example, degumming at 40 C yielded crude
lecithin containing 6365% AI, whereas at 60 C, the yield increased to 6875%.
Although above 60 C some darkening of the lecithin may occur, a higher tempera-
ture (e.g., 7080 C) produces a more consistent lecithin AI on a day-to-day basis.
Agitation is an important factor in batch degumming. The AI content of the
crude gums increases with agitation, presumably because at low agitation rates,
more oil is entrained in the gums (118).
Three types of centrifuges are in common use for oillecithin separations as
described by Podbielniak et al. (127) and Sullivan (128): tubular bowl, disk
bowl, and concentric plate. All of these centrifuges can be hermetically sealed,
thereby protecting the process streams from the harmful effects of air (115). The
newer disk-type centrifuges have a solid ejecting feature that allows the discharge
of solid impurities on a regular basis. These centrifuges are also equipped with a
discharge control valve that can be adjusted to vary the AI content of the gum or
sludge phase. Because the hermetic centrifuges are capable of delivering sludge
of lower oil content than the conventional open bowl-types, more neutral oil is
available for refining, and lecithin with higher AI contents can be obtained (115).
The efficiency of commercial degumming operations is summarized in Table 23
(112). Removal of phospholipids in commercial operation ranges from 75% to
96%, with an average of 87% (112).
For more specific information on the parameters of the degumming operation,
the reader is referred to Brekke (129), Flider (33), List (115), List and Erickson
(130), and List et al. (118).
Novel degumming approaches. List et al. (131) reported on a hexane-extracted
crude soybean oil that had been degummed in a reactor by countercurrently contact-
ing the oil with supercritical CO2 at 10,000 psi at 60 C.
The phosphorus content of the crude oil was reduced from 620 ppm to less than
2 ppm. Degummed feedstocks were fed, without further processing (i.e., bleaching),
388 LECITHINS
TABLE 23. Removal of Phosphorus by Commercial Degumming

of Crude Soybean Oila (112).
Phosphorus (ppm)
Processor Crude Degummed Phosphorus Removed (%) Mean (%)
A 733 167 77.2 82.8

683 80 88.3
B 867 53 93.8 92.3
684 63 90.7
C 711 89 87.5 84.8
588 105 82.1
D 615 40 93.4 95.9
713 12 98.4
E 623 102 83.6 79.7
580 141 75.8
a
Plants located in Illinois, Iowa, Minnesota, Arkansas, and North Carolina. Two samples from each plant
separated by at least 2 weeks.
directly to a batch physical refining step consisting of simultaneous deacidification

deodorization (1 h at 260 C, 13 mm Hg) with and without 100 ppm citric acid.
Flavor evaluation showed that the supercritical CO2- processed oil had the same
flavor scores, both initially and after 60 days of aging and light exposure tests,
as the commercially refinedbleached soybean oil control, deodorized under the
same conditions. These results would indicate that bleaching with adsorbent clays
may be eliminated by a supercritical CO2 countercurrent processing step. As a
result of the considerable heat-bleaching that takes place during deacidification
deodorization, colors of salad oils produced under the above conditions typically
ran 3Y > 0.1R.
A degumming process has been described by Dijkstra (132), wherein the wash-
ing water is recycled to the oil feed and used to dilute concentrated alkali. This pro-
cess does not generate an aqueous effluent and can be used for both acid and alkali
refining, thus allowing refiners to change gradually from alkali refining to physical
refining.
A novel degumming process has been described by Jirjis et al. (133), wherein
vegetable oil miscella is fed to a conditioned polymeric microfiltration membrane
to produce a permeate stream containing decreased weight percentages of phospho-
lipids, and a retentate stream containing increased weight percentages of phospho-
lipids. The membrane is specifically conditioned for the removal of phospholipids
by treating with intermediate solvents, and it has an average pore size of about 0.1
to about 2 microns. Both product streams are then desolventized using traditional
methods. The resulting vegetable oil stream is claimed to have a phosphorus level
of less than 30 ppm.
Bleaching. The color of soybean lecithin can be attributed to many factors: car-
otenoids, melanoids, and porphyrins (125, 134) in the product, age, quality of the
source material, pretreatment prior to crushing of soybeans, thickness of flakes and
temperature during extraction, conditions during degumming, and lecithin proces-

sing conditions (33).
Natural lecithin often has a brown color, although with advanced soybean pro-
cessing technology, the color may approximate that of unbleached soybean oil (65).
Traditionally, one referred to an unbleached product as one that has not been
treated with bleach. A single-bleached product was treated with only one type of
bleach (usually hydrogen peroxide), whereas a double-bleached product usually
was treated with two types of bleach (i.e., hydrogen peroxide and benzoyl perox-
ide). Although these grades continue to exist by name, the bleaching methods used
to manufacture them are no longer uniform. High-quality, relatively light-colored,
unbleached lecithins are now available through modern manufacturing practices.
Additionally, todays double-bleached product may have been treated with only a
small quantity of one type of bleaching agent. Products are presently bleached to a
color specification only, regardless of bleaching techniques or quantity (7, 33).
Lecithin may also be bleached by replacing a portion of the degumming water
with peroxide and carrying out the bleaching and degumming simultaneously. This
method is less efficient, however, than bleaching the gums directly (33).
Although the specifications by the National Oilseed Processors Association
(NOPA) Year Book and Trading Rules, 1998-1999 (135) recognizes color grades
based on unbleached, bleached, and double-bleached lecithins, this nomenclature
is technically incorrect as it is more descriptive of the process rather than of the
product.
From a regulatory point, bleached products are traditionally grouped with the
unbleached forms of crude lecithin. No distinction is made between the bleached
and unbleached forms as far as Generally Recognized as Safe (GRAS) status is
concerned (136).
Laboratory studies by List et al. (118) have shown that with 1% hydrogen per-
oxide, complete bleaching occurs in 30 minutes at 60 C; in commercial operations,
where less efficient agitation occurs, up to 1 hour is required (33).
Bleaching with peroxides involves oxidation of the carotenes and the other color
bodies within the lecithin. There is no evidence that bleaching with either hydrogen
peroxide or benzoyl peroxide functionally modifies lecithins. Bleaching seems only
to affect the pigments, which are not functional constituents. Scholfield and Dutton
(134) reported that although hydrogen peroxide destroys all color bodies to some
extent, its greatest effect is on lutein, the principle pigment (75%) found in soybean
lecithin.
The color of most lecithin products will darken on prolonged heating. Color sta-
bility can be achieved, however, by avoiding exposure of lecithin to temperatures
over 60 C. There are now heat-resistant lecithins on the market that maintain their
light color for extended periods even at elevated temperatures (7).
Drying. After centrifugal separation and bleaching, the gums (containing 25
50% moisture) are dried (to 0.30.75%) as soon as possible to prevent microbial
activity.
The drying operation serves not only to remove moisture, but also to lower the
peroxide value. Peroxide destruction is rapid at or near temperatures of 100 C.
390 LECITHINS
Two types of dryers are commonly used throughout the industry. The sludge can
be dried in batch dryers operating under vacuum (2060 mm Hg) and equipped
with rotating, ball-shaped coils through which warm water is circulated to maintain
the lecithin at 6070 C (140158 F). Although these dryers require longer
residence times (35 hrs), they are popular among European processors because
less charring is apt to occur (137). In domestic lecithin processing plants, continu-
ous, agitated film evaporators are the standard equipment. Evaporators operating on
a vertical or horizontal axis are available. Film evaporators operate at temperatures
ranging from 80 to 105 C, with vacuum of 25300 mm Hg. Residence times are
very short, usually 1 to 2 minutes.
Dry lecithin is highly viscous, and the viscosity increases drastically and then
falls off as the moisture content increases. Comparative conditions used for drying
lecithin in the two types of drying apparatus are given in Table 24 (59, 130).
Because of the sensitivity of lecithin to heat, drying conditions are critical and
the product should be cooled to 5560 C before additional processing, and/or to
3550 C before storage and packing (33). Shelf life of dried lecithin products in
suitable containers is more than 1 year at 21 C (3).
Fluidizing. Fluidizing additives such as soybean oil, fatty acids, or calcium
chloride can be added to adjust the viscosity. The viscosity of dried crude lecithin
can also be decreased by warming it to a maximum of 60 C. The dried crude
lecithin product (unbleached or bleached) can also be used to prepare a variety
of grades of lecithin by removing the oil to increase the phospholipid content, or
by separating the oil-free lecithin into alcohol-soluble and alcohol-insoluble
fractions.
Besides calcium chloride, the viscosity of lecithin products may also be modified
by the addition of other mono- and divalent ions, phosphoric acid, or acetic anhy-
dride. Monovalent ions, such as sodium or potassium, are attracted to the negatively
charged base groups, which tend to increase the crystalline order, thereby increas-
ing viscosity. On the other hand, divalent calcium and magnesium reduce the crys-
talline order and thus reduce viscosity. These techniques are used to produce fluid
lecithins containing 6670% AI without the addition of fatty acids (33). In commer-
cial practice, fluidized lecithins usually are made by calcium chloride addition to
the gums, by the inclusion of fatty acids or vegetable oil, or with the aid of special
proprietary diluents.
TABLE 24. Average Process Conditions for Drying Lecithin Sludgea (59, 129).
Continuous, Agitated-Film
Process Variable Batch Dryer, Bollman Typeb Evaporator
Temperature

C 6080 8095

F 140176 176203
Residence time, min. 180240 12
Absolute pressure, mm Hg 2060 50300
a
Starting product: sludge with 50% moisture. End product: lecithin with less than 1% moisture.
b
Vacuum dryer with rotating, ball-shaped coils heated with warm water.
Fluidization with phosphoric acid is not recommended because darkening of the

product and hydrolysis may occur. Degumming with acetic anhydride results in
fluidized lecithins possibly because PE is acetylated by the reagent. Nonedible
lecithins may be fluidized by the addition of acidulated and dried soapstock.
Plastic lecithins are available in several forms and are typified by high AI, low
AV, high moisture, or their content of certain minerals. One, or a combination of
these, can produce a plastic lecithin. Oil-free lecithins are plastic, because of the
removal of their nascent oil, i.e., residual soybean oil. They are generally powdered
or granular in form.
Nonhydratable phospholipids. According to Myers (138), about 90% of the
phospholipids are removed from the oil by water degumming. Although most of
the remaining phospholipids are removed by alkali neutralization, Braae et al.
(139) report that soybean oil and several other types of vegetable oils often
contain some phospholipids that are not removed by alkali neutralization and
washing.
The impact of enzyme activity on the nonhydratable phospholipid content of
crude soybean oil was investigated by List et al. (140). Evaluation of flakes sub-
jected to live steam and whole beans treated by microwave heating to inactivate
phospholipase D suggests that heat, moisture, and enzyme activity are important
factors contributing to the formation of nonhydratable phospholipids in extracted
crude oils. Approximately 810 minutes of microwave heating is required to
completely destroy enzymatic activity.
List et al. (141) later found that four interrelated factors promote nonhydratable
phospholipids (NHP): (1) moisture content of beans or flakes entering the extraction
plant; (2) phospholipase D activity; (3) heat applied to beans or flakes prior to and
during extraction; and (4) disruption of the cellular structure by cracking and/or
flaking. Thus, NHP formation can be minimized by control of the moisture of beans
and/or flakes entering the extraction process, inactivation of the phospholipase D
enzyme, and optimizing temperatures during the conditioning of the cracked beans
or flakes (141).
In a subsequent study, List and Mounts (142) indicated that the adverse effects of
storage conditions, excessive moisture levels, and elevated temperatures cannot be
completely overcome by inactivation of phospholipase D prior to solvent extraction
of the flakes.
Zhang et al. (143) reported the effects of an expander process on the phospho-
lipids in soybean oil by comparing the differences in phospholipid compositions of
the oils and the lecithins produced from expander and conventional processes by
HPLC. The phosphorus content indicated that the expander-processed oil contained
more phosphorus (985 ppm) than the conventionally processed oil (840 ppm). How-
ever, the phospholipids in the expander-processed oil were more hydratable than
those in the conventionally processed oil. After degumming, the phosphorus con-
tent in the expander-processed oil and conventionally processed oil were reduced
by 93.2% and 78.6%, respectively. The expander-processed lecithin contained
74.3% AI matter, and the conventionally processed lecithin contained 65.8%. There
was also more phosphatidylcholine in the expander-processed lecithin (39.8%,
392 LECITHINS
based on AI) than in the conventionally processed lecithin (34.2%), and the
phosphatidylethanolamine was lower (12.4% vs. 18.1%) and the phosphatidylino-
sitol contents were almost the same.
Braae (144) believes that the nonhydratable phospholipids are present as calcium
and magnesium salts. These phospholipids can have a deleterious effect on oil qual-
ity. They can be removed either by treatment of the oil (at 7090 C) with a small
quantity of concentrated phosphoric acid (0.25%) ahead of the neutralization step
(145) or by refining the oil with a mixture of lye and sodium carbonate. The phos-
phoric acid pretreatment apparently also aids in the removal of deleterious iron
compounds in the subsequent processing of the oil, i.e., caustic refining, bleaching,
and deodorizing of the oil (113). On the other hand, although such pretreatment
aids in the lowering of refinery losses and results in low phosphorus and iron
content in the degummed oil, the resulting lecithin is dark and low in acetone-
insolubles (33).
The use of acetic anhydride as a degumming adjunct has been described by
Hayes and Wolff (146148) and Myers (138). In this process, 0.1 wt % of acetic
anhydride is mixed for 15 minutes with crude soybean oil (phosphorus 750 ppm)
that has been preheated to 60 C (140 F), followed by stirring the mixture for
30 minutes after the addition of 1.5% water. The reaction is completed within
minutes. After centrifugation and water washing, the phosphorus content of the
oil ranged from 2 to 5 ppm.
The advantages claimed for this treatment are thought to be that the caustic refin-
ing step can be omitted, and thus the loss of neutral oil because of saponification is
eliminated, and higher yields are obtained from both finished deodorized oil and
lecithin. On the negative side, the disadvantages found were that equipment and
piping must be constructed of type 316 stainless steel to handle the corrosive
materials; more care is required in deodorization of the oil; and the process will
not produce a satisfactory product from highly colored vegetable oil such as corn
and cottonseed oils, nor from some lots of soybean oil (130). Also, according to
Evans et al. (149), the process removes phosphorus but not iron, one of the metallic
pro-oxidants that can give soybean oil a poor flavor. Lecithin produced from this
process, however, is claimed to be similar to that prepared in the conventional
manner.
Other degumming agents considered include acetic, oxalic, boric, and nitric acids
(150) and surfactants (151). However, none of these are currently used in lecithin
manufacture. Ringers (152) obtained good results in a two-step degumming
process wherein an edible acid, presumably citric, was used. Soybean oil can
also be degummed by heat, but this practice is confined to oils going into industrial
uses.
Lunde et al. (153) concluded that the sequestering action of fatty oil for metal
ions depends at least in part on the oils phospholipid content and reaches a max-
imum at 0.12 ppm of phosphorus. As a point of information, hexane extracts only
about one-half of the phospholipids present in soybeans (1, 114).
For further information on the nonhydratable phospholipids, the reader is
referred to Hvolby (154), Letan and Yaron (155), and Nielsen (114, 156).
4.2. Manufacture of Refined-Grade Lecithins
Producing high-clarity lecithins. Lecithin destined for certain applications may

require more rigorous than usual initial refining conditions. Clarified lecithins are
carefully filtered in (1) the full miscella, (2) crude oil, and (3) directly as lecithin.
As mentioned before, the filtration is carried out on plate and frame or vertical leaf
filters with manual or automatic cleaning cycles. Filtration simply removes HI
material, producing products of utmost purity and clarity. Such ingredients can
be marketed as encapsulated nutritious supplements, as high-purity pharmaceutical
adjuncts, and as additives for high-technology applications.
Although the dry lecithin can be filtered, many processors prefer to filter the
crude oil before degumming (111). Usually, the crude oil is filtered through large
plate-and-frame filter presses. If the degumming operation is conducted at a solvent
extraction oil mill, then either the miscella, i.e., oil-solvent mixture, or the oil can
be filtered. If all of the fines and filter aid are not removed, the dry lecithin will
appear cloudy.
Brian (111) has made suggestions for the type of pumps, centrifuges, filters, heat
exchangers, and drying equipment most suitable for lecithin production. He has
also provided useful engineering design data on filtration flow rates for the crude
oil, miscella, and dry lecithin, quality of filter aid needed, and the overall heat trans-
fer coefficients for agitated-film evaporators and shell-and-tube heat exchangers
used for heating and cooling the oil.
Producing compounded lecithins. Compounded lecithins are special purpose
products made by the direct addition and/or blending of functional additives, emul-
sifiers, diluents, surface active agents, and so on.
Water-dispersible lecithins may be produced by adding a hydrophilic surfactant
(520%) such as polysorbate or ethoxylated monoglycerides. A mixture of lecithin
and nonionic surfactants (1020%) has utility in applications where water dispersi-
bility is needed. Blending of partial glycerides and lecithin, followed by spray cool-
ing, results in flaked or powdered products (33).
When lecithins are diluted with soybean oil or fatty acids, they have a tendency
to separate. In such cases, substituting a portion or all of the soybean oil with other
oils such as peanut, cottonseed, coconut, or partially hydrogenated soybean oil will
increase stability (33).
Lecithins may also exhibit synergistic actions with some compounds. Common
additives include special oils, polysorbates, mono- or diglycerides, modified
monoglycerides, lanolin and lanolin derivatives, solvents, plasticizers, and other
surfactants.
Producing de-oiled lecithins. De-oiled lecithin represents a special category
where high phospholipid content (above 95% AI) is required. When contacted
with acetone, phospholipids precipitate as a fine, free-flowing powder. After remov-
ing the acetone, de-oiled lecithins are dry powders or granules (33).
Depending on the type and efficiency of the extraction equipment, the acetone/
crude lecithin ratio necessary to achieve a 95% phospholipid product is 1020:1
(v/v). In batch extraction, the tank is charged with acetone prior to crude lecithin
394 LECITHINS
addition. Crude lecithin is then introduced into a crystallizing vessel with agitation
until an acetone/crude lecithin volume ratio of ca. 5:1 is achieved. Only the best-
quality fluid crude lecithin should be used for the preparation of de-oiled lecithins.
The mixture is then agitated for 2030 min., after which time the phospholipids are
allowed to settle. The triglycerideacetone miscella is then removed and the vessel
charged again with fresh acetone for the second extraction. A single batch may be
extracted 24 times to obtain the desired phospholipid concentration (95% mini-
mum) (33).
In a continuous extraction, crude lecithin and acetone are simultaneously
metered into a vessel. Within limits, acetone consumption can be decreased by
increasing residence time in the continuous extractor, increasing raw material
efficiency (33).
After extraction, the de-oiled lecithin is recovered by filtration as a cake, con-
taining 2550% acetone. According to Flider (33), the acetone concentration of
the cake is critical for optimal granulation. Too little acetone will result in the
formation of a high concentration of fines and powder. Too much acetone will
result in a salt and pepper effect (i.e., a mixture of coarse and fine particles)
caused by agglomeration of the fines and powder during granulation. The fines
and powder output is 550% of the total de-oiled material, depending on production
conditions (33).
After granulation, the remainder of the acetone is removed by drying, preferably
in a moving bed, forced-air dryer. Such dryers are preferred over fluid-bed dryers
because less destruction of the lecithin granules occurs. After drying, the acetone
content of the product should be well below 50 ppm, preferably below 25 ppm (33).
Flider (33) states that mesityl oxide, through an aldol condensation reaction, may be
formed if excess acetone is present. When the lecithin is sufficiently dry, however,
this is not a problem. The dried de-oiled lecithin is sized by sieving through a series
of screens (33).
As the tocopherols are removed from the lecithin during the extraction process,
the de-oiled lecithin has less oxidative stability than the crude product. Also, the
surface/volume ratio of the de-oiled lecithin contributes to reduced stability. Mixed
soy tocopherols are usually added back at a level of 500 ppm to prevent this. A
small percent of an anticaking agent may also be added to ensure that the product
remains free-flowing. A free-flowing de-oiled lecithin can be easily added to other
products (33).
When compared with crude lecithin, oil-free lecithin is more hydrophilic
and seems to have better emulsifying activity than its AI alone would suggest.
The removal of odor/flavor components with the oil also produces blander lecithins
(7).
Refined de-oiled lecithin can also be blended with carriers such as cocoa butter,
hard butters, medium-chain triglycerides, or other diluents to obtain products with
more functionality and different physical characteristics. Up to 40% phospholipids
may be incorporated in these carriers without the use of solvents. These products
are usually stabilized against autoxidation by the addition of antioxidants (33).
De-oiled lecithin should be packaged as soon as possible to prevent moisture
absorption. For more specific details on various aspects of de-oiled lecithins, the
reader is referred to Flider (33).
Novel de-oiling approaches. A novel de-oiling process has been described by
Hutton and Guymon (157) wherein a mixture of crude phospholipids and hexane
is fed to a polyvinylidine membrane to produce a permeate stream containing tri-
glycerides and hexane, and a retentate stream containing phospholipids (3540%)
and hexane (6065%). The membrane has a molecular weight cutoff of 10,000
50,000 daltons. The permeate stream is then desolventized using traditional meth-
ods. Bleaching earth is added to the retentate stream for color removal at a rate of
58% of the phospholipid mass portion of the stream. The bleaching earth is then
filtered out of the mixture.
Antioxidants in the form of mixed tocopherols are added to the phospholipid/
hexane mixture. The phospholipids are then desolventized through the use of a
drum desolventizer followed by a fluid bed dryer. Solvent residuals in the dried pro-
duct are less than 5 ppm. The dried flakes are placed in storage bins. From there, the
flakes are ground into powder and then agglomerated into granules. The acetone-
insoluble content of the finished product is claimed to be in the range of 9099.9%.
Another novel de-oiling process has been described by Wendel (158) wherein
supercritical gases are used to produce de-oiled lecithin. The crude lecithin is fed
into a column where the supercritical gas mixture of propane and carbon dioxide
flows at a pressure of 80 bar and temperature between 40% and 55 C. This fluid
then goes to a regeneration column where the temperature is increased to 75 C,
and the lecithin component precipitates and falls to the bottom of the column. As
the lecithin falls, it encounters pure supercritical extraction fluid and the oil com-
ponent is extracted. Oil-rich solvent leaves the top of the column. Through pressure
and temperature changes, the lecithin and the oil are precipitated out of their
respective streams and continuously removed from the process flow. The oil-free,
lecithin-free solvent is returned to the column for reuse.
Producing modified lecithins. The chemistry of lecithin has been reviewed by
Pryde (86) and by Wittcoff (4). Schmidt and Orthoefer (58) have discussed the
manufacture and use of modified lecithin products. The latter class is represented
by chemically or enzymatically modified products that are commercially available
in both fluid and de-oiled forms.
The traditional approach to the modification of phospholipid properties is by
fractionation, isolation, and purification of a single component. Functions of phos-
pholipid mixtures are also modified by partial chemical or enzymatic hydrolysis,
acetylation, hydrogenation, and hydroxylation (5).
Crude lecithin contains a number of functional groups that can be successfully
hydrolyzed, hydrogenated, hydroxylated, ethoxylated, halogenated, sulfonated,
acylated, succinylated, ozonized, and phosphorylated, to name just a few possibili-
ties (1). The only chemically modified food-grade products produced in significant
commercial quantities at the present time are the ones obtained by hydroxylation,
acetylation, and enzymatic hydrolysis (58). Hydroxylated or acylated lecithins
represent chemical modifications to improve the functionality in water-based
systems.
396 LECITHINS
Acetylated lecithin. Acetylation occurs primarily on the amino group of phos-

phatidylethanolamine (146148). The amino group of phosphatidylethanolamine,
when acetylated, receives an acetyl group on the positively charged portion of
the phosphatidylethanolamine, which converts it to a negatively charged lecithin
with improved solubility and oil-in-water emulsifying properties (159). Lecithin
can be acetylated using acetic anhydride either by adding the reagent prior to
degumming or adding it to the wet gums. Acetylated lecithin products are made
from natural soy lecithin hydrates by treating them with low levels of acetic anhy-
dride (1.55.0%). Lecithin hydrates are obtained during the degumming of crude
soybean oil. After the reaction with acetic anhydride, the resulting product is neu-
tralized with a food-grade alkali to raise the pH to 6.5 to 8.0, depending on the
intended application. The amount of acetic anhydride used in the process depends
on the level of phospholipids in the gums, and the intended use of the final product,
requiring different degrees of acetylation for optimum functionality. The same is
true for the choice of alkali (e.g., sodium hydroxide, calcium hydroxide, etc.)
used. The product is then vacuum-dried (film dryer in a commercial operation) to
a final moisture of less than 1.0% (160).
The degree of reaction is measured by determining amine nitrogen content in the
resulting product (usually by formol titration). Maximum (100%) acetylation would
be indicated by a zero amine nitrogen value, whereas a minimally acetylated com-
mercial product has approximately 1.7-mg amine nitrogen/g content. In a typical
commercial operation, the amine nitrogen content is usually in the range of 0.7
to 1.7 mg/g.
The total acetone insolubles content of commercial acetylated lecithin products
can vary from about 52% to about 97%, the remainder being soybean oil (or another
food-grade triglyceride or fatty acid as a natural constituent or added diluent), nat-
ural pigments, sterols, and other minor constituents present in crude lecithin from
the soybean. The acetylated lecithin meets all the compositional requirements of the
U.S. Food Chemicals Codex (54).
Typical specifications for a minimally and maximally acetylated, liquid product
are given below.
Minimally Acetylated, Commercial, Liquid Lecithin Specification

Acetone insolubles (%) 60.0 min.; 64.0 max.
Moisture (%) 0.75 max.
Acid value 24 max.
Viscosity (cP at 25 C) 10,000 max.
Color 17 max.
Peroxide value (meq/kg) 10 max.
Hexane insolubles (%) 0.09 max.
Amino nitrogen (mg/g) 1.65 max.
Divalent metals (%) 0.42 min.; 0.48 max.
pH 6.5 min.; 8.0 max.
Maximally Acetylated, Commercial, Liquid, Lecithin Specification
Acetone insolubles (%) 53.0 min.; 56.0 max.
Moisture (%) 0.75 max.
Acid value 36 max.

Viscosity (cP at 25 C) 2,900 max.
Color 12 14
Peroxide value (meq/kg) 100 max.
HIM (Hexane insolubles 0.8 m-Millipore) 100 ppm max.
Amino nitrogen (mg/g) 1.0 max.
pH 7.0 min.; 7.5 max.
Visual clarity pass
Heat resistance test (Lovibond red) 8.0 max.
Appearance Clear and brilliant at 25 C
Acetylated lecithins have improved fluid properties, improved water dispersibil-

ity, and are effective oil-in-water emulsifiers for a wide variety of food formulations
(56, 58). Moderately and highly acetylated lecithins are resistant to heat and can be
repeatedly heated and cooled without darkening. The intended uses for minimally
acetylated products are in infant foods, coffee whiteners, meat sauces, and gravies,
and for oil-in-water cosmetic emulsions. Moderately and maximally acetylated pro-
ducts are used in cheese sauces, release agents in pumpable and aerosol formula-
tions, and shortenings.
The following patents have been issued on the topic of making and using acety-
lated lecithins:
1. U.S. Pat. 3,301,881 Process of Phosphatide Separation, 1967.

2. U.S. Pat. 3,359,201 Lecithin Product and Method, 1967.
3. U.S. Pat. 3,499,017 Alkaline-Hydrolyzed Phosphatides, 1970.
4. U.S. Pat. 3,823,170 Phosphatides, 1974.
5. U.S. Pat. 3,928,056 Pan Release Product and Process, 1975.
6. U.S. Pat. 3,962,292 Phosphatide Preparation Process, 1976.
7. U.S. Pat. 4,479,977 Method of Preparing Heat-Resistant Lecithin Release
Agent, 1984.
Hydroxylated lecithin. Hydrogen peroxide, in addition to bleaching, can also

hydroxylate lecithin. Hydroxylation imparts hydrophilic properties, improves
moisture retention, and contributes to the formation of stable oil-in-water emul-
sions.
Hydroxylated lecithin is a light-colored product with increased water dispersibil-
ity and enhanced oil-in-water emulsifying properties. Hydroxylated lecithin is use-
ful in many applications in which a water-dispersible lecithin is desired. It is
especially useful in baking applications where it can improve the dispersion of
fats and retard staling.
Hydroxylation of lecithin is carried out by the reaction of crude lecithin with
hydrogen peroxide and lactic acid or acetic acid. Active sites for peroxidation
appear to be double bonds as measured by IV drop and the isolation of dihydrox-
ystearic acid from the reaction mixtures. Hydroxylation is allowed to proceed until
398 LECITHINS
a 10% reduction in iodine value occurs (115). The ethanolamine group is also mod-
ified during hydroxylation (58, 161).
Hydrolyzed lecithin. Crude lecithin is readily hydrolyzed in the presence of
strong acids or bases. Enzymes can be used for very selective hydrolysis. Prolonged
treatment leads to fatty acids, glycerophosphoric acid, or their salts, with mixtures
of amino compounds and carbohydrates (4, 115).
In a commercial process, it is desirable to control the reaction so that just one of
the fatty acids is cleaved from the phospholipid molecule. As acid or base hydro-
lysis is nonspecific and very difficult to control, enzymes are usually preferred for
most applications (58). A number of phospholipase enzymes are available (i.e.,
phospholipase A1 or phospholipase A2).
Haas et al. (162) have studied enzymatic phosphatidylcholine hydrolysis in
organic solvents by examining selected commercially available lipases. Enzymatic
hydrolysis of oat and soy lecithins, and its effect on the functional properties of
lecithin, was investigated by Aura et al. (163). The phospholipase used was most
effective at low enzyme and substrate concentrations.
Partially hydrolyzed lecithins exhibit enhanced oil-in-water emulsifying proper-
ties, particularly in the presence of calcium and magnesium ions. They do not lose
their emulsifying action in the presence of calcium and magnesium ions as rapidly
as do the unmodified types. Enzymatically modified lecithins have been used in calf
milk replacement formulations to improve the emulsification and digestibility of
fats (56).
Enzymatic hydrolysis of the polar head group of a phospholipid molecule can be
carried out with phospholipase C and phospholipase D. Phospholipase D is used to
exchange the amino head group of phosphatidylcholine with serine to form phos-
phatidylserine (164).
Transesterification. Transesterification allows for the incorporation of free fatty
acids into lecithin molecules. Unhydrolyzed lecithin contains two fatty acids, and
the fatty acid moiety can be different at the two positions on the phospholipid mole-
cule. The fatty acid composition can have an effect on the stability and functionality
of the lecithin. Changes in the fatty acid composition can be done through transes-
terification (165). Transesterification using lipases can be used for the addition of
polyunsaturated fatty acids to lecithin to enhance the essential fatty acid profile, or
to improve functionality (166).
Hydrogenated lecithin. Lecithin can be hydrogenated to a stearin-like solid that
has greater oxidative stability and is less hygroscopic than unmodified lecithin, but
it has reduced solubility in the usual solvents. Phospholipids are not hydrogenated
as readily as soybean oil, which at lower hydrogenation pressures and temperatures,
can be selectively hydrogenated (58).
Hydrogenation of lecithin is usually done under conditions to reach iodine
values of 1020 in the presence of a nickel or palladium catalyst and a suitable sol-
vent (e.g., ethyl acetate) at 7585 C under 70 atmospheres pressure. Bromine or
chlorine also readily adds across double bonds yielding products useful in lubricant
formulations. Iodine can be added by warming granular lecithin dissolved in acetic
acid in the presence of iodine and magnesium or aluminum catalyst (58).
Producing fractionated lecithins. Finally, fractionating crude lecithin directly, or

after de-oiling, is another way of creating a variety of products with tailor-made
functionalities. Alcohol, or mixed solvent fractionation, combined with other tech-
niques, can produce lecithin products that have been greatly enriched in particular
phospholipids. Separating the acetone-solubles from crude lecithin increases the
amount of phospholipids in the acetone-insoluble fraction by decreasing the amount
of triglycerides.
Further fractionation can separate the alcohol-soluble phosphatidylcholine from
the alcohol-insoluble phosphatidylinositol. Commercial products that are alcohol-
soluble contain concentrated phosphatidylcholine (4060%) and only low levels
of phosphatidylinositol. The alcohol-insoluble products are enriched in phosphati-
dylinositol (4060%), whereas their phosphatidylcholine content is greatly reduced.
The phosphatidylethanolamine component is approximately equally partitioned
between the two fractions. The alcohol-soluble grades tend to be more oil-in-water
emulsifiers, whereas the alcohol-insoluble grades are more effective in water-in-oil
systems (7).
The PC/PE ratios of alcohol-fractionated lecithins are largely determined by pro-
cessing variables such as alcohol polarity, concentration, lecithin/alcohol ratio, tem-
perature, and extraction time (33). By extracting natural lecithin with a PC to PE
ratio of 1.2:1 with 90% ethanol, an alcohol-soluble fraction with a PC/PE ratio of
8:1 can be obtained (33, 120). The fractions may be blended with other surfactants
or carriers to obtain desired functionality.
To obtain individual phospholipids of greater than 5060% purity, some form of
selective adsorption process is usually required. Adsorption and distribution chro-
matography present these options. Treatment of the alcohol-soluble lecithin with
alumina yields a fraction very rich in phosphatidylcholine and free of phosphati-
dylethanolamine and phosphatidylinositol (167). Although these products are avail-
able only in very limited quantities for highly specialized markets, products such as
a lecithin containing up to 95% PC can be obtained commercially.
Storage and handling. Liquid lecithin can be kept for years provided closed con-
tainers are used and the temperature does not exceed 2025 C. Bleached products
require more careful storage and handling. Color reversion will occur rapidly in
bleached products, particularly at elevated temperatures. Decomposition of perox-
ide is thought to contribute to color reversion in bleached products. In order to pre-
vent this phenomenon, low storage temperatures are recommended (115).
Very low temperatures should, however, be avoided when storing liquid lecithin
products because physical separation of the phospholipids and oil may occur. Phy-
sical separation is more likely to occur in low AI products. When separation does
occur, remixing at 4060 C will redisperse the oil and lecithin phases. In bulk
handling of lecithin, storage temperatures of 3035 C are acceptable. However,
prolonged storage at these temperatures may cause darkening (115).
De-oiled granular lecithin can be stored up to 2 years at temperatures below
25 C. If desired, it may be stored in a frozen state at 0 C, but because of its hygro-
scopic nature, the product should be allowed to come to room temperature before it
is exposed to the atmosphere (33).
400 LECITHINS
5. FOOD-GRADE LECITHIN PRODUCTS, USES
5.1. Functionality
Commercial lecithins are multifunctional food ingredients. The combined hydro-
philic and lipophilic properties of phospholipid molecules give them surface-active
effects in many applications. As surfactants, they can exhibit a variety of functions
common to other surfactants while they also have unique functionalities of their
own.
Commercial lecithin products that were sold many decades ago for applications
such as chocolate and confectionery products, margarine, bakery goods, pasta pro-
ducts, textiles, insecticides, and paints are still active today because of their emul-
sifying, wetting, colloidal, antioxidant, and physiological properties. Lecithins
multifunctional properties and its natural status make it an ideal food ingredient.
The major applications and functional properties of lecithin products are shown in
Table 25 (7).
TABLE 25. Functional Properties (7).
Ingredient in Function(s)
Margarine Emulsifier, antispattering agent

Confections and snack foods
Chocolate Crystallization control, viscosity control, anti-sticking
Caramels
Coatings
Instant foods
Cocoa powders Wetting and dispersing agent, emulsifier
Instant drinks
Instant cocoa
Instant coffee
Protein drinks
Dietetic drinks
Coffee whiteners
Milk replacers
Cake mixes
Puddings
Instant toppings
Commercial bakery items
Breads Crystallization control, emulsifier, wetting agent,
release agent (internal and external)
Rolls
Donuts
Cookies
Cakes
Pasta products
Pies
Cheese products
Pasteurized processed Emulsifier, release agent
Cheese and cheese food
Imitation cheese
FOOD-GRADE LECITHIN PRODUCTS, USES 401
Ingredient in Function(s)
Meat and poultry processing

Meat and poultry glazes and basting
compounds Browning agent, phosphate dispersant
Pet foods Dietary supplement, release agent, emulsifier
Bacon
Dairy and imitation dairy products
Infant, milk formulas Emulsifier, wetting and dispersing agent,
anti-spattering agent, release agent
Milk and cream replacers
Egg replacers
Imitation eggs
Whipped toppings
Ice cream
Flavored milks
Flavored butters (garlic, etc.)
Basting butters
Miscellaneous products
Peanut spreads Crystallization control, emulsifier
Salad products
Flavor and color solubilization
Packaging aid
Polymer package, interior coating Release agent, sealant
Can interior coating
Sausage casing coating
Stocking net
Processing equipment
Frying surfaces Internal (in product) and/or external release agent,
lubricant
Extruders
Conveyors
Broilers
Dryers
Blenders
Evaporators
As a variety of methods are available for modifying the emulsifying properties of

commercial lecithin, the potential for improved, tailor-made, functional products is
unlimited. The main functional properties are emulsification, antispatter,
instantizing/wetting/dispersing, release/parting, viscosity modification, and baking
applications.
These functional characteristics are primarily derived from the chemical struc-
tures of lecithins major phospholipids (Figure 1) (7). Phospholipid molecules con-
tain two long-chain fatty acids esterified to glycerol, as well as a phosphodiester
bonding a choline, inositol, or ethanolamine group. A phospholipids fatty acid
end is nonpolar and thereby lipophilic (or fat loving). Conversely, the phosphodie-
ster, with the above-mentioned constituents, is zwitterionic (or dipolar), which
402 LECITHINS
explains the hydrophilic (or water loving) properties of this portion of the molecule
(65).
Because of their charged nature, the phospholipids are susceptible to the ionic
environment in which they function. Based on testing in the laboratory, Dashiell
(31) suggests that salt concentrations greater than 2%, and pH less than 4, contri-
bute to a detectable loss in lecithin functionality. Similar results have been reported
elsewhere (73).
The following commercial lecithin modifications were described in a publication
from Central Soya Co., Inc. (168).
1. The use of oil free lecithins as emulsifiers, lubricity enhancing agents, and
blending aids.
2. Producing low-viscosity, fluid lecithins as wetting, dispersing, and release
agents.
3. Hydroxylated lecithins with enhanced emulsification, dispersing, and wetting
properties.
4. Highly filtered lecithins for use in health food applications.
5. Special heat-resistant lecithins for release applications.
6. Lecithin/distilled monoglyceride blend for bakery applications.
7. Enzyme-hydrolyzed lecithin for bakery, and emulsification applications.
Weete et al. (169) have reported on the improvement of lecithin as an emulsifier

for water-in-oil emulsions by thermalization. Various forms of lecithins can be
heated under certain conditions of time and temperature to greatly improve their
properties as emulsifiers for water-in-oil emulsions. Viscosity, discontinuous
phase-holding capacity, stability, and water retention were greatly enhanced in
emulsions containing thermalized lecithins as the emulsifier compared with those
prepared with corresponding amounts of nonthermalized lecithins. The improved
emulsification properties of thermalized lecithins appeared to be, at least in part,
because of an increase in diglycerides and free fatty acids resulting from the
thermal degradation of phospholipids.
Emulsifying properties. One of the major functions of commercial lecithins is to
emulsify fats. In an oil:water system, the phospholipid components concentrate at
the oil:water interface. The polar, hydrophilic parts of the molecules are directed
toward the aqueous phase, and the nonpolar, hydrophobic (or lipophilic) parts are
directed toward the oil phase. The concentration of phospholipids at the oil:water
interface lowers the surface tension and makes it possible for emulsions to form.
Once the emulsion is formed, the phospholipid molecules at the surface of the
oil or water droplets act as barriers that prevent the droplets from coalescing,
thus stabilizing the emulsion (159).
Commercial lecithins are used in both water-in-oil (w/o) and oil-in-water (o/w)
emulsions. For w/o emulsions, like margarine or ready-to-use frostings, oil-loving,
lipophilic lecithins are typically used. For o/w emulsions, like sauces or infant for-
mulas, water-dispersible, hydrophilic lecithins are typically used (7, 31). The use of
lecithin in oil-in-water emulsions requires the modification of lecithin to increase

its apparent hydrophilicity. The techniques available are somewhat limited for
food-grade lecithins. The commonly employed methods for producing water-
dispersible lecithins include acetylation, hydroxylation, enzyme-hydrolysis,
fractionation, de-oiling, or blending with other hydrophilic emulsifiers.
The manner in which lecithin is modified to achieve increased hydrophilicity
will greatly affect its emulsification properties. Different modifications will create
lecithin products with different apparent HLB (hydrophile-lipophile balance)
values, a term used to convey the approximate degree of water dispersibility
(hydrophilicity) of lecithin products (31). The higher its HLB value, the more water
dispersible the lecithin product. In o/w emulsions, the type of fat to be emulsified
may require a specific type of hydrophilic lecithin for optimum emulsion stability.
Dashiell (31) provides a short listing of fat types, and the corresponding class
of lecithin found to give the most stable emulsion in model systems of water/fat/
emulsifier.
Standard-grade (crude) lecithins are excellent water-in-oil emulsifiers. However,
modified lecithins can function to emulsify either water-in-oil or oil-in-water emul-
sions, depending on the type of lecithin modification and the specific parameters of
the system. These system parameters can include pH, types of components, com-
ponent ratios, solids content, and others. Unlike crude lecithins, hydroxylated
lecithins are stable in acid systems (pH 3.5). Fractionated lecithins can be man-
ufactured for specific emulsion types. As lecithins emulsifying activity is partially
dependent on its phospholipid ratio, changing the ratio can alter its emulsifying
capabilities (7).
Emulsifier/stabilizer systems are normally used to make stable food emulsions.
Thus, lecithin is generally not called on to handle the entire emulsification, but it
works in combination with other emulsifiers and stabilizing polymers such as pro-
teins, starches, and gums (31). Lecithin will break up (emulsify) the particles, and a
stabilizer (water-soluble polymer, etc.) will hold the particles in a dispersed orien-
tation when a stable emulsion is formed.
Marrs et al. (170), showed that a combination of lecithin and carrageenan pro-
duced more stable o/w emulsions with corn oil than either the lecithin or carragee-
nan alone. The emulsifying properties of the lecithin/carrageenan combination was
thought to be caused by interactions in the aqueous phase between the negatively
charged sulfate ester groups on the carrageenan with the positively charged amino
head groups on the PC and PE of the lecithin. This stabilization occurs when the
carrageenan component is of the kappa or iota type, both of which form macromo-
lecular networks in the presence of cations.
The rate of creaming of an emulsion is governed by Stokes Law (171). Conse-
quently, in an o/w emulsion, fat separation can be delayed by reducing the fat dro-
plet size, by matching the densities of the dispersed and continuous phases, and/or
by increasing the viscosity of the continuous phase. The amount of energy applied
to the system when the emulsion is created (for example, with a high-pressure
homogenizer) determines the initial fat droplet size of the emulsion, and the pre-
sence of an emulsifier on the surface of the droplets, prevents the dispersed oil
404 LECITHINS
droplets from coming together and creaming out. The viscosity of the continuous
phase can be increased by the addition of gums, starches, or other stabilizers.
Dashiell (31) reported that with high levels of good-quality protein present, the
selection of a system-specific lecithin (one giving the best emulsion in an oil/emul-
sifier/water system) becomes less important. In fat creaming tests, in fat/water/pro-
tein/lecithin emulsions, results showed that with limited protein in a whey-
stabilized (low-protein) system, more functional lecithins gave a clear advantage.
In a casein-stabilized system (containing abundant protein), differences between
lecithin products were less dramatic. Agboola et al. (172), showed that the presence
of 0.25% of a de-oiled, hydroxylated lecithin, stabilized o/w emulsions formed with
whey protein hydrolysate after retorting.
Solubilization. Most lecithins can aid in the production of microemulsions, an
example being oil-soluble flavors in aqueous systems. Although standard-grade
lecithins do not disperse in water, many modified or fractionated lecithins are
water-dispersible, and they can be used to produce microemulsions. Standard-grade
lecithin can be blended with other surfactants (e.g., ethoxylated monoglycerides) to
produce synergistic emulsifier blends that are also effective in producing micro-
emulsions.
Solid particle dispersions (Sols). Many lecithin products are still the best and
most effective surfactants for dispersing sols. This seems to be because of lecithins
affinity for solidsliquid surface interfaces. Phospholipids seem particularly
attracted to particles containing metals and metal salts. Examples of food sols
include some liquid chocolates, instant drinks, frosting mixes, pigmented foods, and
others. The nonfood applications include paints, inks, and other pigmented coatings.
Foams. Refined lecithins have been employed as effective foam control agents.
Examples include whipped toppings, ice creams, and many types of candies.
Refined lecithin products have also been employed as effective defoaming agents
in foams caused by powdered proteins in water. This is an excellent example of the
system specificity of lecithin products (7).
Wetting/instantizing properties. Lecithin products are effective wetting agents
for a wide variety of powdered or granular products. Lecithination of powders
for improved wetting, and control of dusting problems, is widely practiced. Instan-
tizing effects can be obtained by including the proper lecithin product in a food for-
mulation. Specific lecithin products that are compatible with the various
manufacturing techniques used for instantizing are commercially available.
Lecithin products have been formulated to instantize many types of food pow-
ders to achieve rapid wetting and dissolution. As powder compositions can vary
greatly (from hydrophilic to lipophilic), proper lecithin selection is done on some-
thing of an empirical basis (31). Certain general principles apply, however. If a
powder is hydrophobic, or contains a significant amount of surface fat, typically
a water-dispersible, hydrophilic lecithin is used to reduce the surface tension
between the powder and the water so that the powder wets and disperses easily.
If a powder is hydrophilic, like protein concentrates or isolates, typically a lipophi-
lic lecithin is used to control the rate of hydration of the powder so that it wets and
disperses without skinning or forming large lumps.
Manufacturing techniques employed in producing instant products include

spray-coating dry powders with fluid lecithin products, cospray drying powders
with more hydrophilic lecithins such as the oil-free forms, or hydroxylated lecithin,
and agglomeration of the powder with an aqueous dispersion of a hydrophilic
lecithin. Some types of powders, for example, starches, gums, and chocolate drink
mixes, require agglomeration with an aqueous dispersion of lecithin to achieve
optimum wettability and dispersibility.
Examples of foods that can be instantized with lecithins include cocoa powders,
instant drink mixes, powdered coffee whiteners, milk replacers, cake mixes, pow-
dered instant puddings, and instant soups and sauce mixes.
Release/parting properties. Lecithin functions as the active ingredient in a wide
variety of food-grade release formulations. Products for institutional and retail use
are available in aerosol and nonaerosol forms containing from 0.5% to about 15%
lecithin (31). Common ingredients in release formulations are as follows (31) (from
a Central Soya Co. market survey, 1986).
Vegetable oil (all major vegetable oil classes are used)

Hydrogenated vegetable oil
Mineral oil
Lecithin
Flour
Amorphous silica
Artificial flavor
Artificial color
Beta-carotene
Preservatives
Antioxidants
Antifoams
Water
Lecithin usage levels in commercial release formulas are limited by a tendency

of the lecithin to separate from some oils. The tendency of ordinary lecithins to
darken, polymerize, and foul on heated metal baking surfaces also limits their
use level in commercial release formulas. Moderately and highly acetylated
lecithins, however, are resistant to heat and can be repeatedly heated and cooled
without darkening (173).
Lecithins can be dissolved in oil, dispersed in water, or used as is, in release
applications. Oil-free lecithins can be dry blended into breading, coatings, and spice
or seasoning mixes for release of the coated food product from the food-contact
surface. In food products that have a high surface area-to-volume ratio, like pan-
cakes or fortune cookies, lecithin can be added directly to the product formulation
to achieve release from the cooking/baking surface. Effective release depends on
406 LECITHINS
the presence of lecithin between the food and the food contact surface. If this is
achieved, the food product should not stick to the food contact surface.
Crystallization control. Lecithin can control crystallization in various food sys-
tems. In foods containing sugars or fats, the presence of as little as 0.5% lecithin
can produce altered crystal sizes and structures that can have positive effects on
product texture and viscosity. This is important in cookie fillings, butter-containing
maple syrups, ice cream toppings, and similar products (7).
5.2. Specific Food Applications

As mentioned previously, soybean lecithin is used in food because of its emulsify-
ing, wetting, release, and other surfactant qualities. Relatively small amounts of the
lecithin are needed, often only 0.1% to 2% in foods. These use levels are more or
less consistent with those of chemical surfactants (7). At these low levels of usage,
the color, flavor, and odor of the lecithin normally are not noticeable. When lecithin
is used in conjunction with synthetic emulsifiers, it sometimes has a synergistic
effect, and thus lesser amounts of the synthetic emulsifiers need be used.
General food applications of lecithin include margarine, confections, snack
foods, soups, instant foods, bakery products, simulated dairy products, processed
meat/poultry/seafood products, and dietary applications. The most widespread
uses of crude lecithin products are in confections and margarine (7, 174).
Margarine/shortenings. Standard-grade (crude) lecithin is the classic emulsifier
in margarine and is added at the 0.10.5% level to the fat phase. It is commonly
used in conjunction with mono- and diglycerides. The lecithin prevents weeping
or bleeding of the moisture present, reduces spattering, promotes browning dur-
ing frying, increases the shortening effect when margarine is used in baking, and
helps to protect the Vitamin A in fortified margarine from oxidation (174). More
complete and uniform blending of shortening occurs when 0.5% to 1.0% lecithin
is added (160).
Confections. There are three major specific properties for lecithin in confections:
emulsification (e.g., caramels), anti-stick/release properties, and viscosity modifica-
tion (e.g., chocolate) (175). None of these properties stand alone. For example,
emulsification in caramels will influence shelf life and texture. In chocolate,
viscosity modification will alter production costs and texture of the finished
product.
Addition of 0.250.35% standard-grade lecithin to the chocolate used in candy-
making reduces its viscosity markedly, enables the manufacturer to apply a uniform
coating and thus use lesser amounts of expensive cocoa butter, decreases the time
for grinding and mixing the various ingredients, and produces a more stable choco-
late. Stabilized lecithin-containing chocolate has improved handling characteristics
and is more resistant to fat-and-sugar bloom or graying. Use of lecithin in other
fat-containing candies also prevents graining, streaking, and greasiness.
Studies by Sinram and Schmitt (176) have shown significant improvements in
dark and milk chocolates using a fractionated, phosphatidylcholine-enriched soy
lecithin as compared with a standard soy lecithin or no lecithin at all. The effect
of reduced viscosity and improved yield value in chocolate depended on the type
and dosage of lecithin as well as the fat content of the chocolate.
Addition of 1% to 2% of lecithin to peanut butter gives a smoother, creamier
spread. The peanut butter does not separate under wide temperature variations.
Bakery products. Lecithin is a useful emulsifier in baked goods such as bread,
cakes, sweet goods, biscuits, and crackers. Standard fluid lecithin is not readily dis-
persible in water, giving it limited functionality in a dough or batter where water is
a key component. Modified lecithins, which are water dispersible, provide many
benefits to baked goods including improved shelf life, a stronger gluten complex
in yeast-leavened dough, reduced dough stickiness, improved tenderness, better
release, and reduced checking in products such as crackers and thin bread sticks.
Occasionally lecithin is incorporated into shortenings (solid or fluid) that are
used in baking, but it is more frequently added as a separate ingredient. It can
also be added as part of a dough improver. Gaubert et al. (177) patented a baking
improver composition that contained 20% to 30% (by weight) lecithin. It is an
easy-to-handle dough improver where the lecithin acts as a binding agent and an
emulsifier.
In yeast-leavened dough, the addition of 0.10.3% commercial lecithin improves
water absorption, ease of handling, fermentation tolerance, shortening value of fat,
volume and uniformity, and shelf life (89). If enzyme modified lecithin is used, it
extends shelf life by retarding staling or starch retrogradation (99, 178). Lecithin is
employed in cake formulations, such as box mixes, so that they will wet rapidly
when mixed with water. In biscuits, crackers, pies, and cakes, 1% to 3% lecithin
(on shortening basis) promotes fat distribution and shortening action, facilitates
mixing, and acts as a release agent (179).
Bread and rolls. There is a general consensus about the beneficial effects of
lecithin on dough-handling properties [Aberham (180), Kuntze (181), Pomeranz
(179), Puchkova et al. (182), Pyler (183), and Zapryagaeva et al. (184)]. Lecithin
also has a beneficial effect on baking performance. It is commonly believed that
surfactants, such as water-dispersible lecithin, form lamellar-type, ordered struc-
tures in the water phase of the dough. Those ordered structures improve the stability
of the film surrounding entrapped carbon dioxide (185, 186), resulting in increased
volume and improved crumb structure (178). Over 50 years ago, it was confirmed
that the addition of lecithin also improved the extensibility, dryness, and machin-
ability of dough, producing bread that has improved symmetry, grain, and texture
(187).
Researchers have shown that the native phospholipids in wheat play an impor-
tant role in the baking quality of flour (188), and the addition of enzyme modified
soy lecithin can make further improvements (99). Chung and Pomeranz (189)
reported that fractionated lipids, especially phospholipids at 0.2%, provided a sig-
nificant increase in loaf volume when shortening was added. Johnson et al. (190)
reported that adding soy lecithin (PC) to chlorinated, petroleum ether-extracted
flour, at 0.2% flour weight, improved volume and grain beyond that obtained
with the unextracted flour. Chung et al. (191) found that petroleum-extracted polar
lipids were required at a level of 180 mg per 100 g of flour (H.R.W. 12% protein) to
408 LECITHINS
produce bread of desirable volume. Polar lipids were 50 times more functional than
protein in improving loaf volume. Cole et al. (192) studied the effect of phosphorus-
containing lipids (polar lipids) and soy (PC) lecithin on the quality of cookies baked
from defatted flour. They found that although those fractions containing lecithin
completely restored cookie quality, a phosphorus-free lipid fraction did not.
Enzyme-modified lecithin has the ability to form a complex with the amylose
portion of starch, and the straight portions of amylopectin (193195). By forming
a complex with starch, enzyme-modified lecithin effectively slows starch retrogra-
dation and staling. In a study, two water-dispersible lecithins were evaluated against
hydrated monoglycerides as starch complexing agents to prevent staling in white
pan bread (196). The staling indices demonstrated that water-dispersible, de-oiled
soy lecithin gave no improvement in softening versus the control to which no
emulsifier was added. An enzyme-modified lecithin, however, gave a significant
softening response. This behavior was attributed to starch complexation by the
lysophospholipids.
Many workers have demonstrated the synergistic effects in bread making of
lecithin in combination with mono- or diglycerides and other surface-acting agents.
According to Hampl and Tvrznik (197), the use of lecithin in combination with
monoglycerides (1) improves quality characteristics of the raw materials, (2) opti-
mizes technical processing, (3) reduces shortening requirements, and (4) improves
overall quality of the final product, including freshness retention and nutritive value.
Haarasilta et al. (198) patented an enzyme product containing de-oiled lecithin, or
lecithin spray dried with a carrier, for use as a dough improver for bread.
Pomeranz et al. (199) have also studied the effect of 0.5% commercially avail-
able lecithins on the quality of bread made from untreated and petroleum ether-
extracted flour, at three different shortening levels (0.0%, 0.5%, and 3.0%). The
best results were obtained with alcohol-soluble soy phospholipids containing a
2:1 mixture of PC and PE in both untreated and petroleum ether-extracted flours.
When added to petroleum ether-extracted flours, 0.5% alcohol-soluble phospholi-
pids replaced 0.8% extracted free flour lipids and 3.0% shortening. Excellent results
were also obtained with hydroxylated lecithin, but only with shortening present. In
a separate study, Glabe and Anderson (200) tested carrageenan and hydroxylated
lecithin in continuous mix bread. Their results indicated that hydroxylated lecithin
increased dough stability and loaf volume when carrageenan was present.
Pyler (183) reports that hydroxylated lecithin improves dough extensibility. It
has been suggested (179) that hydroxylated lecithins are particularly valuable in
bakery products because of their apparent synergy with mono- and diglycerides
in addition to their high dispersibility in water systems in contrast to the oil solu-
bility of most lecithins.
Adler and Pomeranz (201) have shown that the addition of lecithin to soy flour-
enriched bread can improve its consumer acceptability in the absence of shortening.
Even in the presence of shortening, an improvement was observed with the use of
lecithin (202).
Mizrahi et al. (203) described the improving effect of soy lecithin on bread con-
taining soy protein isolate. The use of soy lecithin in conjunction with sucrose
esters exerted an improving effect on bread quality in high-protein breads made

with soy flour (204).
There is a patent on the synergistic effect of hydrophilic lecithins (HLB-8 or
higher) on lipophilic surfactants such as glycerol monostearate (GMS), used pri-
marily in bread and other bakery foods to retard staling (205). Not only does the
use of a hydrophilic lecithin result in improved shelf life, but also in improved
dough conditioning as exemplified by increased loaf volume, improved symmetry,
grain, and texture. Larsson and Eliasson (186) tested the effect of added lecithin on
the dough rheology of flour milling streams. They used liposomes formed with a
high phosphatidylcholine lecithin, and they concluded that flours containing less
polar lipid would be most improved by lecithin addition.
Shogren et al. (206) have found that lecithin, and other dough-conditioning sur-
factants, counteracted the deleterious effect of up to 15% wheat bran that was added
as a source of fiber in bread.
Cookies, crackers, waffles. In the processing of cookies and other baked goods
containing significant quantities of fat, lecithin promotes even dispersion of the fat
throughout the dough (1).
Lecithin aids in the dispersion of fat in semisweet dough, and it improves the
emulsification during the creaming stage of short dough (207). It also extends
the fat. Lecithin is easily mixed into cracker and cookie dough to modify the con-
sistency, and to make machining easier by reducing stickiness in the finished
product. Greasiness of cookies with high shortening content is often reduced by
adding small amounts of lecithin to the dough.
Matz (208) also has reported on the improvements obtained with the use of
lecithin in the production of cookies. Cookie dough is drier and more machinable
with the use of lecithin. Lecithin improves the dispersion of fat so that it more read-
ily mixes with sugar, flour, and other ingredients. Improved emulsification also
reduces mixing times. Overdevelopment of the dough can result in lack of tender-
ness in the cookie. The release quality of lecithin improves the extrudability and
release from the die, improving definition of impression.
Kissel and Yamazaki (209) studied the effects of lipid extracts from wheat flour
and soy and safflower lecithins on improving cookie spread when the cookies were
made from protein-fortified wheat flour. Soy lecithin was most effective, and it also
improved the quality of cookies made from weak flours such as sorghum and millet
(210).
Waffles made from dough that contains lecithin show better iron-grid release
and easier handling. The waffles are stronger, crisper, do not become soggy, and
retain freshness better (211). Pomeranz (179) reported that the use of lecithin in
waffle formulations improves release from the grill, provides strength and crispi-
ness, and reduces sogginess. He also cited the use of lecithin as a mold release agent
(i.e., 0.3%) in low-fat formulations such as ice cream cones.
Cakes, cake mixes, pancakes, doughnuts, sweet rolls, fillings, etc. Incorporating
lecithin in cake formulations substantially improves the quality of cakes (212).
Hydroxylated products that have an intermediate degree of saturation were found
to be best. Lecithins have been recommended as emulsifying agents for cake mixes
410 LECITHINS
to assure easy pan-release and to prevent the cakes from falling or dipping in the
center (213, 214). They also improve volume, crumb structure, tenderness, and
shelf life. When lecithin was incorporated at a rate of 1% to 3% into a powdery
foaming agent for cake and fry batters, it homogenized the monoglycerides and
provided the foaming property (215). Enzyme-modified lecithin could be used
effectively.
Wolf and Sessa (216) have advocated the use of lecithin in cake doughnut for-
mulations at 0.5% to 1.0% (based on mix weight) to accelerate mixing of the batter.
Prolonged batter mixing results in less tender crumb in the finished product.
Lecithin is also beneficial in white cakes, and others that contain only egg whites,
by acting to replace the phospholipids normally coming from egg yolk in the
formula.
According to Wolf and Cowan (71), the emulsifying properties of phospholipids
find extensive use in cake mixes and instant foods. Adding 0.5% to 1.0% lecithin
promotes wetting, thereby speeding up mixing of cake-doughnut mixes. Adding
lecithin improved keeping qualities, grain, and texture in sweet-dough products
(coffee cakes, sweet rolls, etc.), and produced shorter dough in these items.
Incorporating lecithin into pie crusts reduced mixing time, produced flakier
dough, enhanced release, contributed to uniform browning, and aided as a moisture
barrier to protect the crust (179).
Lecithin acts as an aid for the blending of unlike ingredients. An excellent exam-
ple is the formulation of cream fillings for sandwich cookies. The use of low levels
of lecithin significantly improves the ease of blending and mouthfeel of these
products, which consist mostly of low-polarity shortening and high-polarity sugar.
The lecithin serves as an intermediary to significantly reduce the stiffness and
mixing time of the filling (31).
Reduced fat baked goods and extrusion. Lecithin is well known for its lubricat-
ing effects. In 1947, Pratt (217) concluded that lecithin increased the effect of short-
ening in bread. Since then, it has become widely recognized that lecithin imparts a
lubricating effect to dough and finished baked goods. Lecithin also has a positive
effect on lubricity in extruded products. It contributes to increased throughput and
decreased clean-out time. Lecithin improves product flow, and it does not have a
negative effect on density. In the production of extruded, fat-free pretzels, down
time for cleaning the cutting knives can be significantly reduced with the use of
0.50.75% (flour basis) de-oiled lecithin. De-oiled lecithin can also be used to
replace monoglycerides in some extruded breakfast cereals and special pastas (218).
De-oiled lecithin is also recommended to increase the lubricating properties of
reduced fat dough. It helps reduce stickiness for improved production yields and
reduce stress on pumps, belts, and motors.
Lecithin is included in several patents for reduced fat bakery products. In a
patent for a low-fat cereal-grain, food composition, lecithin is preferably included
when an egg-like substance is used. The lecithin, along with other emulsifier com-
ponents and gums, presumably functions to incorporate air, which would otherwise
be incorporated by fat (219). Another patent describes the production of reduced-
fat, low-fat, and no-fat baked goods, in which a substantial portion of the shortening
or fat is replaced with an emulsifier composition (220). The emulsifier composition

preferably contains 10% to 25% by weight of at least one type of lecithin. Specific
reduced-fat products that can be made with the emulsifier composition include fer-
mented crackers, unfermented crackers, cookies, and brownies.
In a patent obtained by Gautchier and Dyer (221), a fat-sparing composition is
described that may be used to extend vegetable fat in cookie filler cremes and other
cremes. The patent states that lecithin may be advantageously used in that compo-
sition. In a similar patent, Abboud (222) used lecithin in a fat replacement compo-
sition for ready-to-spread frosting.
Instant foods. Lecithin has been used as a wetting agent and emulsifier in instant
foods. Foods including cocoa powder, instant drinks, instant cocoa and flavored cof-
fee, powdered protein drinks, coffee whiteners, instant puddings, cake mixes, and
instant toppings are widely employed applications for specific lecithins. The most
common method to incorporate lecithin is as an external coating on the powder par-
ticles. The particular lecithin to be employed largely depends on the hydrophilicity
or lipophilicity of the powder system (7).
In recent years, instant food manufacturers have become interested in oil-free
lecithins. These products are granular, but they can be fluidized by the addition
of water or fats. Additionally, fluid products are now available that are based on
oil-free lecithin and have obvious advantages, because of their blandness and hand-
ling properties, for instant food manufacturers (223).
Pan and food release agents. Lecithin-based release agents are employed in
many applications such as frozen waffle manufacture, bakery products, pizza bak-
ing, and pasta products. Most industrial griddle frying fats are formulated with
lecithin, solely for its release functionalities (7). The products may be spray- or
brush-applied to achieve a thin film capable of promoting easy release of baked
items from pans and belts. Bakery release agents may contain 26% lecithin in a
variety of oil bases, and they may also be formulated with particulate matter to pro-
vide an additional mechanical release.
The simplest of all food release agents are found in the category of pan oils or
griddle greases. These products contain low levels of lecithin (0.51.0%). Most of
the release action is provided by the mechanical barrier established by liberally
coating the cooking surface.
Retail release agents for home use are marketed as aerosols and occasionally as
pump sprays. In this setting, the release agent will be used for everything from sim-
ple release tasks like pancakes and fried eggs to more challenging systems like
cakes or muffins. Consequently, retail products are formulated with relatively
high levels of lecithin to provide extra release for difficult applications.
As stated previously, moderately and highly acetylated lecithins exhibit heat-
resistant properties that are very desirable to have in many release agent applica-
tions (173). A natural crude lecithin is subject to thermally induced reactions
that are responsible for the darkening and formation of insolubles that occur after
prolonged heating. There are several viscosity grades of heat-resistant lecithins
available, and lecithin viscosity varies with temperature. Low-viscosity lecithins
can be easily sprayed without dilution, or prepared as part of a spray release system.
412 LECITHINS
Typical formulas for lecithin-based release systems are pan bread: oil 98%,
lecithin 2%; aerosol spray: oil 70%, lecithin 8%, propellant 22%.
Release formulas for cakes, cookies, and other difficult specialty products often
include 515% lecithin, 110% particulates (flour, silica, etc.), and various types of
oils (mineral, vegetable, etc.).
Lecithins can be directly applied to the surfaces of griddles, continuous oven
conveyers, flame broiling equipment, and other cooking surfaces for better release
and ease of cleaning. The thinnest layer possible should be used for surface release.
Lecithin prepared according to U.S. Pat. 4,479,977 (173) is a very effective
release agent when applied to a surface in a very thin film, or used in spray pan
release systems. For use on a grill, grinders, extruders, pans, or skewers, spray
coverage should be applied with lower misting rather than with an air sprayer.
Continuous, multipurpose ovens that are used to precook foods may use water-
filled dip tanks for cleaning and rinsing the conveyor belt. An aqueous release sys-
tem, containing a water-dispersible lecithin, is added to the dip tank to facilitate
release of the food from the oven belt, as well as promote better rinsing and clean-
ing during cooking. A 10% aqueous dispersion of lecithin is commonly used for
this application (224).
An alternative to the use of dip tanks for continuous band ovens is to blend 2
10% heat-resistant lecithin in liquid oil or melted shortening, and spray-apply the
blend to the conveyor belt with an air spray system (224).
Refined fluid lecithins are also used to prevent the sticking of high-moisture
sliced and shredded products like cheese. Specialty fluid lecithin products are
sprayed or wiped onto sheets of processed cheese prior to slicing and stacking.
Effective separation of the cheese product requires an even, very thin distribution
of a low-viscosity lecithin applied as a fine mist to the moving sheet of processed
cheese. Lecithin also works well in separating the slices of certain natural cheeses
where the manufacturing process allows the lecithin to be applied as the cheese is
sliced. For separating cheese slices and shredded cheese products, 1 kg for approxi-
mately 500 m2 (equivalent to 45,000 slices) of a low-viscosity, sprayable lecithin is
used (224).
Dairy-type foods. Another major application for lecithin products is in dairy and
imitation dairy products. Bily (225) has shown that the addition of de-oiled lecithin
to milk during cheese manufacture resulted in an increase of mozzarella cheese
yield by 3.58%, and an increase in cottage cheese yield by 8.90%, primarily by
increasing the moisture content of the finished cheeses. Turcot et al. (226), con-
firmed that as the phospholipid content of buttermilk increased, the moisture con-
tent of low-fat cheese increased in spite of cheese manufacturing modifications.
Drake et al. (227) showed that de-oiled lecithin improved process cheese texture
without negatively affecting flavor or acceptance. Trained sensory panelists deter-
mined that reduced-fat cheeses containing lecithin were more similar in texture
attributes to full-fat control cheeses than reduced-fat cheeses without lecithin.
Sipahioglu et al. (228) demonstrated that yield loss and the increase in hardness
of feta cheese associated with fat reduction, was overcome by the water absorption
capacity of starch and lecithin. The presence of lecithin in low-fat and reduced-fat
cheeses improved the yield, primarily by increasing moisture, and reduced the hard-
ness, while improving the flavor and texture.
Lecithin is used to improve the wettability and dispersibility of various milk
powders including whole milk powder (229, 230) and caseinates (231). Oldfield
et al. (232) demonstrated that lecithinated whole milk powder had increased coffee
stability, with decreased coffee sediment levels over a water hardness range of
0308 mg/L.
Enzyme-hydrolyzed lecithin has been shown to improve the heat stability of
recombined milk products (233, 234), and almost all infant formulas contain either
hydrophilic or de-oiled lecithins as fat emulsifiers (235). Other dairy applications in
which lecithins are used include frozen desserts, whipped toppings, and yogurt.
Processed meats. De-oiled lecithin is used as a key ingredient for the emulsifica-
tion of fat in canned or frozen meat-containing products. When properly formu-
lated, the lecithin can dramatically reduce or eliminate fat-capping in products
such as canned chili, sloppy joes, gravies, Mexican meat fillings, and other meat
products containing a sauce or gravy (236). A combination of lecithin, textured
soy concentrate, and a starch has been recommended for reducing fat separation
in canned meat products that are cooked in the can (237).
Oil-free lecithins, or lecithin/distilled monoglyceride combinations, have been
recommended for reducing water purge in frankfurter formulations that contain
high levels of water (238, 239). Although the lecithin is functional in low-fat or
high water frankfurter formulations, it has failed to function in some full fat-
containing emulsion meats (240). It was concluded that the presence of de-oiled
lecithin in comminuted pork emulsions contributed to the destabilization of the
emulsions. This destabilization of meat emulsions by higher-HLB surfactants
was also seen by Cheong and Fischer (241).
U.S. Patent 4,434,187 (242) covers a meat curing composition that contains de-
oiled, powdered lecithin. The purpose of lecithin is this application is to prevent
separation of the brine solution. Lecithin is also used in release agents for meat cas-
ings and nets. Hammer, et al. (243) recommended that an aqueous-based release
agent for cellulose sausage casings, which contained 512% lecithin, be applied
at a rate of 450800 mg/m2, based on the lecithin weight.
Egg replacers. Lecithins are used in conjunction with dairy and vegetable pro-
teins in an attempt to functionally mimic the lipoprotein complex of egg yolks. A
coagulable egg replacer based on whey protein, polyunsaturated fat, and lecithin
has been described (31). Another formulation included soy and wheat flour blended
with oil, lecithin, carrageenan, and polysorbate 60 to replace up to 75% dry or
liquid eggs in a variety of mixes and prepared foods (31). Dashiell (31) also
reported on a lipoprotein complex formed from soy isolate, oil, carbohydrate,
and various emulsifiers, which is claimed to be useful for whole or partial replace-
ment of egg yolks in baked goods.
Nutritional and health-related applications. Lecithin has long been known in the
worldwide nutritional community. Dietary supplement lecithin is generally derived
from soybean lecithin. Because of its composition of various phospholipids, vita-
mins, and fatty acids, lecithin is involved in numerous physiological actions that
414 LECITHINS
TABLE 26. Basic Physiological Functions of Lecithin/Choline.
Function Description
Cell membranes PC provides structural stability for cell membranes, provides a

reserve supply of choline, and acts as a second messenger.
Fat transport PC is a main constituent of the membrane surrounding the fat
transporting molecules called lipoproteins.
Methyl metabolism Choline is one of a few B Vitamins that participate in methyl group
metabolism. Methyl groups (CH3) are components of numerous
important biological compounds.
Cholinergic Choline is a principal component in the synthesis of acetylcholine.
neurotransmission
range from the molecular to the organ level. Among the phospholipid composition
of lecithin, PC is considered the most nutritionally significant. Indeed, choline
phospholipids are involved in a myriad of essential metabolic reactions, are impor-
tant structural components of cell membranes, and are important mediators and
modulators of transmembrane signaling. Table 26 provides an overview of the
physiological functions of lecithin/choline.
Lecithin is the main dietary source of choline. The U.S. National Academy of
Sciences (NAS) Food and Nutrition Board recently underscored the importance of
lecithin in human nutrition by assigning choline (the vitamin component of PC) a
dietary reference intake (DRI) in 1998 (244). Although it is true that lecithin and
choline are present in a variety of foods, the specific amount of lecithin and choline
in foods is currently unknown. At this time, no comprehensive analysis of choline
in the food supply exists. Such an analysis is, however, underway and it is estimated
that the extensive USDA database of foods will be fully analyzed for choline con-
tent by 2005. The analysis will include all forms of choline found in food.
It is known that the richest sources of lecithin/choline are high-fat/saturated fat/
cholesterol-containing foods such as egg yolks, organ meats (liver, kidney), and
whole milk. Analyses from the 1970s estimated choline intake to be 7301040
mg/day in adults consuming a typical western diet (244). However, since the
1970s, dietary recommendations have strongly advocated reducing intake of foods
high in fat, saturated fat, and cholesterol. Although this dietary guideline policy has
resulted in lower fat intake for many Americans, it is likely that lecithin intake
has also decreased considerably in recent years. Indeed, a number of sources
have reported steep declines in consumption of eggs, red meat, and whole milk
over the past 25 years (245).
Choline has been shown to be essential to the body. In a landmark study in 1991,
Zeisel et al. (246) showed that healthy men with normal folate and Vitamin B12
status fed a diet deficient in choline have diminished plasma choline and PC con-
centrations and subsequently developed liver damage. In other words, when other
nutrients are adequate, the body is not able to produce choline in quantities suffi-
cient to prevent liver damage as assessed by elevated serum levels of alanine-ami-
notransferase (ALT), a critical liver enzyme. These data served as the supporting
TABLE 27. DRI for Choline (244).
Adequate Intake Adequate Intake

Level for Males Level for Females Upper Limit
Group Age (mg/day) (mg/day) (g/day)
Infants 06 months 125 125 Not possible to

establish
712 months 150 150 Not possible to
establish
Children 13 years 200 200 1
48 years 250 250 1
913 years 375 375 2
1418 years 550 400 3
Adults 19 years 550 425 3.5
Pregnancy 450 3.5
Lactation 550 3.5
evidence for the establishment of the DRI levels for choline. The adequate intake
level for choline was determined as the point at which ALT levels returned to the
normal range. The adequate intake level was originally set for adult males and has
been calculated/extrapolated for other populations. Tolerable upper limits for cho-
line have been set for various age groups. High doses of choline, in the form of
choline salts (choline chloride or choline bitartrate), have been associated with
sweating, salivation, a fishy body odor, and hypotension (low blood pressure).
Table 27 shows the AI and UI values for various populations (244).
Although healthy individuals eating an omnivorous diet are not likely to be at
risk of choline deficiency, some groups such as vegetarians, athletes, dieters, and
pregnant and lactating mothers can deplete choline stores. Lowered serum choline
concentrations have major consequences including hepatic, renal, vascular, neuro-
nal, and infertility problems (5, 247). Further, there is evidence that supplemental
choline (intake levels >DRI levels) may be beneficial to the health of some indivi-
duals. The following sections discuss the role of lecithin and choline health and dis-
ease prevention.
Liver health. As noted above, a biomarker of choline deficiency is elevated ser-
um ALT levels, which is an indication of liver damage. One of the many functions
of the liver is its role in fat metabolism. Without PC, the liver is unable to synthe-
size lipoproteins. Of particular importance in liver is the synthesis of very low-
density lipoproteins (VLDL). With diminished VLDL production, the liver is not
able to export lipid. This results in an accumulation of fat in the liver. Lipid accu-
mulation in the liver leads to various stages of liver disease such as liver cell death,
fibrosis, cirrhosis, and liver cancer (248250). The role of choline in liver disease
was underscored in the early 1990s when it was determined that patients on
extended total parental nutrition (TPN) treatment developed fatty livers (251).
At that time, TPN formulas did not include choline. Adding choline (in the form
of lecithin) to TPN formulas reversed fatty buildup in these patients, and a
416 LECITHINS
long-term study strongly suggests choline is an essential nutrient during long-term

TPN (252). This finding led to the series of experiments that eventually determined
the essentiality of choline, and to the choline DRI values.
Because of the link between lecithin/choline and liver health, these substances
are being studied for their therapeutic potential. In humans, the primary cause of
fatty liver is overconsumption of alcoholic beverages. Results from several studies
indicate that the phospholipids found in lecithin, particularly PC, may significantly
reduce liver damage caused by alcohol consumption (253, 254), and this effect was
not seen when choline was supplemented in the form of choline salts. Interestingly,
an editorial on diet and liver disease stated that the most exciting finding regarding
diet and liver health is that supplementation with PC protects the liver from the
damage associated with long-term alcohol consumption (255).
Other studies in animal models show that a choline-deficient diet promotes liver
carcinogenesis (256261). In fact, choline is the only known nutrient for which
deficiency is directly linked to liver cancer in the absence of any known carcinogen
(262, 263). Choline deficiency is therefore considered to have both cancer initiating
and cancer promoting activities.
Effects on blood cholesterol. Many studies conducted from the 1960s through the
1980s investigated the relationship between lecithin administration and serum cho-
lesterol concentrations. In 1989, Knuiman et al. reviewed 24 of these studies (264)
and concluded that most of lecithins cholesterol-lowering benefits were caused by
its content of polyunsaturated fatty acids, primarily linoleic acid. They stated that
soy lecithin, therefore, was no better than soybean oil in lowering total serum cho-
lesterol. However, this research group failed to consider the effects

Bailey's Industrial Oil and Fat Products, 6 Volume Set

Uploaded by

Copyright:

Available Formats

Bailey's Industrial Oil and Fat Products, 6 Volume Set

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bailey's Industrial Oil and Fat Products, 6 Volume Set

Uploaded by

Copyright:

Available Formats

BAILEYS INDUSTRIAL

OIL AND FAT

Baileys Industrial Oil and Fat Products is available online at

A John Wiley & Sons, Inc., Publication

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Library of Congress Cataloging-in-Publication Data:

Printed in the United States of America

R. G. ACKMAN: Canadian Institute of Fisheries Technology, Dalhousie University,

DAVID D. KITTS: University of British Columbia, Vanuouver, British Columbia,

HARIKUMAR NAIR: Bioriginal Food & Science Corp., Saskatoon, Saskatchewan,

Flavor Components of Fats and Oils; Lipid Oxidation: Measurement Methods;

JAMES P. WYNN: Martek Biosciences Corporation, Columbia, Maryland, Oils from

Volume 1. Edible Oil and Fat Products: Chemistry,

1.3 Polymorphism in Fats and Oils ............................................ 1:77

1.5.6 Patterns and Trends in the Production and

1.8 Lipid Oxidation: Measurement Methods .............................. 1:357

1.10.8 Selection and Training of Panelists .................. 1:422

1.12.3 Natural Antioxidants ........................................ 1:503

1.14.11 Adulteration ..................................................... 1:574

Volume 2. Edible Oil and Fat Products: Edible Oils.. ......... 2: I

2.2.9 Nonfood Uses of Standard Canola Oil ............. 2:109

2.5.5 Regulatory Considerations: Cottonseed Oil

2.8.4 Refining and Fractionation ............................... 2:371

2.10.10 Bleaching, Hydrogenation and

2.13.3 Physical Properties of Soybean Oil .................. 2:584

Volume 3. Edible Oil and Fat Products: Specialty

3.4.6 Cancer ............................................................. 3:87

3.6.7 Improved Oil Composition of a Transgenic

References .................................................................... 3:227

3.11.3 Fish Oil Fatty Acids and Gas-liquid

References .................................................................... 3:440

3.16.7 Effects Related to Triacylglycerol

Volume 4. Edible Oil and Fat Products: Products

4.1.3 Applications of Frying Oil ................................. 4:3

4.2.8 Low-calorie Spreads ........................................ 4:68

4.5.7 Non-lauric CBS ................................................ 4:170

4.7.7 Pie Crust, Biscuits ........................................... 4:219

4.9.15 Criteria for Fryer Selection ............................... 4:294

4.11.2 By-products of Oil Refining .............................. 4:405

Volume 5. Edible Oil and Fat Products: Processing

5.1.13 Deodorization and Physical Refining ............... 5:42

5.5 Packaging ............................................................................. 5:231

5.8.5 Commercial Deodorizer Systems .................... 5:376

5.12 Margarine Processing Plants and Equipment ..................... 5:459

Volume 6. Industrial and Nonedible Products from

6.1.2 The Role of Coconut Oil in the Oleochemical

References .................................................................... 6:100

6.4.14 Health and Safety Factors ............................... 6:176

6.7.5 Viscosities, Pour Points, and Oxidative

References .................................................................... 6:349

Index .................................................................................. I:1

Baileys Industrial Oil and Fat Products is available online at

A John Wiley & Sons, Inc., Publication

compounds. Changing perceptions of what is nutritionally desirable in fat-based

2. COMPOSITION AND STRUCTURE

2.1. Fatty Acids

Fatty acid Common name Formula Chain length

4:0 butyric CH3(CH2)2CO2H short

TABLE 1. (b) Occurrence.

( 2 PV AV) being a better index of oxidation than either PV or AV alone.